(Page 2 of 6 pages for this article  <  1 2 3 4 >  Last »)

Filed under: GentryMedia Sister Sites •ProVideo Coalition •

LIGHTING STRATEGIES: Rough Guide to Illuminating a Bounce Card

Filed under: GentryMedia Sister Sites •ProPhoto Coalition •ProVideo Coalition •Lighting •Production •Tips •Training •

LIGHTING STRATEGIES: Exploiting a Single Light Source

Filed under: Cameras •GentryMedia Sister Sites •ProVideo Coalition •Lighting •Tips •Training •

BOOK REVIEW: “How to Shoot Movies Without Shooting Yourself in the Foot”

Filed under: GentryMedia Sister Sites •ProVideo Coalition •Lighting •Production •Tips •Training •

LIGHTING STRATEGIES: Placing the Fill Light for Faces

There are only two lights that naturally come from the sky: sunlight and moonlight. Every other high-angle light source is artificial. There are an awful lot of artificial sources that are above eye level, but even more that are at or below eye level (table lamps, candles, etc.). There’s always some sort of ambience from the flat thing that the actors are standing on, typically the floor or ground. Soft light from below is one of those things that’s all around us that we never notice until it’s gone. Stand in a room with a black or dark floor and overhead lighting and you’ll notice very quickly that something is missing.

The same thing happens in sets where there’s no ceiling: there’s no ambient light reflected from the ceiling on hair, clothes and walls, and we can detect that something is wrong. That’s why it’s normal for us to put some sort of white material over sets with no ceilings—typically a 4’x8’ piece of foam core or two tossed on top of the grid—to recreate the ambience that results from a white ceiling in a normal room.

Soft light from below can result in upward shadows, particularly hands on faces, but if you’re shooting in a sunlit room that’s probably fine. In fact, it probably looks better than fine because you’re recreating what really happens in a sunlit room. It feels very natural, and almost nobody will be able to tell you why it works because very few people consciously notice the ambient soft light that permeates our world. They will, however, notice if it’s gone.

Light from below the lens creates great eye lights because the slope of the cheek is much less than the slope of the brow. A light from above may be blocked by the brow and never reach into dark eye sockets, but there’s not much to block a light from below. Eye lights are usually distinct from fill lights as they tend to be harder point sources that create a very distinct glint in the eye. An old eye light trick involved removing the lens from an Inky (150w fresnel light) and putting a lot of scrims in front of it, so the Inky became a bright point source that didn’t cast much light into the scene but reflected wonderfully in eyes. These lights were often placed either to one side of the lens or directly below the matte box.

This might be the time to mention that I generally fill from as close to the lens as possible. Fill light near the lens axis has a wonderful smoothing quality because it casts very tiny shadows. Skin imperfections become quite noticeable when lit from the side, but fill light from the lens eliminates or reduces these shadows and helps them to disappear.

Because light from near the lens axis casts smaller shadows it’s possible to get away with a physically smaller fill light closer to the lens when you can’t get away with a larger soft light farther away from the lens axis. Lights closer to the lens axis appear softer than they really are, because the shadows they cast are smaller from the lens’s perspective: most of the shadow is cast behind the object, away from the camera.

Here’s soft light from over the lens:

This is a great position when filling large spaces. It’s not unusual for me to fill a large room by putting a 4’x8’ piece of foam core behind and over the camera and up against a wall. Fill light from below is very pretty but that much upward-facing light isn’t appropriate for all circumstances, such as night interiors. Also, fill light from a distance behind the camera won’t cause actors to brighten considerably if they approach the camera, which would happen if the fill source was directly beneath the lens.

Notice that the fill light doesn’t have much of an impact in our virtual model’s eyes. The only time that might be a real concern is if there’s no key light reflection in them and you’re looking for an extra kick to make them come alive, in which case you could add an eye light.

There is one danger to this technique: if the fill source is too small or too high you may get a distracting downward chin shadow. Here’s an over-exaggeration of that effect:

Argh. I’m not a fan of multiple chin shadows. They look “lit” to me and I’m not a fan of the “obviously lit” look.

Here’s a great fill position we haven’t spoken about yet:

Here the fill light is immediately next to the lens on the side opposite the key. Light from next to the lens will always reach into eye sockets, because if the lens can see eyes then the eyes can see the light immediately next to the lens. The only downside is that, like most of the techniques we’ve spoken of here, there will be two tones on the face: a bright (key) side and a dark (fill) side.

There’s one more technique, though, that can be a lifesaver. This is where the power of fill light becomes really apparent. Turn the page so we can talk about filling from the key side…

Filed under: GentryMedia Sister Sites •ProVideo Coalition •Lighting •Production •Tips •Training •

LIGHTING STRATEGIES: What Makes Soft Lights Cast Soft Shadows?

Filed under: Cameras •GentryMedia Sister Sites •ProVideo Coalition •Production •Tips •

For You, a Panel Discussion

Filed under: GentryMedia Sister Sites •ProVideo Coalition •Lighting •Tips •Training •

LIGHTING STRATEGIES: Soft Light vs. Hard Light

I’m a big fan of soft light from below, but students often resist this approach because it is supposedly “horror” lighting. As part of a basic lighting demonstration lighting teachers often show that hard light from directly below a face looks like something you’d see in an old Hammer horror flick, and everyone dutifully takes note that light from below the lens denotes evil.

Uh, no.

When I was a camera assistant I worked a lot with a DP who lit a lot of scenes by putting foam core beneath the lens and bouncing light off of it. The result felt very natural and real, as if sunlight were coming through a window and bouncing off the floor.

I’ve gone through a lot of lighting phases in my career where I’ll try a style out for a while and then move on to something else. Right now I’m lighting a lot of shots with soft low light because, to me, it feels real and unlit, whereas light from above can feel artifical or predictable. I’m not a slave to this style and I do frequently light with sources placed above the lens height, but I do find a lot of situations where soft light from below is more interesting and believeable than lighting from above.

Here’s an example:

This guy could be sitting at a table in a restaurant, near a window where sunlight is coming in and pounding the table in front of him. He could also be sitting at a dinner table where an overhead light is bouncing off the table cloth. Back when I shot mostly corporate videos I found that the fastest way to light the traditional “We’re having a corporate meeting” setup was to hang a light over the conference table and aim it straight down onto the tabletop. I’d scatter some papers around to bounce the light onto the people, and then I could put the camera anywhere and get a decent-looking and realistic shot of anyone around the table without having to relight each setup or avoid shooting my lights. The tabletop looked a bit bright but people will buy that look in that environment.

Same thing here: maybe she’s lit by sunlight hitting a piece of paper she’s holding in front of her, or light is bouncing off the shirt of someone standing in front of her. There’s a famous scene in the film Peggy Sue Got Married where a nighttime basement scene is lit by one light coming through a window. It falls directly onto one of the characters in the scene but the other character is lit only by the reflected light bouncing off the first character’s sweater. When the characters move apart one of them recedes into darkness as the bright sweater moves away.

That’s one of the things that I love about soft sources: they can create a feeling of space. I’m not talking about massive sources that are used as shadowless fill or as a base light for a set; I’m talking about things like sunlight hitting a table or a character walking past a lamp with a lampshade. The soft vs. hard quality and brightness of a nearby soft light will change depending on how people and objects move in relation to it. If a room is lit by a single shaft of sunlight hitting the floor and a character is moving around the room outside of that shaft, the light falling on that character from the hot spot on the floor is going to change as they move around the room. The farther they are from the hot spot the less directionality there will be, but as they get close to the hotspot the bounced light will become brighter and much more directional. If they cross from one side of the room to the other they’ll go from being frontlit to being backlit or unlit.

Soft sources are a great way to define spaces. I call this kind of lighting “volumetric lighting” because it helps the audience feel the volume of a space. Hard sources tend not to do this so much as their character changes less as you move around them: they always cast sharp shadows, so the character may become brighter or darker and the shadows may shift as they move within the light’s beam, but there’s less feeling of movement in relation to the light source.

Let’s pretend our subjects have walked away from the sunlit surfacing they were sitting at:

Imagine this look happening in the same shot as the previous look: this is position two, and the previous images were position one. Can you get a sense of where these people are in relation to the lit surface they were sitting over a minute ago? Do you have a sense of where they are in relation to that surface?

Here’s what happens if we leave the light where it is and make it a small source:

That’s a very, very different look. You can get some sense of where they are in relation to the light source, but that’s a very harsh and artifical light source. That may be appropriate for your scene, and if so—use it!

As always, the ideas I’m expressing are guidelines only. I’m not dictating how you should light, only showing you the possibilities and how I interpret them artistically. Experiment, decide what you like, and then incorporate what you’ve learned into your own style.

Art Adams is a DP with a soft light touch. His website is at www.artadamsdp.com.

Filed under: GentryMedia Sister Sites •ProVideo Coalition •Lighting •Production •Tips •

Pulse Width Modulation is NOT Your Friend

Filed under: GentryMedia Sister Sites •ProVideo Coalition •Lighting •Production •Tips •Training •

LIGHTING STRATEGIES: Placing a Hard Key Light

This is classic “Rembrandt” hard lighting. It’s not strictly “Rembrandt” style because the light is very hard: it casts sharp shadows and doesn’t “wrap” around surface details. The nose shadow does fall along the smile line and connect to the cheek shadow, which are both hallmarks of Rembrandt’s style.

This virtual face is very forgiving as its nose is fairly even and free of bumps and deformities. It’s also not very long. Hard light reveals distortions in nose structure quite easily, and as it is often our responsibility as DPs to make the performers look their best we must be aware that certain lighting setups may not work for everyone.

In this case, though, the classical hard light placement of 45-degrees up and 45-degrees to the side works well. Traditionalists might argue that this is an extreme way to light a woman’s face, and that the key is better placed closer to the camera:

This shorter nose shadow is considered kinder to women, especially those who are concerned about how their noses photograph. If you think such women don’t exist then you’ve never worked with Barbra Streisand. I haven’t, but I knew someone who did and they told me that she would only allow herself to be photographed from her left side. She has a long nose and it points slightly left, so by showing the camera the left side of her face her nose was always pointed into the lens instead of across it. That had the effect of making her nose look straighter than it really was.

A small nose shadow can make a nose look shorter than it is.

This is the kind of thing that endears you to actresses. Pay attention: you never know where your next job will come from, and if you take care of the people you photograph they may have the opportunity to take care of you.

This virtual model also has very shallow eye sockets. The deeper the eye sockets the lower the key light has to go to reach into them. This is not a requirement, only a comment regarding classical portraiture. We don’t always need eyes to be lit, but if an actor or actress has deep-set eyes and the director wants to see them then the lights will have to drop lower, or another light must be added.

The important thing is that there’s always a bit of shadow on the face. I can spot rookie directors very easily: they’re the ones that complain that “Her face isn’t evenly lit; one side is darker than the other!” Yes, that’s intentional: a face that’s lit dead-on doesn’t look three dimensional. That shadow creates depth, and the lightness or darkness of that shadow creates mood. People who complain about shadows on faces haven’t been paying visual attention to nearly every image they’ve ever seen in a motion picture.

Having said that, lighting “flat” can be very interesting. I’m going to save that for another article.

Bringing the light further around results in this:

This is still a nice look, and is very kind to someone who has nose structure issues. The closer the key light is to the camera the fewer wrinkles and bumps it will see as their cast shadows become shorter. (These can also be erased with fill light. More about that later.)

Last but not least, we can place the light directly over the lens:

This can be a very flattering look but it depends on the face. If the person has a wide face you’ll just emphasize its width, and that’s rarely pretty. Sidelight makes a face look thinner; front light can make a face look fatter.

Here are some things I hate:

I’m not a big fan of horizontal nose shadows. I’ve eased up on this opinion over the years as I’ve found situations where it’s not necessarily a bad thing, but a hard sideways nose shadow is more often than not a rookie mistake. Sometimes it works…still, pay attention to this, as it’s a very dramatic look that may not look flattering in a lot of situations.

The hard upward nose shadow can work in certain situations, such as if the actress is lit by a candle that she’s carrying, but there are a lot of situations where this just doesn’t do nice things to an actor’s face. Soft light from this angle can be very interesting, but we’ll get into that later. Save this look for serious drama involving actors with perfect faces.

How to adjust your lighting for men, and some tricks of the trade, can be found on the next page…

Filed under: Cameras •GentryMedia Sister Sites •ProVideo Coalition •Lighting •Production •Tips •Training •

The Simplest, Fastest Interview Lighting Setup—Ever.

Let’s look at that diagram again:

Here are two examples of the look that I get from this setup:

This was a three light setup: one light, bounced, on the subject, and two more on the background, which consisted of art glass.

This setup used a single light bounced off a 4’x’4’ piece of foam core, plus a negative fill card with a smaller piece of foam core clipped to its face. Camera for both setups was a Sony EX3.

These are screen grabs from a documentary I’ve contributed to over the last few years. The fill is a bit brighter than I normally like but that’s what was necessary to match looks established by other camera people.

There are certain magic numbers and ratios in the film industry, and 4’x4’ is one of them. The 4’x4’ bounce or diffusion frame is very common in the industry because it does beautiful things in close quarters, particularly to faces. When I first tried this setup, placing a 4’x4’ bounce card 2’ to 3’ from the subject and lighting it with a 650w fresnel, the results were exactly what I was looking for. The nose shadow was very, very soft and gentle, and it almost didn’t matter where it fell because it’s difficult to see. The source wraps beautifully around the average face and light reaches easily into both eyes. The reflection of the light source causes skin to appear to glow from within. This quality of light, from a large source at close distance, works well on almost everybody.

There were still a couple of problems to solve. The first one was contrast: soft light goes everywhere, and a small room with white walls reflected a lot of unwanted light back onto the subject, resulting in a very flatly-lit face. I had to add some contrast, so I introduced a 4’x4’ negative fill card.

“Negative fill” implies an active approach to removing light, whereas the reality is that you’re passively replacing highly reflective surfaces with something darker to eliminate stray light. Right now, for example, I’m sitting in front of my computer with a window to my right, and although I’m directly lit from one direction I’m filled from every other by light reflecting off walls, the white ceiling, and the light rug on the floor.

For example, here I am sitting at my computer and lit by natural light (shot on my iPhone):

Here’s what happens if I hold a large black card on the left side of my face, blocking ambient light from that side of the room:

The side of my face nearest the card is darker, as a reflective wall outside frame right has been blocked by a non-reflective surface.

Here’s what happens when I block ambient light from the ceiling:

Now the top of my face is darker as I’ve placed a dark surface between my face and the white ceiling.

In 3D computer animation terms this is called “radiosity,” and it was a big deal when it was introduced in the 1990’s. An algorithm examined all of the surfaces in a 3D model, looked at where the virtual light sources were placed, and added ambient light into the scene to show what the environment would really look like when every surface became a passive reflector. Architectural firms jumped on this technology so they could see, for example, what would happen in a white hallway if they installed red carpeting. (Result: sun hitting the carpet would turn the walls, and everything else in the environment, red.) The reason radiosity is so useful in 3D modeling is because this is what happens in the real world: every lit surface around you, at this moment, is lighting you to some degree.

One of your tasks as a cinematographer is to decide whether this is desirable or not. If not, you have to figure out how to eliminate it or replace it.

Unless I’m lighting an interview in a big room with lots of space between the subject and the nearest wall, I add a 4’x’4 black card on the fill side to reduce ambient light and increase contrast. Ambient light looks okay to the eye, but once we frame a shot it can take on a very different feel. It can be the wrong color, or fail to reach into eye sockets, or just look sloppy. (That’s why I found the old Dogme 95 movement so amusing: a lot of “natural light” looks awful once you put a frame around it. Light is a storytelling tool and should not be ignored.)

A really dark fill side, however, is not always desirable. My solution is to use the black card to remove the ambient light that I don’t like, and then use a smaller bounce source to add the ambient light that I do like. My fill source of choice on fast-paced corporate and documentary shoots is:

(wait for it)

Copy paper.

You can always find copy paper in an office, and in a pinch I’ve used my call sheet. It doesn’t matter if it has type on it, just as long as it’s primarily white and reflective. I’ll usually place one sheet on the black card, as far forward as possible and at head height, to create a nearly invisible fill. Putting it forward, closest to the camera, prevents the front of the face from falling into shadow: a bounce that’s placed toward the back of the card, on the side of the subject, will light the subject’s cheek and ear but will leave a very dark area around the fill-side eye. That’s usually not very flattering, so bringing the fill around the front of the face both eliminates that overly-dark eye shadow and hides the fill as a separate source as it no longer casts a noticeable shadow of its own.

If I want more fill, I tape another piece of copy paper next to the first sheet.

There are exceptions to this lighting method, as there are to everything. People with very round or flat faces may require the removal of fill light to create contrast, as the source wraps around their faces too much. A large source doesn’t work well on people with reflective glasses. And it really doesn’t work well with people who are crazily animated and sit forward into the beam of light that’s lighting the 4’x4’ bounce. None of this happens terribly often, and 90% of the time a person who sits into this lighting setup will look great.

There are a couple of things to watch for:

Foam core has a little bit of specularity to it. The light it reflects is soft, but there is a little bit of a hot spot that may cast a shadow. This can be remedied by covering the foam core’s shiny surface with a matte material, like muslin. Generally this hot spot isn’t a problem.

The light source usually wants to be a little higher than the average subject height:

While the nose shadow is very soft it’s not nonexistent, so raising the source throws it down a little bit into the smile line.

The black card serves two purposes. The first is negative fill, but the second is to cut direct light from the lamp itself off the subject. All lamps leak a bit, and stray unwanted light wandering through a set is something that I really, really hate. I always walk the set, if it’s large, or sit where the subject sits in an interview situation, and look around to see if I’m being struck by any unwanted light. In this setup there are two forms of unwanted light that I see most often: light from the glowing fresnel lens of the light, as seen through the gap between the barn doors and the instrument, or a reflection off the barn door farthest from the subject. The paint on black barn doors is shiny, and folding the far barn door the wrong way can mean catching the light from the fresnel lens and reflecting it directly onto the subject.

The best way to solve all of these problems is to back the lamp behind the negative fill card and use it as a flag, such that the subject can’t “see” the stray light. (If they can’t see it, they aren’t being lit by it.) If you can’t back the lamp far enough to hide the barn doors behind the negative fill it’s often enough to open that far barn door all the way, eliminating the reflection.

Backing the lamp behind the black card solves the issue of direct unwanted light on the subject. If this isn’t possible, make sure the gap between the lamp and the barn doors is wrapped with black wrap and open the barn door nearest the camera so the reflected light goes elsewhere.

Sometimes I add negative fill over the top of the subject, but not very often. Negative fill on the fill side is often more than enough unless the ceiling is very low and very reflective.

I’ll often light interview setups with two lights: one on the foreground and another on the background. This makes for very fast setup and breakdown. I don’t use back lights in interviews very often anymore, as I prefer the subtlety of placing the subject against a background of a different tone for separation, but this is a matter of taste. (And it’s a matter for another article, as back lighting is an art in itself.)

Turn the page and I’ll describe the basic interview tools I used for years when shooting hundreds of talking head interviews…

Filed under: GentryMedia Sister Sites •ProVideo Coalition •

The Future of Technology is You

Filed under: GentryMedia Sister Sites •ProVideo Coalition •Lighting •Production •Tips •Training •

Fill Light: The Underdog of Lighting

In film school we were taught that the fill light determines the density of the shadows. What we weren’t taught was how important shadows are, and that while the “key” light certainly has directionality, the fill light can as well. Where shadows fall is just as important as how bright they are, which means that we need to look at not just intensity (which I’m not really going to cover here; that’s determined by your own taste) but at placement as well.

Here’s a typical three-point lighting setup. This is both a great learning tool and an awful formula to follow, because while most lighting breaks down into this in one way or another it is severely limiting if this is all you know how to do. Still, it’s a good starting point.

Top view.

Side view.

NOTE:

In these examples I’m using a simple hard-source key light. I’m not advocating this approach in every situation; it’s just that I’m focusing on fill light in this article and key lights are a different and equally complex subject. The fill examples I give should work well in conjunction with any key source.

The reason the fill light is at the same height as the key probably stems from early studio cinematography when much of the lighting was placed high in a grid. Beauty lighting was often done from the floor (more on that later) but live television and grand feature film sets were almost always lit from the air.

In this scenario the key light is placed so that the nose shadow falls along the “smile line,” the line between the corner of the base of the nose and the edge of the mouth. The position of that light can make the nose shadow long (connecting to the cheek shadow for classical Rembrandt lighting) for “masculine” lighting or short when lighting for glamor. The fill light was set up opposite the key simply to fill in the shadows left by the key.

In the diagram above, where the key and fill light are the same size but differing intensities, the fill light will cast just as hard a shadow as the key light. The fill light’s shadow will be less obvious because it is less bright but it will still be present. There may be a dark area under the chin where neither light reaches but that can occasionally be helpful in hiding what a friend of mine calls “the gobbler,” pertaining to loose skin that collects under the chins of mature leading ladies.

PROS:

Simple, basic lighting suitable for use from an overhead grid or in a location with high ceilings.

CONS:

Very “theatrical” looking. Not very naturalistic.

Multiple shadows indicate competing light sources, possibly drawing attention to the lighting.

Possibility of two opposing nose shadows instead of one.

No guarantee of getting light into deep eye sockets.

This setup hides the fill shadow somewhat because the fill shadow is softer. Broadening the fill light is always a good idea because a soft shadow will be much less apparent when competing with a stronger key light. Our eyes are sensitive to brightness over everything else, and seeing two sharp shadows from a key and a fill light can be distracting even if the fill shadow is much dimmer. By reducing the edge contrast of one of the lights we can better conceal it.

PROS:

Simple, basic lighting suitable for use from an overhead grid or in a location with high ceilings.

Fill shadow is better hidden, which feels more “natural” and less obviously “lit.”

CONS:

Can still look “lit” under certain conditions. The bigger the fill source, and the softer the fill shadows are, the less likely this will be.

Still may not reach into deep eye sockets.

A variation on this technique is to lower the fill light and put it at an opposing angle to the key light. This is an older method of lighting, from the old studio days when it was more important to make actors look pretty than it was to make the lighting look natural. This can be a very glamorous look but can result in dual nose shadows and a weird upward-cast shadow on walls or objects behind the actors if the light is too hard.

PROS:

Can be very pretty.

Puts a nice catch light in eyes. (Lights from below or near the lens are more likely to reach into eye sockets than lights placed high.)

CONS:

Not a very naturalistic look.

Upward-cast shadows on walls and near objects, which become more acceptable if the source is broadened/softened.

Bounced fill.

Diffused fill.

Putting a broad fill light directly over the camera hides the fill shadow almost completely. The shadow falls downward, behind the actor, and is largely hidden from view unless we can see the floor in the frame. We may see some shadows on the fill side if the actor waves their hands around in front of their body, but such shadows are usually minimal. If the source is large enough then fill shadows will be neglible. Any shadows that might appear are cast downward, which often feels natural. There are lots of light sources in our environment that cast soft downward shadows. (White ceilings come to mind.)

Lighting over the len axis illuminates just about everything the camera can see until the light intensity diminishes naturally over distance. As the light is illuminating the scene from very near the lens axis it creates a smooth look, because facial imperfections such as wrinkles and dimples have nowhere to cast their shadows. If the fill light came from the side then those facial features might become more visible through casting both a key light shadow and a fill shadow; but light from near the lens axis erases those shadows nicely.

Filling from near the lens axis goes a long way toward erasing skin imperfections and creating or emphasizing the feel of healthy skin.

Lighting from the lens axis is a very powerful concept. Imagine that you could put a light in the back of the lens and illuminate everything the lens can see. The scene would lack a certain amount of dimension because there would be no visible shadows from the perspective of the lens. Depth could be perceived through the gradual fall off of the light deep into the set, but that’s about it.

What you gain, though, is very interesting:

Anything reflective will have a highlight in it. This includes eyes. If a light is near the lens axis then it will reach into any eye in the frame. This is why the “Obie” light was so popular in earlier days: by mounting a small light source near the lens the actors were guaranteed a catch light in their eyes at all times.

That highlight will always face the lens and be in the center of a round object. This can create a sense of roundness in a 2D image, particularly in faces.

The only place a fill light over the lens won’t reach is a surface that faces downward that the lens can see but that the fill source can’t because it is higher than the lens. For this reason it’s a good idea to keep the source lower than higher, although the farther back the fill source is the higher that can be. Here’s why:

Filling a small area vs. filling a large area: a fill source that is farther away from the actor will reach a larger area, while a fill source closer to the actor is limited in what it illuminates. Note that while the height of the source changes based on distance, the angle is roughly the same.

For this reason I almost always fill from as close to the lens axis as possible. If I’m lighting a big space the source will be above and behind the camera; for a smaller space, or for an actor’s close up, the light will be very close to the matte box.

That covers the basics. Turn the page for some advanced tricks…

Filed under: Cameras •compression •Distribution •GentryMedia Sister Sites •HDSLR •ProPhoto Coalition •ProVideo Coalition •Lighting •Production •Software •Tips •

Blue Nile Shines Thanks to the Canon 5D and Apple Color

The first trick was to find a nice restaurant to shoot in, something that shouldn’t be hard to do in San Francisco. When I first met director David Fine and producer Kimi Milo at Piperade Restaurant, near SF’s financial district, I couldn’t believe my eyes: the place was very nicely decorated, it was well laid out… and it was all ours. The management was very amenable to letting us do just about whatever we wanted.

The interior of Piperade, shot during the location scout. We blacked the windows out on the shoot day and also removed the hanging bottle sculpture on frame right. The podium on the left had to stay in place so we shot around it.

“How did you find this place?” I asked Kimi. “The San Francisco Film Commission? A profressional location scout?”

“Nope,” she said. “I thought it was a nice place so I asked if we could shoot here. They said yes.”

As a very successful businessman once told me, “If you don’t ask, the answer is automatically ‘no.’ If you ask the chances of a “yes” answer increase dramatically.”

Director David Fine has a relaxed intra-table conversation with production manager Kimi Milo. On the shoot day I left the track lighting on, but dimmed way down, for the wide shots. I wanted to see the lights glimmer out of focus in the background but I didn’t want them to actually light anything.

It was pretty clear that we had to shoot away from the front windows to take advantage of the restaurant’s depth. Cross-shooting would force us to look at closer backgrounds, but since the focal length would be longer for those shots the backgrounds would drift nicely out of focus. The trick was to find a way to quickly light a moody night restaurant interior with a small crew and still make our day.

Originally Seedwell asked me if I wanted to shoot this spot on a Sony F3, but I opted instead to use their Canon 5D. The F3 is a much easier camera to work with from an operator/assistant standpoint, but as a DP I like the quality of blown-out highlights on the 5D more than I do on the F3. Moire can be an issue, but I’ve become very good at spotting it on the camera’s LCD monitor and I don’t run across it nearly as often as I thought I would when I discovered the issue years ago.

Using the 5D made focus a bit more difficult for my assistant but it freed up some more money for lighting and grip, which is where I really wanted to spend it.

I mentally toyed with a couple of different lighting setups. The fast method would have been to stage a line of Kino Flos behind the bar and create a big, broad light source that would have been very pretty. What it wouldn’t give me is the foreground/background contrast that I saw in my head when I imagined a moody restaurant interior at night. I really like finding ways to increase contrast between the foreground and the background and the fastest method I’ve found so far is simply to get the lights close to the action so that they drop off quickly. This isn’t always optimal as the lights start to block in the actors, and they often have to be moved around on a shot-by-shot basis, but when working in a small location with a low ceiling and a small crew that’s really the only way to do it.

Yes, that’s an actual slate. For a long time everyone tried to use iPad slates but the basic wood-and-plastic model saves a lot of time. It’s always a good idea to record double-system sound with the 5D as the audio recording is the 5D’s biggest weakness, primarily because of the audio-in connector.

The biggest pain for the wide shot was finding a place for the fill light. The brunette actress was easy to light: outside frame right we put an open-face tungsten 2K through a medium Chimera (to reduce spill) and then put that through a frame of Lee 129. I’d have loved to do the same thing to the other actress but there was no way to hide a large source in the shot. As the camera never came around in front of the blonde actress we hid a 1K fresnel behind the vertical wooden beams on the very left of frame. That edged her face enough that it looked lit, and the hardness of the light was hidden because we never saw the shadows it cast.

I love Lee 129. It’s denser than 216 or grid cloth. I love that the diffusion fully becomes the light source without any hint of specularity. Specular (or “hard”) light has its place, but when lighting beautiful women for digital capture soft light seems to work best on faces. I love how the reflection of a large source in skin makes it glow.

I tried putting the fill light, a 4’x8’ bead board lit with an open-faced 2K, directly over the camera but it just didn’t look right, which is odd as that trick almost always works. The way the light fell off past the hero table made the environment feel obviously lit. In the end we put the fill over the right side of the camera and wasted some of the light off the foreground table. There’s a faint hair shadow from the fill light on the blonde actress’s face, but for a quick wide shot it worked fine.

We executed the dolly move on a 5’ Kessler HDSLR slider. We couldn’t actually move the camera over the foreground table the same way the original commercial’s crew did, but by running the slider diagonally right next to the table and panning the camera during the move we managed to create a shot that was comparable in feel.

Key grip Ernie Kunze stabilizes the slider.

To light the rest of the restaurant we backlit the other diners by placing an open-face 2K in a doorway at the right rear corner of the restaurant. We diffused it with a frame of Lee 250, which spreads the beam but retains a little bit of specularity, and we put some color on it, probably minus green or a salmon color. I’ve been using reddish back or fill light in restaurant locations a lot recently; for some emotional reason I equate dim red lighting with fine dining.

As an accent we put a 4’x4’ daylight-balanced Kino Flo on the floor in front of the back wall, and we put more daylight Kinos behind the bar to add a bit of color contrast. There’s a saying in interior design, “Warm colors advance and cool colors recede,” and it seemed right to use cooler lighting in the background when it was motivated.

One or two tables in the background had tweenies armed over them to put a splash of light into the middle of the tablecloth. This filled faces a little and created occasional splashes of hot light around the room. If we hadn’t done that the background would still have looked okay but it might have felt a little “muddier.” It’s nice to have highlights scattered amongst the shadows as it makes the shadows feel darker and more intentional. It also creates what I call “volumetric” lighting, where the bounce light from the tables lights only the people around each table, creating a sense of volume and depth in the shot.

There are a couple of practical downlights raking the brick wall that we left on and dimmed down. We also dimmed the track lighting way down for the wide shot, just to add some out-of-focus highlights.

The final light that we placed was a 3200K Kino Flo behind the hero table, placed on apple boxes so that it was just below the back edge of the tabletop from the camera’s perspective. This created hot but soft highlights on the underside of faces, as if a light was bouncing off the table cloth or the floor. This worked well for the tug-of-war sequence also: this spot was all about a diamond ring, so I intentionally lit arms and hands to focus the audience’s attention on the jewelry.

This shot used nearly the same lighting as the wide shot, except that the hard light on the blonde was swapped out for another open-face 2K through a 2’x3’ frame of Lee 250 and a 4’x4’ frame of Lee 129. (We only had the one medium Chimera, so we duplicated the look using multiple frames and a few more flags.) Both lights are to the sides and slightly behind the talent so that the soft sources reflect in each actress’s skin and create smooth, beautiful modeling. Each key light is also acting as a back light for the actress closest to it. You can see the rear daylight Kino casting some blue on the rear wall as an accent.

Notice how the Kino Flo hidden behind and below the table is drawing attention to the talents’ arms and hands. 3200K Kinos look a little cool compared to 3000K tungsten lamps, and I used that as a bit of color contrast.

Soft light from below almost always feels natural to me. Both sunlight and lamp light turn floors into light sources, and it’s nice to have a light source that’s beautiful and doesn’t come from above the frame. Light from overhead can feel a little too obvious.

For this setup the 4’x4’ Lee 129 frame is placed just off frame left. For fill we used a second frame of Lee 129 placed directly next to the first frame, basically over the head of the brunette actress, with an open-face 650w tungsten unit through it. Instead of creating a separate fill source we really just extended the key source by making it wider and dimmer. The light wraps beautifully around the actress’s face and then drops off to darkness. I find this look so much more interesting than simply making one side of her face bright and the other side less bright, which is what happens with a traditional key+fill lighting setup. Two opposing lights mean two opposing tones, whereas a big soft source that wraps results in many, many more tones.

The backlight on her hair is a 3200K 4’x4’ Kino Flo, oriented vertically—probably with 1/4 CTO on it to bring the color in line with the 3000K tungsten lamps. The lamp is vertical for two reasons: (1) so it doesn’t wrap around her face from behind and highlight the tip of her nose, as it might if it was oriented horizontally; and (2) to create a uniform highlight in her hair. Hair catches backlight because it’s shiny, and if I were to use a smaller light I’d end up with a small highlight. Imagine every strand of hair as a round shiny mirror: I’m simply putting a large source behind her to reflect in the full length of hundreds of long, small mirrors.

Director David Fine (foreground) guides us toward our comedic destiny. In the background you can see: the key light just over the monitor (a 2K open face lamp through a 2’x3’ frame of 250 aimed into a 4’x4’ frame of Lee 129), the fill light (immediately to the right of the key and consistenting of an Arri open-face 650w unit aimed through Lee 129), and a Kino Flo scratch light on the right of the table. I don’t remember what the Dedo light was for, I think that was left over from the wide shot. David is watching the action on a Flanders Scientific 1760W monitor, which is my current favorite field monitor.

Same lighting setup, only reversed. The glows on the back wall are from practicals in the restaurant ceiling, and the highlight on the tablecloth at the rear left is a tweenie rigged overhead and dimmed down.

Same idea, only closer. That frame of Lee 129 is creating a wonderful glow in her skin. The one thing I regret here is not adding an eyelight so her eyes didn’t go completely dark when she looked down. I haven’t been a fan of dedicated eye lights but I may start moving my lighting in that direction.

The Canon 5D on a Sachtler Video 20p head, in turn mounted on a 5’ Kessler HDSLR slider. We did a couple of nice slow moves during this setup where I slid back and forth under the fill light (the frame of Lee 129 over the camera). The key is hidden behind the fill frame, although you can see the C-stand holding the 4’x4’ frame just to the right of the camera. Also on the right side of the frame, in the deep background, is the 2K and Lee 250 diffusion combo that’s backlighting the rest of the restaurant.

I used some filtration in front of the lens—a Tiffen Soft FX 1/4 + 1/4 Black Promist combo—and while it looked quite noticeable on set it turned out to be very, very subtle when we saw the footage in post. There’s something about the Canon 5D’s HDMI output that exaggerates both diffusion and highlights. I’ve noticed on several shoots that highlights that appeared burned out on set don’t look that way at all when viewing the footage on anything other than the camera. It’s very odd.

I shot this spot at ISO 400, T2.8, and a 1/50th shutter. ISO 400 is what I was used to back when I shot film, and it’s a nice compromise between keeping light levels fairly low and reducing image noise. I’ve shot the Canon 5D at ISO 800 before and it works well in a pinch, but I prefer the cleaner look that I get from lower ISO’s. If we’d had any HMI’s on the set I’d have gone to 1/60th shutter as I tend to be overly paranoid about flicker.

I graded this at Seedwell on their Final Cut Pro 7 system, using Color 1.5, where I added a node tree based on these directions in the Apple Color manual. The approach was something I’d not heard of before: using edge detection, create a mask for all the strong edges in the shot, then invert the mask and soften everything that’s not a hard edge. It worked remarkably well, as you can see below:

This shot was all about the jewelry. I lit the hand using low, raking soft light that half-lit each finger, emphasizing both the shape of the fingers and the texture of the aligned fingers. Gaffer Alan Steinheimer held a piece of silver card overhead to create a glint in the facets of the diamond ring. Originally we tried moving the silver card around to create a moving glint but it also changed the exposure on the hand and gave away the gag. Holding the card in place gave us just enough of a highlight on the ring to make it come alive without drawing attention to our lighting gag.

The edge-detection node tree trick in Color smoothed the skin nicely but preserved the hard edge of the diamond, which really draws our attention to it.

The Kino Flo behind and below the tabletop helps draw attention to the actresses’ hands and arms during the tug-of-war. The background bokeh is… well, interesting. It’s not terrible, but there’s a little bit of a cat’s eye effect as the highlights approach the top edge of frame. We shot this on a stock 24-70mm Canon zoom wide open at T2.8, so overall the look isn’t bad at all for a fairly inexpensive lens that was never designed for moving picture use.

We lit the bar with daylight Kino Flos for a little color contrast.

These shots cut together quite nicely. I really like how the key is very, very soft but blends gently into the fill and then goes away completely. Soft light with high contrast is very striking, and even though one side of each actress’s face drops almost completely to black the effect is very flattering and suites the mood of the piece.

We grabbed this shot at the end of the day. We were a little rushed due to time but comedy is often about the reactions of innocent bystanders to a bizarre situation, and this waitress looks as if she knows exactly what’s happening, has seen it all before, and is a bit bored by it. David wanted to capture some of the tug-of-war in the foreground so we used a 180mm prime to get this shot. The widest T-stop was T4 or so, and we didn’t have time to do a lot of additional lighting, so we bounced a light off a card placed off frame left and let the background light give her a nice hot edge. In the grade we brought her exposure up some more and then darkened the people to her left and right to create the feeling of a localized light source aimed just at her. She is the focus of this shot, and as it’s a short clip I wanted the audience’s eye to go straight to her.

The fact that the background is a little darker and she’s a little brighter than the other shots doesn’t really matter: if the audience is focused on those details then we’ve done something horribly wrong elsewhere.

Gaffer Alan Steinheimer works his dimmer magic. Alan has a small dimmer system that he’ll bring along occasionally to help us dial in light levels and color temperatures quickly. I’m being speared by the monitor stand, but for some reason I don’t seem concerned.

Turn the page for a bit of summing up and final hints/tips…

Filed under: GentryMedia Sister Sites •ProVideo Coalition •Lighting •Production •Tips •Training •

You’ve read my writing, now hear my talking

Filed under: Cameras •GentryMedia Sister Sites •ProVideo Coalition •Lighting •Production •Tips •Training •Web Video •

Anatomy of a Spot: T-Mobile

Filed under: Cameras •GentryMedia Sister Sites •ProVideo Coalition •Post Production •Production •Tips •Training •

DSC Labs Hawk Chart: The Simplest Color Chart That You Can’t Live Without

Filed under: Cameras •GentryMedia Sister Sites •ProVideo Coalition •Lighting •Production •Tips •Training •Visual Effects •

Arri Alexa and Rosco LitePads Come Through for OnLive’s First National Spot

Director Eric Peltier’s original script called for a series of mysterious shots set in a dark house lit only with a weird flickering light. The light is eventually revealed as coming from a TV that has turned itself on, and eventually the OnLive logo appears. The spot was meant as an online teaser to prime the pump for a more aggressive OnLive digital ad campaign. As we were working on a bit of a budget the director cast his own home, a beautiful old house in Berkeley, California, as the hero location. On the first shoot “day” we pre-lit during the afternoon and then shot until about 1am.

Gaffer Alan Steinheimer.

As camera assistant Paul Marbury built our Arri Alexa in the late afternoon, my gaffer, Alan Steinheimer, who had not yet worked with one, stopped by to take a look. The camera was aimed into the living room of the house, which was unlit but for skylight coming through a wall of windows on the far side of the room. The monitor showed significant detail both in the dark living room interior as well as the sunlit exterior. It was only due to fast action on my part that Alan’s jaw narrowly missed hitting the floor with great force. “I’ve never seen a camera do that before,” he said.

Director/editor Eric Peltier.

My primary concern at the beginning of the shoot day was to create a gel pack that felt like moonlight. Every HD camera camera responds to color in its own way, and often in multiple ways depending on the color matrix chosen, and I didn’t have a sense of how the Alexa saw blue daylight when white balanced at 3200K. I tend to prefer moonlight that’s more cyan than blue as some of the prettiest blues have a bit of green in them, and sometimes the addition of green adds a “silvery” quality to the light. We did some dusk color testing with the camera pointed at a white ceiling into which we aimed a 2k open-face tungsten light, which we then covered with 1/2 CTB to see how the tungsten-balanced Alexa responded.

The resulting blue cast was pretty subtle, so we doubled the 1/2 CTB gel and added 1/4 plus green to it. This got me a bit closer to what I’d been looking for but I felt the blue was still a little muted. Instead of adding more blue we doubled the plus green and got a rich, almost silvery moonlight that felt emotionally correct for the spot. That determined our moonlight formula, which became uncorrected HMI 5600K light with the addition of 1/2 plus green gel.

Both the Alexa and the Sony F35 handle blue with great subtlety, and that’s a trait I greatly appreciate. In days of old certain cameras, Sony cameras in particular, were way too responsive to blue, to the point where anything with blue in it would simply turn blue. A cyan tie or a purple lampshade turned the same shade of bright blue. The release of the Sony F35 and F23 saw a much subtler handling of blue hues, where objects that contained blue no longer simply became blue, and Alexa does equally well in this area.

Director Eric and I line up a shot.

Our shooting plan was to start inside the house, just inside the front door, and gradually work our way into the living room, where the mysterious flickering TV set lay in wait. The character of the flicker was meant to be abrupt and harsh, and initially I toyed with the idea of using small tungsten lights on dimmers to create the flicker effect. Tungsten filaments take time to heat up and cool down, however, and I sensed the attack and decay weren’t abrupt enough to sell an electronic effect. Tungsten lights flicker well enough to create convincing fire effects, but we needed something much, much sharper for a flickering television. It occurred to me that the best way to achieve this would be to hang a video projector over the TV and feed it from the same video source. The TV wasn’t bright enough to illuminate a room on its own so the projector would push the video image deeper into the room in perfect sync with the TV.

Director Eric loved the idea and we acquired a video projector within the hour.

First, though, we did some shots in the vestibule just inside the front door, looking through some glass pane doors into the living room. As budgets have been somewhat reduced in the last year or two we’ve been adapting to methods and techniques that are more “budget friendly”; in this case we traded a dolly for a custom-made 4’ slider built by local grip Todd Stoneman. By using wide angle lenses (primarily a Duclos 12-16mm T2.8 zoom) and putting something in the foreground of every shot we made excellent use of all four feet.

In this shot you can see hints of flickering projector light on the ceiling.

We lit the room primarily with a 4k HMI fresnel placed outside the living room windows. I think we also had a 1200w PAR outside the off-screen window on frame right. We used the fresnel for hard shadows that raked the living room furniture through some very nice multi-paned windows along one wall, while the PAR—which casts much less distinct shadows—merely contributed to the overall ambience. I think in this shot we had an additional small PAR aimed through another window at the sliding door.

As bright as the living room looks in this shot the only light placed inside the room was the video projector. This allowed us to use wide angle lenses without a lot of tweaking and relighting.

We shot a number of inserts around the inside of the living room—-a couple of laptops sitting on a table, stereo headphones draped over the arm of a chair—but the one additional shot from this night that made it into the final :30 is that of a collection of toy robots:

In retrospect I probably could have defocused the projector a little to make the flicker more mysterious, but it’s a pretty darned cool shot just the way it is. The robots are lit with a whisper of additional light from a single 3200K Kino Flo bulb taped to the edge of the shelf.

The rough cut containing these shots was so successful that the client asked us to return the following week with an actor to shoot additional footage, much of which replaced the most of the shots we captured on the first shoot day. With the exception of camera assistant Paul Marbury, who wasn’t available and was ably replaced by Rod Williams, the original crew returned and we repeated our fairly simple lighting setup. This is where the fun REALLY begins. Turn the page…

Filed under: GentryMedia Sister Sites •ProVideo Coalition •NAB 2011 •Tips •Training •

The Secrets of the Chroma Du Monde, Explained Live (on tape) at NAB!

Filed under: GentryMedia Sister Sites •ProVideo Coalition •Production •Tips •Training •

A Mix of Film and HD Doesn’t Scare Arri’s Alexa

Even with only one bulb on the 6k space light was a touch bright, so we placed a 4’x4’ double net underneath it. The next-to-last touch was a tweenie, hung on another goal post directly in front of the talent with Lee 216 clipped to the doors. This wrapped the toplight around the front of the actor’s face and gave him a nice eye light. And, last but not least, I wanted to add just a little bit of fill from below for the closeup at the end of the move. I asked dolly grip Ernie Kunze to rig a 1’x1’ bead board directly beneath the lens, and he rigged a white platypus clamp to a C-stand arm before heading off to find the 1’x1’. I stopped him: the white platypus clamp added just the right amount of fill, so we left it alone and forgot about the bead board.

Producer Sean Cope has a chat with actor Douglas Olsson. A bunch of naked 2k space lights illuminate the cyc through a 20’x20’ frame of light grid cloth. The overhead light for the talent is a 6k space light with one bulb on, further treated with a 4’x4’ double net. At the upper right corner of the frame is a tweenie with Lee 216 on the doors to wrap the top light around the actor’s face and give him a little eye light. This light was actually in the top of frame at the beginning of the move but as it was clear of the actor it was easily removed in post. (Moving it out of the shot completely would have compromised the eye light.) The fill for the closeup is the white platypus clamp directly beneath the lens. We didn’t have to put anything in it, it was perfect on its own.

Another view of the lighting setup, during lunch with the house lights on and everything else turned off. This shot shows the large black teaser hung by the grip crew to keep the light from the 6k space light on the set and away from the client and the lens. The goal posts holding both talent lights are visible.

The eye light from the tweenie is plainly visible in the talent’s eyes. Because the diffused tweenie is basically an extension of the soft overhead light, we only see one shadow even though there are two lights in play.

The dolly shot took place over about 90 seconds, and we landed 2’1” in front of the actor’s face. Having been a camera assistant, I try not to give my assistants impossible shots when I don’t have to. I rated the Alexa at EI 400, which I prefer over 800 as it’s a cleaner, less noisy look, and we lit the stage to T5.6 1/2. This left camera assistant Rod Williams plenty of time to make phone calls, send emails and run to craft service for snacks during the first 75 seconds of the move before having to concentrate completely for the last 15 seconds. Even at T5.6 1/2 he had only 1 1/2” of depth of field at the end of the move, as the dolly slowed and came to a halt.

The white platypus clamp adds just a touch of fill to the closeup. You can see it reflected subtly in the lower half of the actor’s eyes.

We adjusted the speed of the dolly during the move, starting out at a higher rate of speed and then slowing as we approached the end of the track. This actually made the dolly move appear consistent in speed, rather than feeling very slow at the beginning and very fast at the end. Ernie worked out a system where he counted dolly track ties and delivered very consistently timed moves. The actor had recorded the lines the day before and we played them back during the shot for timing.

We brought the normal engineering tools in the form of a waveform/vectorscope, but we used them only to confirm exposure. There was no need to color balance the camera beyond setting the color temp at 3200k, and we controlled the contrast entirely through lighting. I had my crew arrange large black solids on the white stage nearly up to the talent’s feet, where we left a gap to allow a little bounce light from the floor to illuminate his hands from below. Blocking the white floor eliminated all fill except for what was provided either by a light or the portion of floor we left uncovered at the actor’s feet. (I didn’t want “accidental” fill; I wanted to control exactly what was filled, how much, and from where. Placement of the fill light can make a HUGE difference to a shot.)

The shoot went very quickly and smoothly, and we shot around 30 takes. Although the plan was to cut away to stock footage the director felt that it was worth the time to finesse the talent’s performance, and then he pulled what he wanted from a handful of takes. Within a week the project was delivered to a very, very happy client, who projected the piece once as an introductory piece during a large company conference.

My DIT, Jay Farrington, told me a very funny story toward the end of the day. Jay is one of the owners of Chater Camera, who provided the camera gear for the shoot, and during the day he’d been receiving phone calls from another rental house who was trying to round up another set of Ultra Primes for a project that was filming the next day. “I’m sorry,” said Jay, “but my last set is with me on a job right now. You can have it this evening, but I can’t release it any sooner.”

Later in the day he got a return call from the other rental house: “We found another set, all we need is a 40mm! Do you have one available?” “You’re in luck,” said Jay. “That’s the ONLY LENS we’re using all day!” I’m not sure what the response was but I suspect it was along the lines of “Aaaaargh!”

DIT Jay Farrington objects as camera assistant Rod Williams assigns blame. Hidden behind Rod is a KiPro deck. As the Alexa was unable to play back footage at the time due to software limitations, Jay recorded the HD-SDI output for playback and as an editorial backup.

Thanks to a great crew, and a great camera, we had a very productive and non-eventful day. I like shoots that keep me on my toes, but every once in a while it’s nice to do good work while being slightly under stimulated. This project went like a dream.

Director/Creative Director: Justin Curtis

Agency: GYRO:HSR

Talent: Douglas Olsson

Producer: Sean Cope

Production Company: Sean Cope Pictures

Production Manager: Susan Gavet

Post: Teak Digital

DP: Art Adams

Gaffer: Charles Griswald

Dolly Grip/Key Grip: Ernie Kunze

Swing: Rick Edmondson

Camera assistant: Rod Williams

DIT: Jay Farrington

Camera package: Chater Camera (Arri Alexa and Ultra Primes)

23.98fps, EI 400, T5.6 1/2, ProRes 4444.

Art Adams is a DP wHo LiKeS a LoT oF cOnTrAsT. His website is at www.artadamsdp.com.

Filed under: GentryMedia Sister Sites •ProVideo Coalition •

Where I’ll Be at NAB

Filed under: GentryMedia Sister Sites •ProVideo Coalition •Post Production •Production •Tips •Training •

The Secret Art of Slating: 25 Tips to Help You Slate Like a Pro

The only time it’s acceptable to shoot a slate out of focus is when the shot is locked off for visual effects.

16. If you’re simply ID’ing a take, which happens when it is being recorded without sound or if the sound is being recorded single-system with the picture onto tape or a hard drive, put the slate into the shot with the clapper closed. If the clapper is open then the editor will assume that it’s going to close and they’ll look for audio to sync up to it, but if it’s closed then they know not to bother. For example, the shot above was captured without sound.

17. Sometimes slating has to happen at the end of the shot. Hold the slate upside down to indicate that the slate is happening at the end of the take. (These are all visual cues that allow editors and their minions to very quickly figure out what’s going on with both picture and audio.)

After the director calls “Cut!” it’s a good idea to yell “Tail Slate!” or “Tail Sticks!” to remind the operator and first assistant not to stop the camera yet.

18. Make sure the slate is well lit. Shooting in the dark is no excuse for a dark slate. Ask the electricians to set up a small slate light that they will turn on for slating, or invest in a good flashlight with a wide beam.

19. When the shot is very, very tight there are a couple of things you can do:

If there’s room, simply hold the slate so scene and take fill the frame. When the camera rolls, wait a second and then move the clapper down into frame, say “Mark!” (or “Marker!”) and hit it.

If the shot is so tight that you can’t get both the scene and take in the frame, hold the scene box on the slate in the frame and wait for the camera to roll. Count to one and move the slate horizontally so the take number fills the frame and count to one again. The lower the clapper into frame, say “Mark!” (or “Marker!”) and hit it.

Insert slates are miniature slates, although often without clappers. If an extremely tight shot requires a clap and you don’t have an insert slate with a clapper, take the clapper off a regular slate and use it. Hold the insert slate in the shot as the camera rolls, count to one, and then hold the sticks in frame, say “Mark!” and clap ‘em. (Sometimes you can remove the sticks from a full size slate and squeeze the insert slate into the slot, toward the open end of the sticks.)

In the absence of an insert slate you can write the scene and take together in the “take” box (“17A/1”) and only shoot that portion of the slate. It’s a good idea to include all the information on the slate when possible, but the bottom line is that post only cares about the scene and take information. They probably know who the director and DP are.

20. Slating for multiple cameras on film was pretty simple: you’d “bump” a slate on each camera, meaning you’d roll a second of film on each camera’s slate, and then at the beginning of the take you’d take one set of clapper sticks out in front of the cameras and say “A and B common mark” before hitting them. (This tells the editor that A and B cameras are rolling and this sync mark works for both.) That doesn’t work in digital because bumping a slate means creating a separate file, which defeats the purpose of ID’ing a take.

The solution is to hit each camera’s slate separately. Bang the slates in sequence: “A camera mark!”, “B camera mark!”, etc. Hopefully there are enough people around to help if there aren’t enough assistants to do the job. Often the A-camera first assistant can slate their own camera as they usually have a wide shot.

When slating multiple cameras keep your slate in the frame from the beginning of the shot. This not only creates a thumbnail of the slate for the editor but it blocks the camera’s view of other slates, preventing the editor or assistant editor from trying to sync the audio to the wrong camera’s slate.

Each camera will have its own slate with a large “A” or “B” or “C” letter on it somewhere. The letters are often different colors: red for A camera, blue for B camera, etc. (There’s a color system for additional cameras that I don’t remember anymore.) Once again, this is another visual tool that helps an editor quickly figure out what’s going on.

(21) Occasionally you’ll have odd slate numbers: R17A means you’re reshooting 17A, 17A-TV means you’re shooting a “clean” version of the scene for TV, etc. Visual effects projects often have a very bizarre nomenclature for identifying scenes and takes, often because each completed shot requires many separate visual elements and all the elements must be tracked. The script supervisor or visual effects supervisor will guide you in this.

(22) Roll (or card) numbers are often ignored by union crews. I’m not sure why, but assistants frequently buy slates that don’t have a place for roll numbers and union editors typically don’t ask for that kind of information. (See the slate at the top of this page.)

(23) Make sure you spell the director and DP’s names properly. If you want to impress everyone, get a label maker and print out neat labels for each new shoot. It looks very professional, and that goes a long way toward impressing those around you. Neatness counts. (Large budget projects will custom order engraved slates with the project title and director/DP information.)

ROOKIE MISTAKES that you’ll want to avoid:

(24) DON’T TAKE THE SLATE AWAY FROM THE CAMERA. There’s a tendency for rookie second camera assistants to hook the slate in their belt and walk around with it. This is pointless and possibly embarrassing, because if you’re not on the set when the cameras roll someone is going to be very angry with you.

Always leave the slate in the first camera assistant’s front box, on the dolly, or some other consistent place near the camera so that the first assistant can easily find the slate and mark the shot if you’re not around. The slate should always be within easy reach of the first camera assistant.

(25) NEVER REHEARSE SLATING. Film students often think they have to hit the slate during rehearsals, when the camera isn’t rolling. You don’t rehearse the slate as it isn’t part of the performance, it’s just an identification tool for editors who have to round up the good takes quickly and cut them together.

Some of these tips may seem unusually intricate but there’s a good reasons for every one of them. In the film business time is money, and knowing how to mark a shot quickly so that an editor can find it quickly saves a lot of money in the long run. The best way to make a producer angry is to force them to spend more money than they need to, and bad slating can waste a lot of money—both on set and in the editing room.

My first union job was loading film on a sitcom with a crew that hadn’t done a lot of sitcoms before. The first assistants refused to do their own paperwork (“That’s the second assistant’s job!”) and the second assistant was a bit crazed managing four cameras on the first few episodes. I often received rolls of film that had nothing on them but a blank camera report and a roll number. I did my best to fill in the camera reports but I often had to send the rolls to the lab with almost no scene and take information.

On the third episode a guy walked up to me on the set and said, “Hey, can I ask you a favor? Can you PLEASE make sure the camera reports have the scene and take info on them? I’m the third assistant editor they’ve hired on this show and I’d like to keep my job.” Upon hearing this the first assistants became a bit more cooperative and that assistant editor was able to sync dailies fast enough to stay employed.

What you do on set affects a lot of people down the line. Get in the habit of doing it right and, in time, it’ll become so routine that you’ll be able to focus on the hundred other things you have to learn to do properly.

Art Adams is a DP who likes to be in sync with his crews. His website is at www.artadamsdp.com.

Filed under: Cameras •GentryMedia Sister Sites •Apps4Phones •ProVideo Coalition •Lighting •Mobile Devices •Pre-Production •Production •Software •Tips •

iPhone Apps: The Short List for the Average Cinematographer

Filed under: Cameras •GentryMedia Sister Sites •ProVideo Coalition •Production •Tips •Training •

Panasonic AF-100: The Good, The Bad, and The Ugly

This is the opening sequence of the piece, and it very quickly revealed the character of the camera. During prep I discovered that the AF-100 has about three stops of overexposure latitude before it clips, and the clip isn’t very clean. Most cameras work hard to make the clipped values look as filmic as possible, which typically means the clipped areas don’t turn into an odd electronic white with sharp edges. The Arri Alexa and RED ONE do a very nice job with clipped highlights, and Panasonic camcorders do a good job in Film Rec mode, but the smaller cameras like the Sony EX1/EX3 and Panasonic HVX-200/HPX-170 don’t handle clipped highlights very well at all. Sadly the AF-100 looks very much like a large sensor version of the HVX-200 in this regard.

The shoot location was in the Sea Cliff area of San Francisco, and there was one major problem that presented itself on the shoot day (we didn’t have the money for a scout): next door there’s a HUGE palm tree that blocks most, but not all, of the sun from the front yard of our hero house. We didn’t have the firepower to bring the shadows up in value, so whenever possible we brought the sunlight down.

The Google Maps view of the location. The palm tree just below and to the right of the “A” is the main culprit. This image was taken in the summer; our sun was much lower, it being January, and was actually hidden from the driveway until about noon.

In this case, however, since the background was in full sun, we had to pop the driver with a shiny board double bounce. The car is in the shadow of the tree, while one shiny board is on the near sidewalk (just off frame left) aimed at another shiny board in front of the car. I lit the driver as frontal as possible as we didn’t have the time or resources to execute both a key and a fill, and an off-center key would have left some really dark shadows. The frontal light is very flattering to her face, and I’m all about lighting faces.

I let her go dark as she exits the car. By that time we’re supposed to be watching the guy in the background, so letting her drop off into darkness helped direct attention to him. I’ve discovered over time that cheats like this can work nicely to direct attention, and it’s also generally good practice to avoid over lighting a shot.

As the actor runs across the street the side of his face clips due to the open sun. The far side of the street is exposed as brightly as possible without clipping (the sidewalk was close to 100% on the AF-100’s built-in waveform) in order to open up the car interior as much as possible, and the camera looks really nice as long as it doesn’t clip.

I should point out that PVC’s Adam Wilt was my camera assistant and did a very nice job on focus, especially as I tended to shoot close to wide open (T2.6) to reduce the depth of field. (Sorry, Adam!)

This is the reverse shot, and we did this toward the end of the day when the sun went behind some buildings to the west. The mixture of sun and shade was too much for this camera to handle so we had to wait for all of the sun to go away.

Here’s my biggest complaint about clipping on the AF-100:

When I saw the brake lights I just had to roll some footage.

Those brake lights aren’t supposed to be orange. They’re an orange red, and while most other cameras would render them as clipped red this camera seems to say to itself, “I see a clipped red channel and a clipped green channel, so I’m just going to mix those two clipped values together.” When you mix red and green you get orange. It’s not an attractive look.

The Panasonic Varicams have a feature called highlight or knee saturation which pumps a lot of color into highlights. It’s not an attractive look because the clips turn a weird over-saturated version of whatever color, or colors, are clipped or close to clipping. During prep we aimed the camera out Shooting Star’s loading door at a car covered with a blue tarp, and when the tarp’s exposure clipped it turned a bright featureless cyan. It was really, really ugly.

I also noticed that, at the default setting, bright colored objects are chroma clipped long before they are luma clipped, so on subsequent projects I’ve reduced the camera’s saturation level by a couple of notches to reduce this effect.

It seems as if this camera uses highlight saturation ALL THE TIME. It would be really nice to be able to turn this off or desaturate and remove detail from the highlights, but this is a consistent problem with low-end Panasonic cameras from the HVX-200 up. Even the HPX-500 does this.

It’s too bad, because under controlled lighting this camera is really gorgeous.

We did this moving shot on a Dana dolly. I hadn’t heard of this nifty little invention until very recently. It’s a ball mount on a small platform equipped with skateboard wheels, and it uses common speed rail as its track. It comes with a couple of spacers that hold two rails at the proper width and it works with just about any length of speed rail.

Operating the Dana Dolly with Adam Wilt’s C-stand arm extension.

There are two things to know about this rig:

(1) It doesn’t lock, so if you walk away from it make sure someone is watching it so it doesn’t roll down the track.

(2) It’s not locked to the rails so it is possible for it to tip over. I’ve not had a problem with that but it is possible.

As a low cost dolly, though, it can’t be beat.

It was pretty difficult to do a fast yet smooth move hunched over such a low track, so Adam suggested attaching a C-stand arm to the front of the Dana dolly in order to exert constant force along the center line of the dolly. Adam set the focus and I watched his monitor to operate the shot.

Here our character starts to scratch a design into the hood of the car with a key. The garage is in shadow but he isn’t, so he’s covered by a 12x12 silk that gives him a nice soft sidelight.

Large sensors are all about shots like these. Adam racked from the actor’s face to the key and back again. I didn’t notice the Alura breathing at all.

The hood of the car was protected by a plastic shield. The AF-100’s large sensor makes this a gorgeous shot. This could be done on a Canon 5D but there’s always the danger of moire, not to mention that it’s nearly impossible to tell if something is in focus on the 5D’s tiny standard def LCD screen—especially if it’s in motion. This wasn’t a problem when viewing the AF-100’s built-in LCD screen.

Adam and I shooting a high angle.

We shot some other angles on this character doing his deed, including one that “looks through” the hood of the car. Ian came up with this idea on the day and we improvised with a sheet of clear acrylic. The scratches are a little hard to see but maybe Ian can enhance that in post. (He’s doing the actual etching of the hood in post as well; the budget didn’t include repainting a car.) Next time I’ll try to do some testing to see if there’s a lighting angle that brings out the scratches.

Uncorrected sunset light.

You’ll notice that this angle is very warm. That’s because we shot this at magic hour and the sun was rapidly disappearing. (We had to start shooting in early afternoon to allow the sun to move around that big palm tree next door.)

Shooting through the “hood.” The black duvetine eliminates reflections from the front surface of the acrylic sheet.

The light on the actor’s face is a 1’x1’ tungsten LitePanel provided by gaffer Luke Seerveld. I did a rough color correction pass to see what happens when the excess warmth is removed:

Quick and easy correction in Final Cut Pro shows that this shot can be easily graded to match the others. (Just not by me.)

This was done quickly using the three-way color corrector in Final Cut Pro. I used the white picker to white balance on the window frames in the background. We’ll do a much more careful pass later with a real colorist.

Ooops—one of our characters caught the other. While the guy is still lit from the sunlit silk off frame right, the woman is lit with a shiny board pushing sunlight through a frame of Lee 250 just off frame right. There’s another shiny board raking the bushes and spilling onto the fence in the background.

She looks beautiful but the skin tones are just on the edge of clipping. I tend to overexpose flesh tones a little to make them “pop” and while the camera isn’t anywhere close to clipping luma it’s on the verge of clipping chroma, which is just as bad. Still, as long as you don’t cross that threshold, the camera looks great. And, as I mentioned earlier, dialing the chroma down a little reduces the odds that you’ll inadvertently clip a color.

You can’t see the subtleties of chroma clipping on the camera’s LCD monitor. Our DIT, Jeff Regan of Shooting Star Video, kept a sharp eye on our “critical” monitor, a 17” Flanders Scientific LCD.

The Flanders is my new favorite inexpensive LCD monitor. Supposedly it uses the same panel as the Panasonic 17” model that everyone knows and loves, but instead of trending toward green, the way the Panasonic does, it trends slightly magenta. It does take a little mental correction to properly interpret color, but a slight magenta cast on flesh tone looks a lot better than a slight green cast. I find I don’t fall into the trap of correcting something that looks too green but really isn’t, as I occasionally do when I see someone with an olive complexion on a 1700 or 1710 Panasonic LCD monitor.

(Side note: in the real world there is no magenta light in the spectrum: it’s simply the absence of green. Apparently it’s very difficult to render green properly on that panel, so instead of adding slightly too much green Flanders apparently opted for slightly too little.)

Here we had enough 12x12 coverage to silk the character and his exit path (he walks sheepishly away around the back of the car) and not quite enough to do the entire background. The setting sun helps us a little but there’s still some leaf clipping in the background. We’ll probably track that and remove it in post.

The AF-100 is a contrasty camera. I had to hide a bounce card behind the actress in the shot above, just to get it close enough to fill the actor properly. (Ian will use either the previous shot or this one in the edit, so the mismatch in background lighting won’t be a problem. That’s what you do on a budget.)

You’ve been very patient so far, and I want to reward you for it. On the next page you’ll find not only my summary, along with some rough color correction tips, but a little idiosyncrasy that may bite you when using the AF-100 with a large lens. It’s nothing to worry about, and it’s easy to fix and prevent, but you definitely need to know about it in case it happens. You’ll never guess what it is.

Read on…

Filed under: Business •Production •Tips •Training •

Career Advice for the Young DP

Filed under: Cameras •Production •Tips •Training •

Phantom Adventures: 1000fps on a Budget

This is raw footage from the first setup, daytime in the living room:

Here’s what it looked like behind the scenes:

The trick was to pour 2000fc (plus or minus) into the set and make it look reasonably natural. (Later on we opted to turn the shutter off and reduce our light levels to 1000fc as none of the action was fast enough to show the difference.) During preproduction I took a look around the internet to see if I could find some examples of naturally-lit interior high speed work and I couldn’t find any. Everything I found was either shot day exterior or photographed within a very, very small and brightly-lit space. I had to start from scratch, so I opted to build up a base level of soft light and then add hard light accents.

We started out by bouncing several 12-light Maxibrutes off a 12x12 UltraBounce outside the right frame, where the right wall of the room would be, to create a base level of ambient light. We then added a couple of Maxibrutes to the left of camera (later, when we moved in for closer shots, they came over the top of camera) through a poly silk, followed by 1.2k tungsten VNSP (very narrow spot) “firestarter” pars to create sunny accents around the room. (Firestarters are called that because, if you put something too close to them, they will.) We also poked a nine-light bulbed with Firestarters over the corner of the set as general backlight.

The sunny window effect was a combination of warm 10k accent lights raking the curtains and additional units lighting the white cyc wall behind.

All the black lights are Firestarters, very hot and very spotty par lights that add “hits” of sunlight around the room. The one in the center is aimed into a mirror. A nine-light Maxibrute backlights the room through a frame of diffusion, probably Lee 250.

The wide shot took a bit to light but worked very well. Initially we captured the kids separately from the dog as we didn’t know how the dog would react to cues. Later we tried tried shooting all three performances at once for the medium shot and got one that worked perfectly. While the dog was comped in for the wide shot, it performed live during the medium shot.

Here you can see the 12’x12’ Ultrabounce that’s providing most of the base light for the set. The fresnel in the center of frame is bouncing off a white card to create soft upward shadows on the bookcase, as if sunlight was bouncing off the floor. There’s another bounce hidden behind the couch. We were clearly working at the limits of the stage.

The rest of the set.

Gaffer Luke Seerveld wonders what he did to deserve all this.

The Phantom HD Gold is an odd camera: it is always, always, always capturing imagery. The camera has a 16gb buffer and image data is always flowing in one end and out the other. “Rolling” the camera simply means that you capture whatever is in that buffer. There’s no “roll camera” and “cut,” only “action!” And then you look at what you’ve got. At 1000fps that buffer holds four seconds of real-time footage that plays back at 24fps over about three minutes.

Imagine the Phantom’s capture buffer as a pipe. Data flows in one end of the pipe and exits out the other. When the capture trigger is tripped, whatever is in the pipe is saved. What’s really cool is that you can set the capture trigger anywhere along that four second pipe: if you set it at the beginning of the pipe then hitting the trigger saves the next four seconds of real time at 1000fps. If you set it in the middle of the pipe then hitting the trigger captures two seconds that are already in the pipe and the two seconds that follow. (This is handy if you want to capture data just before and just after an action. For example, if you’re dropping something in water and want to see its entrance and the subsequent splash, you’d put the trigger in the center of the pipe and hit it just as the object touches water.) You can even set the trigger at the end of the pipe and trigger capture as the event ends, capturing the previous four seconds of action.

In this case my Phantom Tech, Jay Farrington, often set the capture trigger about 1/4th or 1/5th of the way into the pipe, or buffer, as he knew how to gauge his reaction time in relation to the action on screen. As soon as he saw the sweet spot of action he’d hit the trigger, and he nailed it every time.

This is the monitor output of the Phantom HD Gold. The line across the bottom represents the buffer, or data “pipe,” and the triangle indicates at what point the camera will start saving data when triggered. Usually Jay set this a little in from the left, but he varied the settings as needed to make sure he got the shot. The monitor says we’re using a 180 degree shutter but we turned that off at some point to reduce light levels. 1/1000th of a second exposure turned out to be plenty to freeze the action we were shooting, instead of 1/2000th of a second with a 180 degree shutter..

After shooting a circle take the most important portion of the shot is marked with in and out points in the camera and that section only is transferred from the camera to the Flash mag. The Flash mag footage is then played back through the monitor out port (with the character generator turned off!) and captured by the Ki Pro. It’s rare for an entire take to be recorded as three minutes of 1000fps footage is a huge amount of data, and most of it isn’t usable. We only needed ten seconds out of three minutes of playback time.

As we wanted to capture ten second shots we only had about a quarter of a second (240 frames out of 1000 per second) for all the elements of a shot to gel. That’s pretty amazing when you think about it, and even more amazing when you see it actually work. The kids nailed their performances, and we added the dog in the medium shot and got a take where all three living beings in the frame did exactly what we wanted them to do within a quarter second of capture time.

I gave little sister a bit more light in the wide shot as I wanted her to “pop” as the center of attention. I brought that light in a little closer when we did the tight shots as I wanted it to wrap around her face a bit more.

Production manager Vanessa Tomasello, sitting in for little sister, bursts with joy over being on time and on budget.

We used a six light FAY fixture (it looks like a smaller version of a Maxibrute) to give little sister some extra light on the wide shot, and for the medium and close shots we moved it in a bit to make the light wrap around her face and look more natural. Soft light from near the lens is very flattering to faces. The position and angle of the diffusion puts most of the light on little sister, making her the star of the shot, while big brother gets some fill from the same light and a little modeling from the nine light Maxibrute over the corner of the set. The diffusion in front of the FAY light is probably Lee 216.

Director Jono Schaferkotter offers wise words of advice to big brother while I stand in shorts next to a very hot lamp. Behind us is the over-the-camera fill light, a Maxibrute through an 8’x8’ frame of poly silk.

The shot of the controller was done with existing lighting. The trick to most product photography, especially for objects that lay flat, is to light the object from behind, so I found an angle where the controller reflected the diffused nine-light Maxibrute poking over the corner of the set. If the product is shiny I use a big source that reflects in the shiny surface. If the product isn’t shiny and has surface texture then I use a harder light to bring out the texture. The shadows should always fall toward the camera.

Let’s move on to the den for some quiet evening laptop action…

Filed under: Cameras •Production •Tips •Training •

Alexa ISO Settings: The Least You Need to Know

Filed under: Cameras •Production •Tips •Training •

Arri Alexa’s Dynamic Range: It’s All in How You Use It

Yay! Don’t you love looking at charts? It’s a necessary evil, sadly, as there’s too much unknown detail cluttering the average shot to really understand what a camera is doing. A test chart is a known quantity: we know how it is supposed to work, we know how it should appear, and if something happens that we don’t expect then we can be fairly sure that there’s something going on with the camera because charts—over the short term—don’t change.

Here’s the DSC DX-1 17-stop chart photographed by an Arri Alexa using an 85mm Zeiss Ultra Prime. The stop is T2.3. The chips are placed at the top of the frame following Adam Wilt’s advice concerning flare: placing the bright chips in the center creates flare in the center of the frame that overlaps the chips and affects how they look, whereas framing the chips high, low, left or right throws most of the flare to the opposite side of the frame, away from the chips. In other words, placing the important part of the test chart to one side throws any flare to the OTHER side. (Very clever.)

In the top image you can see the piece of tape that I ran down the surface of chip #7. We can see the effect of that piece of tape on the waveform next to the label that says “Notched Reference.” That chip fell exactly on 45% in LogC mode, so I picked it as my middle gray reference. It floats up to 50% in Rec 709 mode, so in Rec 709 we’ll call 50% middle gray for ease of analysis.

Middle gray stays at nearly the same value when toggling between LogC and Rec 709, with the upshot that the LogC image-—which is not designed to be viewable on any kind of HD monitor-—still looks okay when viewed on a Rec 709 monitor. (Arri’s FAQ states that an 18% gray card should read at 39% in LogC mode and 38% in Rec 709, but none of my regular technical sources believes that 38% is a proper 18% gray value in Rec 709. Answers vary but range from 42% to 46%. The bottom line is that middle gray changes very little when toggling between LogC and Rec 709, and this seems to make LogC more “monitor friendly” than other log curves.)

There are a couple of things we can discern by looking at the relationships of the different chips to middle gray on the LogC waveform:

Looking at highlight latitude, I see seven stops of exposure latitude before clipping occurs, which is unheard of in an electronic camera. Four or fives stops is typical. This is clearly why Alexa is so good at handling highlights, which have long been HD’s weak spot.

Looking at shadow latitude, I see eight stops of exposure latitude before hitting solid black. This can be seen in the chart itself, where the 15th chip is just barely discernible from black, but whether you’d ever want to expose something important at that value is questionable. My guess is that I’d plan on -5 being the lowest really usable stop, and know that in order to achieve real black I should probably push the exposure more than eight stops below middle gray. Odds are that everything below stop -4 or -5 will be crushed in the grading process, but if you need to hold detail in those tones it’ll probably be there. Just don’t try to pull them up much or you may find more noise in the image than you like.

Log curves are designed for efficient storage, and nothing more. They are not raw, and you are not excused from white balancing when using them. What they do is preserve the maximum amount of usable grading information by storing tonal values according to perceptual steps. If you want to learn more about how Log curves work, read this article. (And stay tuned, as I’ll be talking to Arri shortly about how their LogC technology works.) LogC is a great way to observe what kind of information is really available to the camera, and if you’re headed towards color correction for a broadcast master this is definitely the gamma you’ll want to record in. And do note that LogC is a gamma curve only: the color in the image is unaffected.

The most interesting thing about LogC is that the only time the values compress is when they descend into noise at the right side of the waveform. It is unknown whether this is due to the slope of that part of the log curve or due to the “dog leg” that appears as sensor response drops near the noise floor. Otherwise every stop in the camera’s dynamic range stores the same number of bits as any other, which is seen in the equal steps between most of the chips. This helps to prevent banding during the color grade.

Let’s look at the Rec 709 curve:

This is the same chart, exposed exactly the same way, but with Rec 709 gamma applied. Wow, what a difference. The good news is that this is the Alexa’s what-you-see-is-what-you-get mode, where you can trust a properly calibrated broadcast monitor and vectorscope to show you how the final image will look without grading. It’s shocking, though, to see what you give up at the extreme ends of exposure.

The important thing to remember about Rec 709 is that it was created at a time when cameras could deliver only about five stops of dynamic range-—so that’s all Rec 709 is designed to hold. And it’s rare to find a modern camera with a true Rec 709 curve in it, because there’s not a single camera out there that only delivers five stops of dynamic range. Instead, manufacturers create curves that look good in a Rec 709 environment, in spite of the fact that the cameras capture more dynamic range than Rec 709 is supposed to hold.

The trick has become cramming 10+ stops of dynamic range into a small five stop “bit bucket” in such a way that it still looks good. As you’ve probably noticed from the image above, Alexa does this by applying quite a strong S-curve to the gamma, stretching out the mid-tones so they are nice and contrasty while compressing the highlights and shadows much the same way film does.

If you’re curious as to what happens when the mid-tones aren’t emphasized, look at the log curve above. Cramming 15 stops into a five stop “bit bucket” results in a very flat-looking image: whereas five stops have plenty of room to stretch out, fifteen stops end up crammed closely together and the contrast between each step is reduced. Stretching out the mid-tones, which is where most of the important visual information is, and compressing the highlights and shadows, where our eyes naturally discard detail, is often the best way to shoehorn all that latitude into a small “bit bucket.”

By analyzing the Rec 709 curve and comparing it to LogC, my guess is that stop -5 is about as low as you’d want to expose something that should have some presence in the frame. Looking at the chart itself, chip #12 is the last one I can clearly see, and that corresponds to -5 on the waveform. Stops -6, -7 and -8 have some presence but they’ve already been compressed to the point where pulling any detail out of them in post is probably impossible. They are so close together that there won’t be much contrast between those values and, while they’ll be visible, they may blend together to the point where image detail in those shadows becomes indistinct due to lack of contrast.

What’s interesting is that the LogC and Rec 709 curves look similar between step -5 and black. The black levels are different between the two curves but the lowest stops show about the same amount of distance above black.

The real difference between the curves is in the ability to access highlight detail in post: Rec 709 rolls off and compresses highlights the way film would, but LogC better preserves the differences between the tones and allows a colorist to massage them more easily later. We can see these differences in the charts:  LogC shows the brightness steps at the top of the scale very clearly, but Rec 709 shows the difference between steps +5 and +6 (chart steps 1 and 2) as the last discernible step, with the difference between steps +6 and +7 (chart steps 0 and 1) being too small to see.

One interesting thing is that while the Rec 709 curve fills the entire “bit bucket” from 0% to 100%, LogC only extends from 0% to 95%. In extended mode, where the bits from 100-109% are made available, LogC still only rises to a maximum value of 104%. During our tests Adam and I discovered that there’s only one ISO setting that uses the full range of the “bit bucket,” but you’ll have to wait a bit before we reveal which one.

Here’s a better look at how many stops are available overall in ISO 800:

The vertical charts are recorded in LogC. Chip #15 is just barely visible as brighter than the black dividing line between sections.

Arri has asked me to clearly state that their factory considers the camera to have a maximum dynamic range of 14 stops. Based on what I see in this chart that measurement is slightly conservative as the 15th stop is clearly visible on the chart, but it’s so faint it’s likely not usable. This is a case where practicality wins over marketing, as marketing dictates shouting “15-stop dynamic range!” from the rooftops while practicality recognizes this 15th stop isn’t something that should be relied on.

In a previous article I noted the difference between what I call “paycheck” stops and “gravy” stops:

Paycheck stops are values that you can bet your paycheck on. They’ll appear in the image with enough contrast and detail that objects exposed at that value will be easily discernible.

Gravy stops are the values that lie beyond paycheck stops: they are the tonalities at either end of the dynamic range scale where some presence will be seen, along with some discernible detail, but they shouldn’t be trusted. If there’s some information held in those tones then that’s great, but don’t count on it.

That 15th stop is definitely a gravy stop. Don’t bet your paycheck on it.

NOTE: In some of these chart images you might see some magenta or green fringing around highlights. That appears to be longitudinal chromatic aberration, as changing lens focus caused the colors to shift from green to magenta. This means the color shifts are lens related, not camera related. (The vertical charts were shot on a 24mm Ultra Prime, and I’ve zoomed in to the image to make it visible. Why a 24mm Ultra Prime? Because the 85mm was too long and all the other Ultra Primes were out on rental that day.)

Incidentally, there’s no spectrum of light that’s magenta—so what we’re really seeing is that a portion of the image has either too much green (which appears green) or too little green (which appears magenta).

Turn the page to see what happens when we deviate from Arri’s recommended ISO “sweet spot” of 800…

Filed under: Cameras •GentryMedia Sister Sites •ProPhoto Coalition •Production •

Canon 5D: How much dynamic range does it have, really?

Filed under: Cameras •Lighting •Post Production •Production •Tips •Training •

My Love Affair with Alexa

Left to right: Me, camera assistant Paul Marbury, gaffer/DCS chapter president Simon Sommerfeld, and Leigh Blicher, co-owner of Videofax.

My intrepid crew of volunteers included DCS chapter president Simon Sommerfeld as gaffer; camera assistant Paul Marbury; Adam Wilt as behind-the-scenes documentarian; and Jim Feeley as production assistant. Also accompanying us was Michael Bravin of ARRI and Leigh Blicher, co-owner of Videofax, who provided us with nearly all of our support gear. (Chater Camera supplied two lenses, but the rest of the gear came from Videofax.) Our talent was Colin Stuart, who appeared in both my WEAVE and Facebook projects.

Our goal: to shoot a mysterious man walking through San Francisco, surrounded by shapes that are, at first, completely unnoticeable; eventually the edges of the shapes start to glow and we realize that our character is constantly surrounded by them. The shapes will take on meaning after the visual effects are completed, and I won’t ruin that surprise until that project is competed. Ian and I scouted three locations, two of which we used: The Embarcadero, a street and promenade that winds its way along the east side of San Francisco, along the piers; and Treasure Island, a man-made island built by the military halfway between San Francisco and Oakland. Our third location, on Telegraph Hill, offered a spectacular view of the Transamerica Building. Unfortunately the fog was so heavy that night that the building was completely obscured, so we wrapped after shooting at Treasure Island.

Oh, and we shot entirely without permits. Which wasn’t hard as we basically looked like a news crew with a few more people and a couple of additional cases.

The entire piece was recorded to SxS cards in ProRes4444 and LogC. LogC is a Cineon-based log curve that captures everything the sensor has to offer and makes it available for grading. Log curves are not raw, but are actually more efficient than raw as storing raw linear data wastes quite a lot of space. Log gamma encoding captures more than enough information to stand up to very aggressive grading. (I hope to write an article in the near future specifically about ARRI’s LogC implementation.)

The camera was rated at EI 800, and the shutter varied between 180 and 270 degrees. The white balance was set to 3200k and never budged. The frame rate was 23.98p. I’m going to talk about the shots in the order that they appear in the montage, not in shooting order.

The first shot in the piece occurred well after dark, and not long before our move to Treasure Island around 10:30pm:

Close-up of the Bay Bridge from The Embarcadero near Harrison St. This was a grab shot on a 200mm Nikkor lens, shot wide open at T2. This is the graded version.

This is the same shot in LogC, ProRes4444, ungraded.

Just before our location move we grabbed a number of shots, including one where we shot through the window of a local restaurant. We just wanted to see what we could get. I didn’t bother with a meter reading as there was no point. At one point earlier in the evening the fog read T0.5 reflected. This was shot with a 270 degree shutter.

Talent Colin Stuart in front of the Bay Bridge. 85mm Super Speed, T1.3, 180 degree shutter. Bright side of face read T1.4 incident. Fill is from ambient street lighting. This is the graded version.

This is the same shot in LogC, ProRes4444, ungraded.

This was the first shot that we lit all evening:

That’s gaffer Simon Sommerfeld holding a LightPanel Micro. Most of the light in the shot came from the left side of frame, so I asked Simon to pop up the frame left side of Colin’s face for a little more contrast.  I’ve found great success in following the direction of the ambient light and enhancing it, rather than fighting it, and using the Micro from the left of frame hid the fact that it was a separate and additional source. It increased the contrast on Colin’s face by popping the existing highlights as the ambient streetlight tended to be a little flat.

Left to right: Leigh Blicher and Michael Bravin observe while I look through the viewfinder, Paul Marbury pulls focus between Colin Stuart and the bridge, production assistant Jim Feeley conveys instructions to Colin via radio, Simon Sommerfeld holds the LightPanel Micro, Colin stands on a lens case for height, and Ian McCamey directs from near the camera position.

Around this time I took a few meter readings and discovered that the bright sky under the bridge, lit by the Port of Oakland in the far distance, read T0.7 5/10 and the fog over the bridge read T0.5 1/2. I was a bit baffled because everything looked brighter through the viewfinder, and on the on-camera monitor, than my meter said it should, but at some point I realized that the Alexa was showing me things my meter couldn’t even read.

This is a situation were a large HD monitor would have been very helpful. I couldn’t judge the quality and direction of the fill by eye as it was incredibly dim, and the small on-board monitor didn’t allow me to see a lot of detail. At T1.3 and EI 800 a good-sized monitor is necessary to see what kind of image you’re really capturing as the monitor will be brighter and crisper than what the eye can see on its own.

On a real production I would have flagged a certain amount of the fill off of the right side of Colin’s face for a more dramatic feel. My sense is that night/exterior cinematography with this camera will be more about the grips than the electricians: the crucial aspect of lighting will be removing the sources that you don’t want, and slightly enhancing the ones that you do. That’s not to say that an electrical crew won’t be necessary—far from it!—but we’ll be doing more with fewer lights and spending additional time removing sources rather than adding them.

Handheld with an 85mm lens at T1.3, 270 degree shutter, pulling my own focus.

This is the same shot in LogC, ProRes4444, ungraded.

I’m going to group this shot with a couple of others that we shot around the same time:

Handheld, 85mm lens at T1.3, 270 degree shutter, pulling my own focus.

This is the same shot in LogC, ProRes4444, ungraded.

Same shot as above, but walking with Colin toward the bridge.

Same shot as above, ungraded LogC.

This shot generated some concern on the Cinematography Mailing List as a number of people assumed that they were seeing “jello-cam”: as the camera moves abruptly the out-of-focus lights in the background appear to smear and change shape, as one might expect with a rolling shutter. I stepped through these frames one by one and determined that the effect seen is not that of rolling shutter, but a combination of using a 270 degree shutter and cat’s eye vignetting:

Figure 1: In the frame above we can see some motion blur from the 270 degree shutter, and a number of the lights are ovals instead of circles. The ovals are due to cat’s eye vignetting, where out-of-focus highlights near the edge of the frame are cut off by the aperture opening. Click here to read the best explanation I could find for this phenomenon. Keep an eye on the highlight toward the left of frame, next to Colin’s ear, as we’re going to watch what happens to it over the course of the shot.

Figure 2: The above frame is relatively still and occurs during an abrupt change in camera direction. The motion blur from the 270 degree shutter is gone, and the bright highlight that was vignetted before is no longer cut off as it is now near the center of the frame, where the effect disappears.

Figure 3: This time the bright highlight near Colin’s ear is shaped differently to the first one. It has resumed its cat’s eye shape by drifting left, away from the center of the lens, and the shape is oriented differently from the first image above because the highlight falls above the horizontal mid-point of the lens. In figure 1, above, the cat’s eye shape is different because it falls below the mid-point of the lens.

A line drawn through the short axis of these cat’s eyes passes through the center of the image. The cat’s eye vignette increases in strength the farther it gets from the center of the frame, and the short axis is always oriented toward the middle of the frame.

The arrows bisecting the cat’s eye don’t line up perfectly with the lens center but you can clearly see what’s going on. The combination of motion blur and the changing shape of the bokeh can give the impression of jello-cam but I see no evidence of rolling shutter artifacts whatsoever. The cat’s eye vignette is lens-dependent, and these old Zeiss Super Speed lenses appear to be the culprit. The vignette is easily removed by stopping down a few stops… which is not going to happen when shooting entirely by the light of existing street lamps. (The light on Colin read T1.4 incident when he was directly in front of one of the streetlights.)

If you’re curious as to why the shot was so bouncy, that’s because I was operating it like this:

Left to right: Director Ian McCamey, camera assistant Paul Marbury, me, gaffer Simon Sommerfeld.

Let’s take a quick trip over to Treasure Island:

Colin in front of the Bay Bridge, on the edge of Treasure Island. Lens is the 200mm Nikkor T2, at T2. The “key” on Colin’s face is T2 incident.

The same shot, ungraded, in LogC.

This was shot around 11:15pm on Treasure Island. We’d shot some scenics farther up the road as I was paranoid about moving near the guard shack that controls access to the rest of the island, but the light was better here and we needed to shoot past the end of a wall that threatened to obstruct our wide shot. We worked quickly and quietly so as not to wake the guard.

Here’s the wide version:

Wider shot, same location. Lens is an 85mm Super Speed at T1.3. 180 degree shutter. The “key” on Colin’s face is T1.4 incident.

The same in LogC.

The bright light in the lower right of frame is the stadium lighting in AT&T Park. Apparently there was a game that night.

Here’s the extent of our lighting:

Simon Sommerfeld lights with the trusty LightPanel Micro.

Once again we’re following the direction of the ambient lighting, which is coming from a parking lot on camera left. The LightPanel Micro is making the bright side of Colin’s face a bit brighter to “pop” it a bit. If you’re curious about the fill level…

Me, reading the ambient fill level.

A blowup of the above, showing what my meter reads.

Yikes. Colin’s face almost looks flat-lit in the LogC version of the wider shot. It’s a little crisper on the close-up as that lens only opens up to T2 instead of T1.3, so the fill is a stop darker than it was on the wide shot. It’s amazing what one small light can do, especially if you’re using it to augment the lighting that’s already there. (We moved the Micro in on the close-up to build the “key” level from T1.4 to T2 incident.)

Once again, notice the cat’s eye vignetting on the highlights in the background near the left and right edges of frame.

More low-light goodness can be found on the next page…

Filed under: Cameras •Tips •Training •

Next Stop: The Last Stop! RED MX Latitude Tests

Here’s the test we shot with the camera rated at EI 800:

RED MX, EI 800, Camera RGB, RedGamma.

This is a cropped image from an Apple Color screen grab. I used Color for this test because I could interact directly with the R3D files. I opted to work in Camera RGB in order to see what was happening directly in the sensor’s color space. The camera was set to 5600k preset and I did no further white balancing. You can see the green hue from the spike in the LED illuminant. (The light source is a one-off prototype that will not ship with other Ambi-2 units until the spectrum has been perfected. The light is generally daylight-balanced but has some extra green in it. DSC is sending me a kit to convert this particular Ambi-2 into a tungsten-based unit for future tests.)

It’s important to know not to count that first wedge, the brightest one on the left. As the only solid reference point we have when evaluating a sensor is the point at which the photosites “saturate,” or clip, that first step is exposed so that it just barely clips. Starting with the next step down, the second wedge from the left, I can see 11 further steps and possibly part of a 12th.

It’s pretty hard to see this in RedGamma, so let’s take a look at RedLog instead:

RED MX, EI 800, Camera RGB, RedLog.

RedLog boosts the blacks up to about 20% on a waveform monitor, which makes it a bit easier to see how far down into the murk the wedges drop. Step 12 is directly above the “12” in “2012”, and there’s a very faint 13th wedge barely visible just to the right of that. And, deep in the noise, there’s another step just barely visible beyond that.

That seems to imply 12 stops of reliable latitude, which a couple of barely visible stops beyond that. Hmmm. That’s pretty impressive. Let’s take a look at what the objective data says:

RED MX, EI 800, Camera RGB, RedLog.

We’re looking at luminance only because the green channel tended to clip ahead of the others in RGB parade, due to the green spike in the illuminant. Remember, skip the first step as we’ve clipped that to set an objective starting point. Middle gray and black are much more difficult to define, and are often defined for us by the camera manufacturer based on how they think their camera looks the best. (The RED trusts us to do that on our own, according to our tastes and needs.)

Skipping that top step, I can see 12 steps total. The 13th is barely visible if you really look for it, as is the 14th step—but since they appear to have the same brightness level it’s hard to know whether to count them as they show up as the same tone, instead of separate distinct tones. The last step is just a scan line brighter than black, so I’m going to say that it’s probably too far down in the mud to be of real use. Based on this waveform I’d say that there are 11 solid stops of really useful latitude visible. If there’s detail in those steps it will be barely visible, and both steps will appear to be the same tone.

Just to double-check that, let’s look at the same image in RedGamma, which will push the shadows back down toward black the way we’re used to seeing them:

RED MX, EI 800, Camera RGB, RedGamma.

Now I’m only counting 10 stops. Stops 11 and 12 are visible as lines just above black, but they are of equal value so there won’t be much tonal difference between them using this gamma curve. Still, they’re down there if they are needed—although I suspect that you wouldn’t want to pull them up to make them more visible as they’d be awfully noisy so close to the noise floor.

Just for fun, let’s crank this thing up to EI 2000 and see how many steps we can detect:

RED MX, EI 2000, Camera RGB, RedLog.

RED MX, EI 2000, Camera RGB, RedLog.

The waveform shows us that the second wedge from the left is now effectively clipped, so we’ll ignore that one for now. If I look carefully at the image itself, above the waveform, I can see… wow, maybe 15 wedges from the left! Subtract the top two wedges and we’re down to 12 stops, which is consistent with what we saw earlier, at least visually. The thing that disturbs me about this image, though, is that steps 12 and 13 are very similar in tone, and steps 14 and 15 are basically the same very dark tone. If the steps don’t show up as separate tones, well… it’s hard to say whether they count. There will most likely be some detail in tones 12 and 13, along with a little bit of contrast, but it’s questionable how useful steps 14 and 15 will be in a real-world situation.

If we look only at the waveform, and start counting after the first two clipped wedges, I can see 10 distinct steps. The 11th is a tiny bump on the noise floor that slopes down into the remaining tones that we could see visually, but not terribly well. Keep in mind that this is RedLog again so the blacks are artificially boosted; if we look at RedGamma we’ll see something very different:

RED MX, EI 2000, Camera RGB, RedGamma.

RED MX, EI 2000, Camera RGB, RedGamma.

Starting at the third wedge from the left (the first one that’s not clipped) I see 10 solid steps in the waveform, for 10 solid stops. In the visible image I see another step for a total of 11 visible steps. The second wedge is clipped at EI 2000 but we know that it’s not really clipped at EI 800, so let’s add that back into the mix for 11 really solid stops that are detectable on a waveform monitor, and 12 that can be seen by eye.

I tend to be very conservative when it comes to exposure, dividing the dynamic range of a camera into “paycheck stops” and “gravy stops.” Paycheck stops are the zones upon which you can reliably place information that will definitely be reflected in a normally graded final image, and which won’t cost you your job when something that’s supposed to be visible there doesn’t show up in dailies. Gravy stops are the zones that will probably show up but that I don’t personally count on: if they are there, great; if not, but I don’t rely on them. Based on this model I see 11 paycheck stops and a further 3 gravy stops that aren’t visible on a waveform monitor but that are somewhat visible when viewing an actual image.

If I want something completely black I would expose my shadows below those last two gravy stops. Or I could very easily crush them in post as they are barely visible anyway.

The original RED ONE M sensor clocked in at around 9.5 stops of latitude, so 11 stops of solid latitude is a healthy improvement. The real treat is the increased speed of the camera. At low light levels contrast is relatively easy to control, as shadows seem to fill themselves in. There are many different explanations for this, including the theory that shooting wide open on a lens results in reduced contrast due to a certain amount of added lens flare, but the bottom line for me is that it’s more important to have a fast camera than to have one with 13 stops of latitude. (Although I wouldn’t turn one away…)

By looking at the images above it should be fairly simple for you to decide how you want to use this camera. Latitude isn’t just a technical choice, it’s an artistic choice—and artistic choices vary based on the person making the choices.

Disclosure: DSC Labs provided the charts used for this test at no charge.

Art Adams is a DP who lives in the shadows. His website is at www.artadams.net.

Filed under: Cameras •Production •Tips •

IR Cheat Sheet, Updated to Version 2

Filed under: Cameras •Production •Tips •Training •

Arri Alexa and Far Red: A Problem That’s Already Been Solved

Lets look at some pictures so you can see what I’m describing. These images were shot under tungsten light but with an 80A daylight-conversion filter in front of the lens. Far red is primarily an outdoor daylight issue because that’s when we’re most likely to use ND filters. (I did a quick test under tungsten light and there was no appreciable difference in response, but you never know. Far red and IR contamination in the RED ONE camera show up as red under daylight but blueish red under tungsten light.)

These images were recorded at EI 800 to ProRes422. I used Final Cut Pro’s three-way color picker to set a white and black balance: white was the sheet of paper on the right, and black was the black fabric at the top right of frame. I looked at all the filters through the camera first, and then shot them slightly out of sequence, so please pardon the messy on-screen notes.

Here’s my IR fabric chart with no ND. The blacks are lifted to about 20-30% on a waveform monitor, depending on the fabric patch, which overemphasizes whatever effect far red has. The top middle patch, “Quilters Solids,” shows a little bit of far red contamination, but under normal conditions this fabric would be rendered so dark that this tonal shift would probably be unnoticeable.

Here’s the same chart with a Tiffen ND .30 filter. We can definitely see some red in the middle two patches, with a little red visible in the left column of patches as well. The two patches in the right column are “controls”: they tend not to reflect far red and always remain black. (Note the variety of fabrics that do, and don’t, reflect far red, depending on either the type of fabric or the dyes used.)

With a Tiffen ND .60 filter the far red contamination becomes even more apparent. Note that this is not an overall color shift introduced by the filter, as only some of the blacks are changing color.

An ND .90 filter reveals some fairly serious changes in color. This is more-or-less consistent with what you’d see under the same circumstances with the Sony F35, EX1 or EX3 cameras.

I then put an ND 1.2 (four stop) ND filter in the matte box, to further stress the camera, and started running through hot mirror filters. Hot mirrors use dichroic coatings to very precisely eliminate specific wavelengths of light. Heavy dichroic coatings change color when viewed at an angle, which is why some of the filters with the heaviest coatings, that work the best against IR and far red, are unusable on wide angle lenses as they’ll turn the edges of the frame cyan, where the angle of view through the filter is the most extreme.

This is the Schneider Tru-Cut 750. This filter works great on the RED ONE, but not so good on the Alexa (or the Sony F35/EX1/EX3). Its light dichroic coating cuts too high in the spectrum, way above where the Alexa’s problems lay.

The Alexa seems similar to the Sony F35 in that the F35’s on-sensor IR cut filter is already excellent; adding a hot mirror with a higher cut results in no change because the on-sensor filter is already doing an excellent job. And the issue isn’t a sensitivity to infrared (heat), but a sensitivity to red on the edge of the visible spectrum.

Here’s the Tiffen Hot Mirror. This is also an excellent filter for the RED ONE, but its cut is still too far above the Alexa’s on-sensor cut.

This is the Formatt Hot Mirror. Its dichroic coating is a little heavier, as seen in a slight cyan shift around the edges of the frame when using a wide angle lens, which makes it less friendly artistically but better at cutting far red. Its lower cut removes a little bit more far red than the Tiffen Hot Mirror does.

The reason hot mirrors don’t work on this camera, or on the Sony F35, is that they are designed to cut IR above 720-750nm. The Arri Alexa and Sony F35 have a hot mirror on the sensor that cuts at around 700nm, so if we add a hot mirror that cuts above the range of the hot mirror on the sensor we’ll see little or no change in IR/far red contamination. A 750nm hot mirror won’t do anything when combined with the 700nm hot mirror mounted to the sensor, because the 700nm hot mirror has already done all the work. Clearly the problem is from far red below the 700nm cutoff—and there’s a good reason that camera manufacturers don’t use an on-sensor dichroic filter that cuts lower than 700nm. More on that in a second.

And now, the filter you’ve been waiting for: the Tiffen T1:

All gone. That’s pretty amazing, especially when you compare it to this filter, the gold standard of IR cut filters:

This is the Schneider Tru-Cut 680. You won’t see many of these around because, while they work brilliantly, they use a very, very heavy dichroic coating that vignettes obviously on even slightly wide lenses. It’s a great reference filter but not very usable in the field.

What is so interesting about the Tiffen T1 is that it is NOT a hot mirror (dichroic) filter. Remember, the Arri Alexa already has an excellent hot mirror filter on its sensor already, so unless we want to add a heavier dichroic filter in front of the lens and risk color shifts around the edge of the frame, hot mirrors are not the way to go with this camera. What Tiffen seems to have done is to create a color dye filter that passes most of the visible spectrum without passing the one sliver of spectrum that causes these fabrics to change color. That’s a pretty amazing feat when you consider that dye filters deal in subtractive color: they absorb colors that are not their own, and pass the rest.

The Tiffen T1 filter.

How Tiffen managed to create a filter that absorbs that one little bit of spectrum while passing the rest is beyond me, but they did it without using dichroic coatings which means no risk of color vignetting. Use this filter just as you would any other color filter.

I asked Arri’s Michael Bravin why they didn’t just put a 680nm cut filter on the sensor and wipe out this far red contamination once and for all. His answer:

“Flesh tone. Realistic and pleasing flesh tones require some far red in order to look natural.”

Eliminating all far red through the use of a very low cut dichroic filter on the sensor can have a devastating effect on flesh tone. The T1 rebalances colors to prevent far red from overwhelming far red-reflective fabrics and dyes while protecting and preserving flesh tones.

If you’ve already invested in 4x5.6 Tiffen T1 filters, or a set of Tiffen IR ND filters, for use with the Sony F35 or Panavision Genesis, you should be good to go with the Arri Alexa. If you’d like to learn how to preview fabrics for IR/far red contamination without actually looking at them through a camera, click here.

You’re going to like this camera. A lot.

Disclosure: nearly every filter manufacturer in this article has sent me filters, for free, to test, and they’ve allowed me to keep them.

Art Adams is a DP who, under certain circumstances, really likes to see red. His website is at www.artadams.net.

Filed under: Cameras •Production •Tips •

RED MX IR Tests: The New Sensor is Similar to the Old Sensor

Filed under: Cameras •Post Production •Production •Tips •Training •

Step into the Matrix: What I Learned from Examining RED’s Build 30 Color Science

This is a DSC Chroma-du-Monde chart, which is probably the best designed broadcast color chart ever.

This chart’s design makes it exceptionally easy to discern how a sensor sees color just by looking at a parade waveform. The bottom left corner of the chart contains a green chip, and the column that extends upward from that chip contains green plus increasing amounts of red, resulting in an even split between green and red—yellow—at the top of the column.

Moving right from yellow sees green diminishing and red remaining constant until we reach the middle chip, which is pure red.

From there blue is added until we reach the far right chip, which contains equal amounts of red and blue (purple) and from there to the bottom right corner of the chart red diminishes until we reach a chip of pure blue.

Moving left from blue along the bottom row sees green gradually increase until we reach the center chip–cyan–and then blue decreases until we reach the green chip.

DSC Chroma-Du-Monde chart viewed under tungsten light in RedColor color space. White balance is tungsten preset. Pulled from Apple Color. Note the “arms” on the green and blue channels where they respond to seeing, or not seeing, their own color.

RED CHANNEL: The beauty of the Chroma du Monde chart is that we can see, very specifically, how the camera’s color channels respond to color. In the case of the red channel we can see the waveform peaking where it sees red, and dipping where it doesn’t. For example, look at the blue chip and see how low its trace is: that’s because the blue chip has no red in it. The red chip’s trace is fairly high, as is the entire top row which contains the same amount of red mixed with other colors (green on the left, blue on the right). It’s interesting to note that, in this case, the green chip’s trace is higher than it should be when compared to the blue chip, which implies that the red dye on the sensor may pass some green light as well as red.

GREEN CHANNEL: Notice the “arms” on the left and right side: on the left the waveform peaks because the left column contains the same amount of green in each chip, while it dips over the right column as there is no green (only blue and red) on that side of the chart. Notice, also, how red causes a dip while cyan causes a peak, because cyan contains green and red doesn’t.

BLUE CHANNEL: Same thing again, only in reverse: the waveform trace peaks where the chart contains blue and dips where it doesn’t. Thanks to the chart’s layout the blue channel is almost a mirror image of the green channel.

If a color channel sees too much of another channel (color “crossover”) the overall colorimetry of a camera can be compromised.

What I discovered, back when I was trying to discern how RED was able to reduce blue noise levels so drastically in the RED ONE “M” running Build 20, was that the blue filters on the sensor pass a lot of green light as well. First, here’s a 5600k chart viewed with a RED ONE, M sensor, Build 20:

The circle shows that there’s a nice dip in the blue channel where the green/red column is, which is normal and expected. Here’s the same chart viewed under 3200k light:

This waveform shows the blue channel responding to blue where the chart contains only green and red. From this I was able to surmise that the blue filters on the sensor’s photosites passed a little bit of green along with blue: instead of being a “pure” blue the filter is more of a greenish blue:

This is a dramatization, but it communicates the general idea. Filters operate by absorbing wavelengths of light that are not the same color as the filter, so a blue filter absorbs, and eliminates, any wavelengths of light that are not its hue of blue. If a filter is greenish-blue instead of “pure” blue, then it will pass mostly blue along with a little bit of green. (The exact wavelengths passed by these filters vary from manufacturer to manufacturer and are closely guarded secrets. There’s very little agreement as to what wavelength is a “pure” hue of any color, so different manufacturers have different sensor “recipes.”)

The photosite underneath the filter has no idea what color is being passed, as it can only count how many photons hit it: silicon alone can’t determine light color, only that light is present. A digital camera’s processor “knows” what filter covers each photosite, so by counting the number of photons hitting a photosite and referencing what color filter covers it, the processor creates a value that represents how much color that photosite “sees.”

For example: If photosite number 2745 generates a signal that tells the processor that it is detecting some light, and the processor checks its directory and sees that photosite 2745 is covered with a blue filter, then it will route the signal from that photosite into the blue channel.

The resulting numbers from all the photosites are then processed via a de-Bayering algorithm to calculate red, green and blue values for each pixel, even though each photosite can only detect one color.

NOTE: Photosites and pixels are completely different things. Photosites are the individual light-sensitive points on a sensor, while pixels are “picture elements” derived from photosite data. The number of photosites and the number of pixels in an image don’t have to match; for example, the Sony F35 uses clusters of six photosites (two rows of red, green and blue photosites) to create a single pixel.

Before Build 30, colors that contained green looked dull under tungsten light, although not under daylight. My theory was that the enormous amount of blue in daylight overwhelmed the small amount of green that’s passed to the blue-filtered photosites, so the small amount of green light passed by the filter was effectively overwhelmed and buried.

Under warm tungsten light, however, there’s so little blue that a small amount of green light made a much bigger difference as there wasn’t enough blue light to drown it out. Since a photosite has no way of knowing, on its own, whether it’s seeing blue or green light, it tells the processor that everything it sees is blue. The processor dutifully adds blue to the areas in the image where the blue photosites register light, even though some of those areas are green.

And what happens when blue is added to a bright color like green? It becomes a dull color.

When I looked at the RED MX sensor in RedColor color space I saw no indication that this was still a problem. The MX’s color actually looks brighter and richer under tungsten light than under daylight, which was a bit of a surprise to me. Greens were particularly vivid. Maybe, just maybe, RED changed the filters on its MX sensor such that the RED ONE saw much more accurate color under tungsten light. Certainly the new vibrant colors have led many to think so.

What I needed to see was something close to a “raw” image off the sensor. Camera RGB is simply an image in sensor “color space” that is white-balanced but has no additional color processing to match it to a viewing device (such as a Rec 709-compliant monitor.)

Same chart as above viewed under tungsten light but using Camera RGB color space. Note the the left “arm” of the blue channel is now raised, not lowered.

See anything familiar? Yup, it’s the same problem I wrote about before: the blue channel sees green, and quite a lot of it. I was a bit baffled when I saw this. I hadn’t studied Camera RGB for my previous article on this subject, only RedSpace, and I suspect I should have paid more attention to it at the time.

I hadn’t seen this green-blue issue in RedSpace under daylight conditions in my previous test, and it occurred to me that I should take another look at daylight in Camera RGB and see what popped up.

Wow. The crossover issue is there as well. It’s not quite as bad, but it’s definitely there. Here are some pictures that show the difference between Camera RGB and RedColor and under tungsten and daylight. These were shot on the MX sensor:

Camera RGB, daylight. Note that all the colors that contain green, from just left of the blue chip all the way around the left side of the chart and just short of the red chip on top, are muddy due to blue contamination.

RedColor, daylight. The blue cast is gone.

Camera RGB, tungsten light. The green chips are REALLY muddy and dingy under tungsten light, much worse than under daylight.

RedColor, tungsten light. The chart is more saturated and rich than it was under RedColor in daylight

There’s no sign of the blue-green crossover problem in RedColor. It’s completely gone. It appears that the MX sensor does have similar colorimetry to the RED ONE M if one looks only at Camera RGB, but why does Build 30 make both the M and MX cameras look so much better? Could RED truly be fixing this issue solely in software? And if so, how? A possible solution lies on the next page…

Filed under: Cameras •Lighting •Production •Tips •Training •Visual Effects •

Lighting Against The Wind: The Making of a Mime Music Video

Actress/mime Nanishka Camberos and director Ian McCamey have a brief chat before reverting entirely to hand gestures for the rest of the day.

We decided to shoot this project on Canon 7D cameras for highly technical and artistic reasons: we could get two of them for free.

I’d shot one prior project on the 7D and thought it was a decent, but not great, camera. Even after turning the camera’s saturation down considerably the colors were too punchy and odd for my taste. I’d heard great things about Steve Shaw’s gamma curves for the 5D and 7D, and as I’d used some of his other gamma curves (primarily on Sony F35 projects) with great success I thought I’d give them a try. At $40 for a three-curve package I could hardly go wrong.

Makeup and wardrobe stylist Maria O’Reilly prepares Nanishka for mimery.

The curves worked very well. The 7D is a somewhat contrasty camera and I was concerned that I’d have a hard time holding highlights, given that Nanishka’s mime makeup consisted primarily of white grease paint and that the background was meant to be nearly black. I used the lightest curve possible and the highlights held wonderfully. The colors were rich and natural without appearing artificial, the way they had on my previous 7D shoot.

Director/editor Ian holds up our makeshift time code slate: his MacBook Pro playing a Quicktime movie that contains both the song audio and a full-screen time code reference.

The biggest technical hurdle was finding a monitor that accepted an HDMI signal, as that’s the only live video signal the 7D outputs. Adam Wilt was kind enough to loan us an Ikan monitor prototype that he was reviewing, so we had a live but not very accurate color feed. Ikan monitors tend to be the “cost effective” monitor choice, but unfortunately cost effectiveness always comes at a price: proper viewing required us to be directly in front of the monitor, as off-angle viewing resulted in dramatic color and brightness shifts. It was quite a chore squeezing myself, a director and a stylist within an approximate 10 degree viewing angle.

We used Canon’s EOS utility to control the camera, and view a 1fps live image, on my Mac G4 Powerbook. (Apparently 1fps is the best one can expect over a USB cable.) For some mysterious reason Canon provides this software only to purchasers of 5D and 7D cameras, and we weren’t guaranteed access to the installation disks that originally came with either of the cameras we used. Fortunately a friend sent me this link and I was able to easily download and install the utility. Here’s hoping that Canon recognizes, soon, that those of us who rent their cameras need access to their software too.

Camera intern Ted Allen and I discuss deep and pithy things in front of our monitoring station. On the left is my G4 Powerbook running Canon’s EOS Utility; on the right is the Ikan monitor we borrowed from Adam Wilt.

The EOS utility made it possible to install our custom camera curves as well as control all exposure and look functions remotely. This allowed us to lock off the camera for our visual effects shots and roll it without having to touch it. It also gave us an additional reference image to use alongside the Ikan monitor, and by averaging the two we had a good idea of what the recorded image looked like. (The Ikan monitor was great for viewing motion, and the laptop worked very well for viewing accurate color.)

Ian borrowed one camera from a friend, and our second camera came from Ted Allen, a film student who has been interning with me for the last year. One camera was locked off for the visual effects shots while the other was free to shoot closeups, although never at the same time. We used stock 28-135mm zooms on both cameras, with an f-stop range of f3.5 (wide) to f5.0 (tight). Even though the f-stop changed when we used different focal lengths (we always lost some f-stop when zooming in) I didn’t see any obvious exposure changes in the image or on the camera histogram.

One big surprise was that the 7D didn’t seem to be as fast a camera as the 5D. I had to set the camera at 1/30th sec. shutter and ISO 320 just to get a decent exposure. This played a little bit of havoc with the green screen, as motion blur was enhanced, but Ian said he could work with it–and he did. During the shoot we also saw a fair amount of noise in the image, both on the Ikan monitor and on the laptop, but that noise disappeared in post so it seems we were mislead into thinking there was more noise in the image than there really was by the 7D’s monitor outputs.

The lower part of the background was covered by green cards as one of Nanishka’s characters was meant to lay on the bench while watching her other selves interact in the foreground. In the end this character was dropped and replaced with a clean background plate as Ian decided the story was stronger without the third character.

At one point in the song Nanishka hands herself a pair of sunglasses. This was a tricky maneuver as the glasses had to hit the same point in the same way in both plates. We put a baby nail-on plate on a short C-stand and created a platform for the sunglasses to hit, and then we marked the stand carefully. By using the same stand on the same marks Nanishka was able to take the sunglasses from a PA wearing a green glove, and later hand the same glasses to the PA for the other side of the action, while hitting the same mark each time.

Nanishka prepares to receive sunglasses. The small C-stand next to her is precisely placed so that the sunglasses always land on the same mark, no matter which character has them. Later, in different makeup, she will hand herself the sunglasses using that same platform. Note that the C-stand legs are very carefully marked for position.

Originally we tried using a regular C-stand with the arm extended but realized quickly that it was too big and cumbersome. We had to place and remove the C-stand during Nanishka’s performance, and the short stand was much easier to position without getting in Nanishka’s way. (If she’d blocked the C-stand briefly then that could be fixed through rotoscoping, but if the C-stand blocked her then there’d be no way to remove it.)

Resetting for the hand off. Production assistant Whitney Kahl stands in for Nanishka’s better half. Note her green glove, which will be replaced later with Nanishka’s performance.

If you’ve always wanted to learn to light mimes but were afraid to ask how, turn the page…

Filed under:

A Cine Gear Find: TechScout Touch

Filed under: Cameras •Production •

GearNex: The Next Generation of Gear Head

Filed under: Cameras •Editing •Lighting •Post Production •Production •Training •Visual Effects •

VFX Tell the Story in California State Fair Spots

This was my first shoot in a barn, I’m happy to say.

The barn we chose faces north and south, which meant that if we shot north inside the barn we wouldn’t have to deal with the moving sun. By looking north we were guaranteed to have very little change in the background that we could see through gaps in the barn’s construction, as the exterior in that direction would remain roughly front-lit and overexposed all day long.

We were able to cover the south side of the barn with black visquene and flags, blocking out all exterior light. The crate shots were a series of lock-offs and stray bits of sunlight passing through the barn’s wooden slats and hitting the set would have caused a number of expensive problems in post–problems we couldn’t afford.

Here’s my lighting setup for the interior of the barn:

The 6k PAR, bounced into a 12’x12’ Ultrabounce, created a soft but contrasty base light for the barn interior. The 4k PAR, through a Chimera with a light-controlling grid on it, provided a touch of hair light. The 1200 PAR on the top right provided a nice sunlight kick from the key side, and the lights in the crate provided a magical glow. One remaining 1200 PAR provided a hint of fill from near the lens, bounced off a 4x4 piece of bead board on the key side of the lens.

The 1/2 CTO and 1/4 Minus Green on the 1200 PAR scratch light produced a light salmon color that felt like sunset. A mirror caught some of that light and raked it across the front of the hay bales directly behind the crate. (I frequently use mirrors to try to cut down on lighting costs and setup time, but in the future I think I’ll just add an additional light. Trying to work a mirror into a beam of light without cutting that light off of something important can be a bit of a pain.)

We shot separate passes of the kids crouching down, standing up, and then crouching down again, so that the editor could speed up these actions and make the kids burst out of the crate. The green screen allowed the kids to be manipulated separately from the background. The 4k PAR backlight is visible at the top of frame.

My shooting stop was just shy of T4 on a 35mm Zeiss Super Speed, shooting on a RED (build 30) rated at EI 320. We shot the crate itself in multiple passes.

We created the glowing light between the crate boards by covering the interior of the camera-facing sides of the crate with tracing paper and filling the interior of the crate with open-faced 1k and 2k tungsten lights. Blasting the tracing paper with light turned the paper itself into a light source, which then radiated light through the gaps in the crate’s boards. (This effect was severely enhanced in post.) We also waved flags around the inside of the crate to create shadow elements.

That technique worked great when the crate was empty, but there was no way it would work with people and dinosaurs standing in it. For those shots we replaced the hot tungsten lights with 3200k Kino Flos, rigging them around the inside lip of the crate on all four sides. We took the bulbs, and the metal base that held the sockets, out of the black plastic lamp housing to make the lights lower in profile. They were rigged as shelves just below the lip of the crate using chicken wire.

The crate itself is one pass, and the opening of the crate encompasses numerous other passes.

This is Poppy, the California State Fair mascot. This shot caused us a little bit of trouble as we’d tried to frame for our largest element from the start, as all the crate shots were meant to be locked-off for visual effects, only to discover that our largest element was taller than we’d expected. Poppy bumped the top of frame. We solved the problem by making the frame bigger: our project format was 4K HD, so we switched to 4K 16:9 for this one shot. That gave us just enough room to squeeze Poppy into the frame without breaking our lock-off shot.

Poppy’s yellow fur became quite saturated under tungsten light so we switched off a few of the interior Kino Flo bulbs. That’s my one complaint about RedColor: right out of the box the colors are very, very saturated. (More on Build 30 color later.)

You may have noticed that nearly every shot in this spot has a handheld feel to it. Ian likes rough camerawork because it sells the visual effects better than a lock-off does. It wasn’t so long ago that all visual effects shots had to be locked-off, so the freedom to track objects inexpensively adds production value to low budget VFX projects like this one. Originally Ian wanted to create the handheld movement entirely in post, and I’ve seen him do that kind of thing very convincingly in the past, but I had another thought: if we placed a couple of reference markers in the shot and then spent a minute or so shooting them handheld, it should be possible to lift the motion from that shot and lay it on top of another.

This seemed like it might result in more natural camera movement as well as eliminating the need to create camera shake from scratch. In order to do this we needed a larger frame to work with. Here’s what would have happened if we’d tried to rotate the above shot, of the kids on a roller coaster, without shooting tiles to extend the background:

By shooting four different “tiles,” or shots, including the crate but shooting more of the background than could be seen in the 4K frame alone, we ended up with a “Super Image” as a background:

The light area shows the 4K frame with plenty of room to move around inside the grayed-out 8K background plate.

After shooting all the crate elements I broke the camera loose and shot those four overlapping tiles, and then I had the grips put two C-stand-mounted tennis balls in the shot. I then hand held the rolling camera for about a minute while Ian watched the monitor and talked me through a series of moves and reactions. Then the compositor tracked the tennis balls in post and laid that movement across all the other shots, filling the blank spots around the moving frame with the super image created by the background tiles.

The result is a series of locked-off visual effects shots that look as though they were actually shot hand held in real time. As a bonus, Ian discovered he had plenty of room to rotate the crate for the roller coaster shot. That wasn’t part of the original plan but was a very nice extra.

This is a similar shot that really shows off the lighting. The fill from near the lens really did more for the crate than it did for the people, and if the script didn’t call for a glow emanating from the crate I’d probably use a bigger fill light. In this case, though, the soft sidelight and the tungsten light in the crate blend nicely with the hard kick of “sunlight” from the left of frame. Soft sidelight on its own can be very pretty but that nice hard kick from the key side adds a lot of character.

This is a great example of how keeping the light sources on one side of the frame and letting them blend into each other creates a very natural but rich look. It also shows the power of color contrast, as the HMIs sometimes appear in flesh tones as slightly cool and contrast nicely against the warm light from the 3200K Kino Flos.

I wish I could take credit for the rock and roll lighting in this shot, but I can’t. We didn’t have the time or budget for moving lights on location. The moving lights are entirely a post effect, and are quite well done.

Shooting the dinosaur went smoothly once we got him out of his trailer, but that alone took hours. We had no choice but to stake live chickens to the ground in a trail that led to the set.

But seriously… there was no dinosaur present when we shot this element, and the flying hay is a post effect. It’s brilliantly animated and really sells the shot.

We shot the usual silver ball and gray ball, where the silver ball shows where the light sources are and the gray ball reveals contrast and quality of light. With these two references the “lighting” on the dinosaur can be matched to the lighting in the set.

It took a while to shoot the lid flying off the crate. The grips rigged a rope-and-pulley system overhead and it took us a long while to get it to fly off just the right way. The action was significantly sped up in post. 

The barnyard animals were green screen elements shot the year before by someone else, using a different camera, for last year’s spots. They worked startlingly well.

We had to tone down the interior lights for the jockey, who–naturally–showed up wearing white pants. Fortunately the RED held the highlights very well.

The snap zoom on the crate was done entirely in post and it looks wonderful. I asked Ian why so many projects that reframe RED footage in post look so bad, and he says it’s because they don’t go back to the original 4K file. They zoom in to the image after it has been downrez’d to 2K or 1920x1080 for editing, which results in a softer image. Going back to the original 4K file seems like a no-brainer but I’ve seen it done the other way quite often.

This is the only shot of the crate that was actually handheld. Every other shot of the crate was locked-off due to the visual effects, and the motion was added in post.

One we finished shooting into the barn it was time to shoot the kids in the doorway. This was a tough shot as looking from an interior outdoors into sunlight is always difficult. Competing with the sun is never easy. As we saw the ground in the wide shot I was unable to cheat and put a large net behind the kids. We had to pound a lot of light into the scene in order to bring the foreground exposure up enough to match the background.

The half soft frost takes some of the curse off the hard sun, and the 6K par was probably 6’ or 8’ from the kids through a 6’x6’ frame of full grid cloth. Surprisingly, I didn’t use the full 6’x6’ frame, as filling the frame didn’t give me enough punch. I had the electricians put the PAR through the bottom quarter of the diffusion on the side closest to the camera. I had to keep the light fairly low in order to get into their eyes.

Soft light from near the lens is a very fast way to create very smooth and pretty closeups. Although the size of the source was relatively small, filling only a 3’x3’ patch of diffusion, it looked softer because it was so close to the lens axis. The closer a light is to the lens axis the softer it looks, which allowed me to use a smaller, brighter light source while still retaining a soft feel.

(The boy is the director’s son, who just happened to be perfect for the part.)

I used a polarizer to take some of the curse off the hot sky and to bring color back into the grass. Grass is shiny and tends to reflect blue skylight, which gives it a desaturated look unless a polarizer is used to eliminate the blue reflection.

Here’s what I figured out during the location scout:

I realized that the time to shoot this opening shot was at the very end of the day, when the sun back-lit the kids and gently raked the front of the barn, so I asked Ian and our producer/AD, Tom Ruge, to schedule the shoot such that we could do this shot last.

The good news is that, when it came time to shoot, the sun was in the perfect spot; the bad news is that it was a hazy day and exposing for the barn made the sky blow out into a featureless white.

When Ian and I first tackled color correction (yes, we graded this ourselves in After Effects–it was that kind of budget) he showed me added effects that made the shot a lot more interesting. He had inserted a blue sky and some artificial clouds (there were no distinct clouds in the sky during the shoot), but this necessitated handmade mattes around the kids as they ran toward that barn and covered the effects. As we tweaked things further it became clear that once we’d made our grading changes Ian would have to completely redo the mattes, which is a very time-consuming process. Instead I came up with a cunning plan.

Over the course of my career I’ve discovered that there are some shots that are greatly improved by using the most obvious grads possible. This doesn’t work for every shot, but once in a while it’s the perfect solution. In this case I suggested a diagonal blue grad across the top of frame and an orange one across the bottom:

This played very well and didn’t require any matting. The diagonal where the blue grad transitions into the orange one frames the kids nicely, and the fact that the orange grad trails off on the right side of frame helps keep the eye focused on the left, where the kids are. By raising the clouds above the kids Ian eliminated the need for further matting. (There was also a power line running to the barn that Ian removed.)

This was a 16’ move on a doorway dolly sitting on track. I handheld the camera in my lap, and we rolled at 30fps to give the shot a slightly dreamy feel. We tried 36fps initially but that turned out to be way too slow.

I’d dropped into 3K to do this shot at 36fps (the RED ONE won’t run faster than 30fps in 4K HD) and Ian later asked me why I didn’t switch back to 4K HD when we fell back to 30fps. It boiled down to time: changing to 4K HD meant a lens change, as using more of the sensor area meant the lens would appear wider. At 3K the RED was only using a portion of the sensor, so the 18mm lens I used actually looked more like a 25mm lens. Switching to 4K would have meant that the 18mm lens looked like an 18mm lens, and I’d have had to quickly change to a 25mm lens to roughly match the current frame size. I didn’t think it was worth the time as the sun was setting fast, and I really liked the frame as it was.

I’ve got some comments about RED’s new color science on the next page…

Filed under: Cameras •Lighting •Tips •Training •

Everyone Looks Sexy at 1000fps

Filed under: Cameras •Lighting •Production •Visual Effects •

The Tiger Lillies Finally Set Sail

Filed under: Cameras •Production •Tips •Training •

Random Tips from a Professional Camera Operator

Filed under: Cameras •Production •Tips •

Two New Sharp-Looking Charts from DSC Labs

Filed under: Cameras •Lighting •Production •Tips •Training •

The Making of an Epic Media Project

HERBST THEATER, SAN FRANCISCO

We recorded the opening shot of the piece on the third day of the shoot. The theater location had been a difficult one to find and lock, and my gaffer (Luke Seerveld) and I hadn’t been able to scout it. Of all of the locations scheduled, this was the one that gave us the most heartburn. Without looking at it first and formulating a plan we felt it was going to be difficult to get in, light, shoot, and get out on schedule if we had to light the entire location from scratch.

It turned out to be very, very easy, thanks to some help from the house stagehand, “Rhymes with golf” Rolf, who made our shooting experience very pleasant indeed.

The first thing we did was to look at the ambient illumination through the camera. With all the house lights on my spotmeter told me that we had a very good base light level at T1.3. Once we saw that we were able to relax a bit, as this allowed us to focus on little touches as opposed to lighting the entire space. I had a two person lighting/grip crew, plus a dolly grip for specific days when we had a Chapman PeeWee on the job, so whatever we did had to be simple.

Initially I’d thought of putting a bunch of lights—probably 1200w tungsten PAR “firestarters” (so-called because they will if you put something close enough, like a flag or a human head) and some Source 4 leikos—on the balcony. Putting all those lights in one place would have made them easy to power, and it should have been fairly simple to pan them around and sketch out parts of the theater. When I actually looked at the theater, though, it became clear that lighting from the balcony would not do much for the beautiful red seats: backlighting them might have revealed their presence but would have done little to show off their richness, which was one of the reasons this location had been chosen. It became clear that front light was necessary to bring out the plush red seats.

We put our theater lights on either side of the stage, out of the shot, and created some pools of light that washed across the front of the seats. At first we created symmetrical washes, lighting left and right in even sections, but that didn’t really work artistically. We took Rolf up on his offer to light the balcony symmetrically by repurposing some wall-mounted stage lights, but the lower level looked better with “random” washes of light from Source 4 leikos placed on the stage itself.

I’d envisioned the podium lit by a spotlight of some sort, but once I saw that the podium was made of translucent plastic I had another idea. If we’d been able to add a light haze of smoke to the theater I’d have stuck with the spotlight idea as the shaft of light it created would have been spectacular, but as we weren’t allowed to use smoke it made more sense to put lights on the podium itself. I asked Luke to put two 2’ Kino Flo tubes underneath the top of the podium, turning the frosted surface of the podium into a large light source. (I love putting lights in the shot. It not only looks good but saves time because we have to hide fewer lights.)

We put a 4’x8’ bounce card horizontally on the floor to the left of the dolly track to add a little bit of fill.

As a last touch, Rolf re-aimed a high stage light on frame left to rake across the podium, giving our talent a light to walk into as he approaches the podium. This light also worked as a backlight for our other angles. In this sequence of stills you can see the talent gradually walk into that hard sidelight as he approaches the podium:

As the camera dollied toward the podium, Luke gave Rolf a voice cue to bring up the house lights, revealing the expanse of the theater.

Jono yells “action,” Luke cues the house lights, Rick Edmondson pushes the dolly and Bruce McGregor keeps me sharply focused.

Another take: the behind-the-scenes camera captures a Mole 1k in the foreground bouncing into a 4’x8’ card for base fill onstage.

This next shot is one of my favorites:

I used the GearNex gear head for all of the shots in the theater. It gave shots like this one a wonderfully smooth feel. We didn’t change any of the lighting between this shot and the last one.

Production manager Vanessa Tomasello, on the left, documents this shot for posterity. Key grip Jeff Nealon stands in the background with a lenser, keeping the backlight (our former sidelight in the previous shot) from flaring the lens. Visible behind the talent is one of our seat lights.

The Gearnex gear head was a great help for this shot. For some reason it’s often easier and smoother to find the right pan speed by spinning a wheel than by moving a pan handle. Our stand-in is executive producer Steve Weisser.

Jono watches the onboard monitor as we do a quick rehearsal with the talent in place.

While we’re here, let’s jump to the last shots of the piece, as they were shot at the same time as the opening shots:

And here’s the lighting setup, from the behind-the-scenes film:

The Kino Flo is filling from the key side. The podium light was a bit too stark and contrasty on its own, so the Kino Flo lowers the contrast a little and wraps nicely around the talent’s face.

The bounce card on frame right is casting soft light across the front of the curtains, as well as giving the talent a little hint of edge light. The vertical bounce card on the left is casting a shine onto the left side of the talent’s face. Skin is shiny, and rather than hitting it with a hard edge light it’s often prettier and more interesting to reflect a large source in it. That’s why the card is vertical: it’s reflecting in the vertical plane that is the talent’s face. As reflections follow the physics of “angle in = angle out,” a taller source means that the shine will be visible from a greater number of angles—important in a shot that booms upwards and changes the relationship between the angle in/angle out of the light and the camera lens. (We originally tried a boom-up shot but I think we settled on a simple tilt.)

The flag is keeping the talent’s backlight (one of the stage lights rigged on a truss above the far right bounce card) from flaring the lens. One of the hallmarks of a great grip (in this case, key grip Jeff Nealon) is that flags like this appear before I have a chance to ask for them.

The final shot is lit with many of the same lights, only reconfigured a bit. The Kino Flo is now in front of the podium and over the lens, and the light aimed into the bounce card to the right of frame has been panned to rake directly across the curtains.

This was the first time I’d shot the RED under tungsten light without any filtration, and I’m pretty happy with the results.

By the way, that final shot—with the talent’s audio—was the only shot in this entire piece done at 23.98fps. All else was shot at variable speed 30fps for 23.98fps playback.

The key to working successfully with a gaffer is to learn strong communication skills.

Travel with us now to the next location, the BBC, on the next page…

Filed under: Cameras •Lighting •Post Production •Production •Tips •

Low-Budget PSA’s, Shot on RED, Prove that Budget is Not a Barrier to Excellence

Here’s the first shot of “Perp:”

And here’s the lighting setup:

This was shot in the Meets the Eye conference room. The lighting started out a little more complicated than you see here but it didn’t work for me, so I simplified.

The two lights used are Vista Beams, which are very high intensity 8-tube Kino Flo lights with special reflectors. Meets the Eye has two of them, and I love them. They’re punchy big sources that can run off a wall socket. Originally I had my gaffer, Luke Seerveld, set up some pools of light on the table using tungsten Source 4 lights, but for some reason they turned out looking a bit too contrived and video-like, so I settled on one Vista Beam as a soft key light from near the camera and another aimed at the window in the background that was covered with 1000H tracing paper.

Tracing paper is a beautiful diffuse material that was very popular in the 80’s, back before Chimera soft boxes came to rule the industry. Its only downfalls are that it’s hell to transport and it’s fairly flammable. Otherwise it’s one of my favorite diffusion materials ever as it completely removes any trace of specularity from a light source: it becomes a big, evenly lit glowing source that makes everyone and everything look good. (Lee makes a gel called Lee 129, which is a fire-resistant plastic version of tracing paper.)

The fill side of the actor’s face picked up too much light from the white drawing board on the right side of frame so we blocked the ambient spill with a 4’x4’ floppy flag. The high contrast on the actor’s face, in combination with the bright window in the background silhouetting some clothes and reflecting in the table, makes this a very interesting yet simple lighting setup.

I also framed a dark portion of wall in the foreground of the wide shot to help focus attention on the actor. It might not seem obvious that blocking part of the frame with a black object would enhance a composition, but I think it worked pretty well here.

Rating the RED one a EI 320 makes the footage too noisy for my tastes. I rate it at EI 160 for a smoother, cleaner look that pushes the noise down into the blacks. I used a Schneider 1/2 CTB filter in front of the lens for partial correction back to 5600k from 3200k to keep the RED’s daylight-balanced sensor somewhat happy.

As you can tell, we opted for a “shakey-cam” style of operating. Ian likes gritty images, so not only did I shake the camera a bit but we also over-sharpened the image in the color grade. More on that later.

Left to right: Gaffer Luke Seerveld, camera assistant Adam Wilt, me, our sound person. That’s a RED ONE with a RED 18-85mm zoom. We’re shooting into the MTE conference room from the kitchen area.

Originally I tried to color the back window blue, using daylight Kino tubes, but that was determined to be a little too expressionistic and pretty for what we were doing so we switched the 5600k tubes to slightly warm 2900k tubes. (Kino Flo 2900k tubes match tungsten light sources better, as tungsten lights tend to be 3000k instead of a perfect 3200k.)

A Kino Flo Vista Beam lights a window covered with 1000H tracing paper. In the background a Source 4 leiko gives the actor an edge light. I eventually eliminated that light as it was just too much: simpler was better.

On to our next setup:

Originally we set this up as a low shot, following this gentleman from his car in the background to this final position, framing him against the sky. There were a lot of clouds in that part of the sky, though, and the combination of blown-out clouds and dark skin made the shot look really ugly. We raised the camera and shot downward as a compromise to eliminate the over-bright clouds. (Dark skin looks great if it’s brighter than the background. A bright background tricks the eye and makes dark skin look even darker. A “proper” exposure might, under certain circumstances, look way too dark.)

I wanted a little modeling on the actor’s face as he landed in his final position, so as an experiment we set up a Vista Beam with daylight bulbs at the actor’s end position. It turned out to be bright enough to make a difference in open shade. There was nothing I could do about lighting the actor’s starting position, about 20’ away, because I didn’t have a light bright enough to make a difference at the necessary distance to keep it out of the shot, but the end of the shot was where the payoff was so I didn’t worry so much about the beginning.

The smaller a person or object is in the frame the more you can get away with, and it wasn’t that big a deal to forego lighting the deep background as long as I could make the actor look good for his closeup on his end mark. The end of the shot is where I focus most of my efforts, because that’s usually where the the payoff is. (This may explain why I always set dolly shots in reverse.)

It helped that, in the final edit, the early part of the walk was cut out due to time constraints.

The Vista Beam is visible just past the camera. Gaffer Luke has a blade on a stand in an attempt to cut sunlight off the Vista Beam’s silver reflector. Director Ian McCamey is center frame wearing headphones.

Another angle of the same setup. You can see the Vista Beam tubes reflected in the windshield of the car on the right.

While we graded the previous shot to be very neutral, this shot felt like it wanted to be a slightly unpleasant green color. We carried that theme through on the final sequence:

Originally the main actor was going to stand behind the window the entire time, but Ian thought it would be nice if he sat on the desk in the background, stood up and walked into his closeup. It was a very nice move, and it was lucky that I’d already lit the shot in a way that would accommodate this:

The Vista Beam comprised the main source, and you may wonder how I came up with the idea of putting it on the floor. Very simple: it reflected in the window in nearly every other position. I could have placed it above the frame but I was worried the light might look too nice, as light from a high angle tends to flatter faces and this wasn’t a flattering spot. I put the Vista Beam in the only other place that worked.

The Kino Flo from the side created a nice highlight on the side of the actor’s face, which then transitioned into a “normal” exposure from the Vista Beam, which then tapered off into shadow. It’s pleasing without being flattering. It had the added benefit of lighting the actor at his seated position, lighting him through the full range of his move from background to foreground.

We added a 1’x1’ FloLight, an LED light made by Prompter People. Luke consulted on the design of several of their lights and he has a couple in his kit. They aren’t the greatest LED lights in the world, but they’re punchy and cheap and they work great in a situation like this where you need a small, flat, cheap soft light on someone in the background. The main actor actually blocks it as he stands up.

FloLights tend to be a little green, so this one has 1/4 minus green gel on it.

Adam Wilt waits patiently for me to stop lighting things and shoot. The light on the other side of the window is a 2900k-bulbed Vista Beam through a 4’x4’ frame of opal diffusion. (The thin diffusion helped to blend the separate tubes into one smooth light source.) The 4’x4 tube Kino Flo on the left of frame is also bulbed for 3200k. Not shown, outside frame left, is the FloLight.

You may have been marveling at the amazingly adequate color grade given to this piece. Ian and I did it ourselves, at Rough House, in Apple’s Color. There was no budget for color grading, because initially there was no budget for a RED (MTE gave us a very generous deal on the gear and their facility in order to contribute to the cause), so we winged it. I have a little experience grading my own projects, and Ian has been working in post his entire career, so between us we comprised one fairly competent colorist. I think we did a reasonably good job.

That’s it for the “Perp” spot. On to the next page for “Lineup”…

Filed under: Cameras •Lighting •Pre-Production •Tips •Training •

A Tale of Forbidden Love, Shot on RED

We shot this on day two, in a house in the Marina District of San Francisco. It was at the top of two flights of stairs (blech!) but it was a very nice home so dragging all the equipment upstairs was worth it.

Soft light from below the lens feels very “ambient” to me, so when I want to light a room quickly and in an interesting manner I’ll bounce a light off a large piece of foam core placed somewhere below the lens axis. This is a 1200w PAR aimed into a 4’x4’ piece of foam core placed upright on the floor to the left of the camera. The feel I’m going for is that of sunlight coming through a window and bouncing upwards onto the fireplace. I didn’t have the tools or space to create a nice hard shaft of sunlight so I didn’t bother trying; but I had more than enough to reproduce the feel of hard sunlight bouncing around a room.

The lighting setup for this room was pretty simple:

I wanted the light to appear as if it was coming from the window but I also wanted a source with some directionality to it. I didn’t have much of a crew, or enough grip equipment, or even enough space, to properly shape a large light source, so I opted instead to use a smaller more reflective option: a 2’x3’ piece of foam core covered with Reflectix.

I’ve written about Reflectix before: it’s water heater insulation, available at your local hardware store or anywhere online where you might buy supplies for your massive interior pot farm. It’s highly reflective but with a textured surface, so while it’s a very shiny source it casts a soft “messy” shadow. In a case like this, shooting in a small room with nowhere to put a forest of C-stands and flags, using a smaller soft source that cuts itself is a very practical way to work. For example, in order to keep light off the back wall, all one has to do is rotate the card slightly in the opposite direction.

It’s also easy to rig, as the card can be grabbed with a C-stand arm at full extension, keeping the base of the stand out of the shot.

An alternate method might have been to rig a Kino Flo onto the C-stand, but that is a very different quality of light with different fall-off characteristics. I used a Kino Flo in just this way for one setup, but I’ll get to that in a bit.

I remember this as being lit with natural light, but when I look at the right wall it appears there’s a Kino Flo on the floor. I love the actor’s reflection in the picture on the wall. Notice the amount of fill light that is falling onto his back from the top of the bed. Remember that: we’ll revisit it a bit later.

Right now I’m going through a phase where I’m very much trying to control how I expose footage with the RED, in order to make for consistent dailies and ease of color grading. While shooting this project, though, I opted fully for the “expose to the right” method: rather than filling with more light I simply opened up the exposure as far as I could without clipping a detailed highlight and let the ambient light do its thing. Ambient fill has a definite feel and it’s not right for every project, but as the director wanted a very natural and real look, and since we were moving very quickly, this technique felt like the right way to go.

By watching the raw data exposure bar graph and zebras it’s very easy to push your exposure right to the edge of clipping. The bad news is that, while shots within a scene or lighting setup are consistent, exposures between scenes can vary dramatically and require entirely different color grades in post. It did provide the “thickest negative” possible on the RED, which is always good as it keeps shadow detail above the noise floor.

In a scene like this I took the curtains right to the edge of clipping and used the ambient lighting for fill. For this setup this approach worked very, very well. If the lighting is already there then there’s no need to mess it up.

The kitchen was a little difficult to figure out. Originally I tried using an HMI PAR to create the quantity of light that I wanted because I was concerned that the window to the left would become overly bright as the sun moved around to that side of the house. Rather than start off with a small light and then find myself in a bind when the sun started blasting through the window, I tried to use a large light so I could quickly deal with brightness changes.

I couldn’t quite get the look or the control that I wanted by using such a big light in a small space so I ended up putting a 4’x4 tube daylight Kino Flo on the kitchen counter, standing it vertically against a wall just outside the left of frame. You can see a little bit of spill from the light on the kitchen counter.

I pushed the flesh tone exposure as high as I could, for a nice thick “negative,” and let the ambient fill do its thing. The quality and direction of the fill light isn’t perfect, as it would have been if I’d filled from the key side or from somewhere near the lens, but this project didn’t want to look “perfect.” I would have needed a lot of negative fill, mostly on the ceiling, to accomplish that “perfect” look, and I didn’t have the time, crew or resources.

I really like this frame.

The refrigerator in the background is lit by warm bounce off the floor. There’s a 575w HMI in the back left corner of the room bouncing off the warm wood floor to light the dog.

One problem when working with animals is the trainer often has to stand somewhere near the camera to cue the dog, and that can be problematic in small spaces. For this shot the only place the trainer could be was in front of the light source, so we used a Power Window to brighten the dog’s face. The eye light is from the Kino Flo placed on the kitchen counter.

It’s kind of like “My Dinner with Andre,” only shorter and with less conversation.

The arch in the doorway was a blessing for this shot as it gave me plenty of room to hide lights. The director wanted a bleak-but-real look and edge lighting the actor and the dog, with little or no fill, seemed like the right way to go.

When I looked at this shot later I panicked a little as the arch looks off level and I thought that either the camera wasn’t level or there was an issue with the lens sitting properly on the camera. (I’m obsessive about leveling cameras so I thought I was losing my mind!) Closer examination shows that the camera was level but the walls of the building are leaning. It’s pretty clear from the moulding above the background windows that the building is settling down and to the left.

The only fill in the room is from a window on the dog’s side of the table, and rather than block it I opted to work with it. It gave the dog a little more of a “pop.”

Originally I wanted to do two little sconce light effects, one on each of the background posters, but one of my two 300w fresnels died and I couldn’t match the look of the remaining 300w light with a 650w fresnel. Instead I used two Kino Flos to give the walls a little bit of illumination. In the future I’ll probably light walls like this a little brighter, as there were very few bits utilized in imaging that dark tone and it ended up a bit noisy.

I protected the background window highlights just in case we wanted to see some detail outside, but we opted to put a Power Window around them and blow them out in post as that looked better.

I first tried a big source for this shot, but a 6’x6’ grid cloth just off frame right was too big a light for the space. It went everywhere. A 4’x4 tube Kino Flo covered with light grid cloth worked perfectly. We aimed the light away from the back wall in order to give it a little shape. The back wall went orange in post instead of orange-ish red so we added a secondary color correction to fix it. (Bonus trivia: men prefer reddish oranges, while women prefer blueish oranges. It’s a fact.)

The hallway illumination in the background is natural light from the living room.

We did the same thing for the dog’s closeup.

The table gives a real sense of depth as it extends from beneath the lens all the way to the subject, where it catches the subject’s reflection. The flower vase provides a sense of geography and helps to unify the shots.

This shot was lit with two lights: one is a vertical 4’x4 tube daylight Kino Flo down the hall to the right that lights the dog, and the other is a tweenie hidden on the floor to the right side of the door bouncing into an 8 1/2"x11” piece of typing paper behind the toilet. The glow behind the toilet helps set the toilet off from the background, plus I was hoping to get a little bit of backlight on the urine stream. (Yes, I tried to backlight a urine stream. It’s not really urine, it’s water from a plastic bottle the actor is holding.)

The remaining light is from a small window over the toilet.

The predominant light source in this shot is a window over the bathroom toilet. We put a flag in front of the bathroom door, just out of the left of frame, to eliminate that as a bounce surface and make the shadows on the actor’s face a little deeper for this closeup. He’s still getting a lot of fill from the ceiling.

This shot was lit entirely by the Kino Flo sitting vertically on the floor of the hall about five feet out of left frame.

Once again, here’s the lighting diagram for this shot (you saw it earlier at the top of the page):

This lighting setup worked for all of our bedroom shots except for those that were silhouetted against the window. I exposed the scene by pushing the white pillow highlights to the edge of clipping. In the closeup I actually clipped the blue channel a touch but as the other channels held plenty of detail it was completely unnoticeable. (Clipping one of the RED ONE’s color channels usually isn’t a big deal. Clipping two is a big deal.)

At least the scene you’ve been waiting for:

The Reflectix gag described above worked for most of the bedroom shots, but for the shots facing the window I needed a different gag. The director wanted complete silhouette, but I wanted to wrap a little of the window light around the talents’ heads to bring out a little bit of facial detail and give them some depth. Here’s the lighting setup:

I’m simply extending the window “source” around a bit more than normal. Adding sources to a naturally-lit scene, while preserving the natural look, can be very tricky. The best method I’ve found is to simply extend or cheat the existing sources as much as necessary. It’s a little bit like keying from the fill side: I’m extending the wrap of an existing source, and since my second light is coming from much the same direction as the first light they tend to blend and compliment each other rather than conflict.

Here you can see the subtle reflections cast by the Kino Flos, both in the top and bottom of the dog’s snout and the actor’s cheek and chin. We crushed the shadows a little bit in the color grade while boosting the highlights, which gives the highlight “shines” a mild bleach bypass feel.

This sequence was shot at 3k, 50fps for dramatic effect. Originally we smeared peanut butter to the side of the actor’s mouth but the dog’s licking didn’t look right, so the actor put a big dollop of peanut butter on his tongue and said “Leth shoot!” THAT’S dedication to one’s craft.

Never underestimate the power of negative fill. We covered the bed’s white duvet with a blank furniture pad to eliminate a ton of ambience bounce light.

That’s it for the beginning of the film/day two. Turn the page for the end of the film, shot on day one…

Filed under: Cameras •Lighting •Production •Tips •

Anatomy of a Spot: Commonwealth Club

Filed under: Cameras •Lighting •Production •Tips •Training •Visual Effects •

World’s Only “Death Oompah” Band Gets Virtual Reality Music Video

The Tiger Lillies don’t pay money for their promotional work. Instead, they pay artistically. They give free reign to whoever offers to do the work. This has led to the occasional surprise. (In this case it will be a very happy one.)

Mark sent me the following Quicktime videos as a reference to the project’s look:

The bottom clip shows Mark standing in for the ship’s captain, who, in the “Rime” and the song, plunges his crew into a living hell while trying to kill an albatross.

These clips gave me a good idea of the look Mark was going for, and I noted that the light source was soft, had some contrast, and came from the right side of frame. Knowing that there was only so much I could do with our small crew I asked Mark if I could use the same rough look for the beginning of the piece, which takes place on a gazebo at dusk. He said yes, so I formulated a cunning lighting plan:

The platform doubles as our ship’s deck and gazebo floor, and the only real surface that appears in the video. Against the far wall is a 12’x12’ Ultrabounce, lit with both 4k and 6k HMI PARs. The diffusion material in front of it is Half Soft Frost, which softens the bounced light further and makes the Ultrabounce appear larger and very evenly lit. (PARs almost never create an even spread of light across a bounce surface, so the additional layer of diffusion smooths out the hot spots. Half Soft Frost is very thin, almost like shower curtain, and absorbs almost no light.) We put the Ultrabounce fairly far away so that its light would drop off less across the platform.

To the right of the Ultrabounce is our “fill from the key side,” a 12’x12’ bounce consisting of a layer of bleached muslin on top of a 12’x12’ griffolyn. Griffloyn is shiny and can create specular “hits,” so covering it with muslin gives it a matte finish. Muslin alone is transparent and needs a backing to make sure all the light bounces forward instead of some escaping out the back.

I love filling from the key side. Filling opposite the key creates two tones: a bright “key” side and a darker “fill” side, while filling from the same side as the key extends the key and wraps it around the subject, creating a much wider range of tones. In essence it creates the impression of a larger key source. It also “hides” in the shadow of the key and frequently doesn’t reveal itself as a separate source. If it does, I feel it to be much more pleasing to have two soft shadows falling in the same direction, as might occur in a room with two windows next to each other on a wall, rather than one shadow falling one direction and another falling in the complete opposite direction, which feels very “lit” to me.

On the left is an 8’x8’ frame of muslin, providing a base fill opposite the key to boost the darkest shadows to a minimum level. A12’x20’ green screen is rigged behind the stage. This is the biggest screen that we could reasonably use in the space, and we shot off the edges a lot. Mark said that he wasn’t terribly worried about shooting off the screen as the platform’s straight edges would be fairly easy to track.

We used Build 20 on a RED ONE, set at EI 320 but rated at EI 160. (I tried setting the camera to EI 160, having habitually rated the RED at Ei 160 but never actually encoding that into the metadata, but the resulting image appeared too dark on the director’s monitor.) I habitually rate the RED at EI 160 to crush shadow noise, and while others rate it at 200 or 250 I feel more comfortable giving it a full stop of additional exposure light. I’ve had very good results so far, but it makes interiors awkward as producers are not used to budgeting for that amount of light. In this case, with a total of 10,000w of HMI light plowing into the Ultrabounce, I got a whopping T3.5 on my meter. As I wanted the flesh tones to pop a little we shot at T2.8 for most setups. (The key side fill was T2.4, and the opposite side fill read T1.2.)

Except for one shot, everything was lit for 5600k daylight. I set the camera shutter to 1/60th, not so much to avoid flicker (all lights were on flicker-free ballasts) but to slightly reduce motion blur and make keying and rotoscoping easier.

Here’s the opening shot, of the band playing on a gazebo at a wedding:

And here’s a moving test composite. There’s a lot of cleanup that has to be done here, but that’s what happens on low-budget VFX shoots: we shoot first and worry about details later. Normally that just adds to the post budget, but given that the director is going to do a significant amount of post work himself, and is having quite a bit of success enlisting post houses to donate their time toward building his reel, it made sense to work quickly and get everything in the digital “can” than to spend long hours making everything perfect and taking excessive advantage of the crew’s goodwill.

We returned a week later for a day spent shooting extras for insertion into other parts of the music video. For one of our setups we shot a man and a woman, a bride and groom, for insertion in front of band footage shot the previous week. (The Tiger Lillies are playing for this couple’s wedding, and the song reduces the bride to uncontrollable sobbing for the last shot of the video.)

I tend to be very analytical in my approach to visual effects, which is a bit of a contrast to Mark’s approach which is very intuitive. I tend to want to precisely match camera movements, focal lengths, camera height and angles, whereas Mark is very comfortable doing all that by eye. The following clip shows how successfully we matched two dolly moves, shot a week apart, just by eye. This is a VERY rough comp:

This apparent distance between the couple and the band makes it less obvious that the foreground dolly shot ends before the background plate dolly shot begins. When the couple dances by they are moving left to right only in front of a still camera—and we shot them in slow motion at 48fps and 3K resolution. They feel as if they are floating past the camera. It all works wonderfully.

The surreal feel of the video is established within seconds, so a little discontinuity actually works to the video’s advantage.

I lit the couple the same way I lit the background plate, although I added two Kino Flos on the floor for a little uplight on their faces. After this shot we reset for the pull out, where the bride sobs as her new husband tries to console her. While the previous shot took place at dusk, this one takes place at night, after the song has concluded. I turned out my big soft sources and lit the bride and groom very simply, with no fill:

I’ve been working in HD so long I’m not sure which part of the meter to read. The camera white balance is set to 7300K to give the light from the daylight-balanced Kino Flos a warm tungsten feel. [AW]

I don’t have an example of how that looks at the moment, but it felt as if a ghostly light was coming from the gazebo. The ghostly part was due to the lights on the ground, which provided the same lovely uplight as in the previous setup but was a little out of place here. It looked great but it wasn’t technically correct, and that added to the surreal feel of the piece.

We shot the couples dancing in the background behind the gazebo two at a time right up against our green screen, and Mark dropped them into the background behind the band as needed.

Here’s the background plate for the end of the video. Try to imagine a reverse dolly move passing between the bride and groom as she cries helplessly and the guests look on, horrified. Ah, good times!

I think it’s pretty obvious that the band has a very rich, dark and funny sense of humor.

As you can see, the lighting for this setup was very simple: we hung a Kino Flo Vista Beam 600 from a Speed-Rail goal post and diffused it heavily with Lee 129. That, combined with the tungsten practicals wrapped around the gazebo and some Super Green Kino Flos on the screen, was all we needed. Although the light is motivated by the practicals, pushing light into the scene from roughly the same direction is often enough to satisfy the viewer’s brain that the light source is real. This doesn’t always work perfectly, but it works often enough to be a valuable trick of the trade.

You may have noticed a green flag on the left side of the frame. Due to the small size of our green screen (12’x20’) and the 24mm Ultra Prime that felt just right for this shot, the gazebo posts fell off the sides of the green screen at the start of the shot. Mark reassured me that he had no problem tracking straight edges with a matte, but that small black lantern passed in front of black visquene that covered the door into the next studio, and I knew it would he hell to rotoscope that lamp as it would simply disappear. I had our key grip wrap a flag in green fabric and place it behind the black lantern in order to separate it from the black background. It probably won’t make for a perfect key but it’s a start, and it’ll make rotoscoping the lamp a lot easier.

I used roughly the same lighting setup for a sequence in which the band sings in the hold of a sailing vessel:

[AW]

The Vista Beam was suspended by rope so it could be swung side to side, simulating the light from a gas lantern that will be added digitally along with the ship’s hull. The dividers on the light create an “egg crate” or “grid” effect that reduces the size of the light source from left to right, splitting it into three smaller sources and emphasizing the feel of a swinging light over the lead singer without reducing the amount of light falling on the rest of the band to the rear. As I’m deathly afraid of the RED’s noise floor I decided to pop the other band members with two additional Kinos, and then bounced some “ambient” fill into the set using a 1200w PAR aimed into a 4’x8’ bounce card below the lens.

Bounce light from below feels very much like “ambient” or natural light, and I love the subtle beauty of a big soft fill source near the lens. The farther along I get in my career the more I feel that placement of the fill light is vastly more important than the placement of key lights. An otherwise ugly or “overly realistic” lighting setup can be rendered beautiful by a fill light of suitable quality placed in just the right spot.

I’m thrilled any time I can create a sense of space with lighting, either by moving a light during a shot or lighting spaces that people or things can move through. Part of the illusion of depth in two dimensions is created by the reaction of a person or object to a light source: changes in brightness, changes in shadow length and edge softness, shadow texture, motion, etc. At a coarse level you can think of lighting in layers: depth can be created by playing a bright foreground against a dark background, or a dark foreground against a bright background, or any combination of alternating layers. In the case of the shot above, the swinging light causes shadows to play differently upon the foreground and background, creating a sense of depth by lighting the two layers in slightly different ways over time. There’s a strong sense of where the light is in space in relation to the people in the shot, even though the image is viewed on a flat plane.

Here’s the shot in motion (but not sync’d yet—this is from a VERY rough cut):

I love that closeup.

I think that was a 24mm Ultra Prime. (I asked Adam Wilt to keep a log of camera details for post—lens, subject distance, focus, camera height, camera angle, etc.—although after seeing some of the rough comps I suspect it won’t be needed.)

The stills crew shoots for the cover of the band’s CD. [TB]

Let’s go up on deck and see what’s happening on page 2…

Filed under: Production •Tips •Training •

So You Want to Work with Cameras

Filed under: CS4 •Tips •Training •Web Video •

ADHD Guide to Flash Video for the Web

The next step is importing our video:

Which takes us to

In this case the file is on my computer, so I’ll select the “On your computer” option and click “Browse” to select my file. Having done that, I’ll click “Continue” on the bottom right. This screen appears:

This is where you can select the player “skin” that you’d like to appear with your video. For ease I’ve selected “SkinUnderAll,” about halfway down the list, as this will show all available player controls underneath the video window.

After selecting a skin, clicking “Continue” brings me to this page:

This is a quick summary of what we’ve done so far. Note that the video path path is relative to the .swf, or final Flash movie. We’ll fix that later. Click “Finish” to go back to the stage:

We still have some work to do. Click on the black box representing the video window and look at the inspector:

Right now we won’t worry about position and size, as the video-player combination is our only asset. The important thing to do here is to click the chain icon to the left of the video dimensions. This locks the video aspect ratio and prevents it from being deformed in the event a user resizes it on playback.

Now click on the stage itself:

I’m only concerned about what appears in the “Properties” section. This describes how big my Flash “stage” will be. The default was smaller than my video size, so I increased it—adding pixels to the vertical dimension to make room for the player, similar to what has to be done when embedding Quicktime into a web page. (The video is 720x404, but the stage has to be 720x441 so it doesn’t cut off the player.)

The “FPS” setting determines the playback frame rate of the entire Flash project, so this will have an impact on the Flash project’s video assets. My video file was cut on a 60 fps timeline and is a 60 fps video file, but it will be played back at 24 fps if I leave this setting alone. (This setting has very little visual impact on the final Flash movie. The end viewer notices nothing unusual.)

Below “Size” you’ll see an option for “Stage”, along with a color picker box. This allows you to color your stage to match the web page in which it will be embedded. You may have noticed that the corners on the bottom of the player “skin” are rounded: the color of the stage will show through between the edge of the stage and the edge of the player.

next page: testing, exporting, and the thrilling conclusion

Filed under: Cameras •Lighting •Production •Tips •Training •

Rough Guide to Flicker-Free HD Shooting

Filed under: Cameras •Tips •Training •

What You Need to Know about Green, Blue and RED

Here’s a still frame from a RED running Build 20. It’s set to Rec 709, and the image was white balanced in RedAlert. It was shot under tungsten light with an 80A filter, which converts tungsten to daylight. The camera was white balanced for 5600k daylight, and was seeing a chart lit by 5600k light:

Here’s the same chart shot under 3200k light, without the 80A filter and using the “tungsten” preset:

Under tungsten light (bottom chart) the greens and yellows appear muted and dull compared to daylight (top chart). The cause for this is easy to see in the following two waveforms.

First, the RED under daylight:

The circled area shows the left column of chips ranging from green to yellow. Yellow is a combination of equal amounts of green and red, so what we see in that left column is a set of colors in which the green value is the same all the way through: the amount of green in the bottom green chip is the same in every other chip up through yellow, just with varying amounts of red added.

The dip in the circled area shows what we’d expect from the blue channel in a part of the chart where there is no blue: it dips, because there’s nothing for it to do. Notice that on the green and red channels that part of the waveform spikes, showing that those channels see their respective colors in that part of the chart.

Now look at the RED under tungsten light:

The circled area is flat compared to the daylight waveform, which indicates that the blue channel sees some blue in that part of the chart under tungsten light. If you scroll up and look again at the chart images you’ll see that the green and yellow column of the tungsten chart has a blue cast to it when compared to the daylight chart.

The trick is to figure out whether this is intentional or not. If the white balance setting is the only way the camera knows what overall color the light is, then I should be able to incorrectly white balance the camera and force it to mix green into blue when it doesn’t need to.

So I shot a chart with the RED white balanced for daylight, but lit with tungsten light:

Then I shot another chart, lit with tungsten light and with the RED white balanced for tungsten, but with an 80A filter to convert the tungsten light to a daylight white balance:

I then took these images into RedAlert and white balanced them back to normal. In both cases I had to increase the effective ASA to 640 in order to boost the least exposed channel high enough to match the best exposed channel.

If the RED was intentionally mixing color channels to eliminate blue noise under tungsten light, then I should be able to white balance the RED for tungsten light, shoot the wrong color of light (daylight), and see the RED mix green into the blue channel to eliminate the blue noise that it thinks is present because the camera is set for a tungsten white balance. Here’s what happened:

If RED was intentionally mixing color channels based on the white balance setting then each set of waveforms should match, because the white balance settings were the same even though I shot under two completely different kinds of light. The waveforms don’t match, which tells me that the channel mixing isn’t intentional at all, but is instead a flaw in the sensor’s color filters. The camera is simply reacting to the light that’s in front of it, regardless of how the white balance is set.

Turn the page to see my explanation of what I think is going on…

Filed under: Cameras •Production •Training •

RED Color Comparison

Filed under: Cameras •Production •Tips •

Tiffen Goes Into Production on T1 Far Red Filter

Here’s my homemade far red/infrared fabric chart:

Read this article to learn how I developed it. (This test chart was shot with a Sony EX3 at Videofax.)

As you can see, the left two columns of fabrics reflect a lot of far red, while the third column doesn’t. The Macbeth chart is present to show whether there are any obvious color shifts when using the T1 filter.

Here’s the same chart with the T1 filter added:

No more far red.

How big a deal is far red on the EX1 and EX3 cameras? The answer is “really big”: I shot this test chart under tungsten light, period. No special camera settings, no neutral density filters. This is just the way it is. The F35 shows a lot less far red contamination but it’s still there—and the T1 filter cleans it up nicely.

My market has no F35’s currently available for testing, but as the original T1 worked fine on the F35 and this filter appears to be a perfect color match (I compared it to the original T1 that I still have in my possession) I have no qualms recommending it for the Sony F35 and the Panavision Genesis.

The black patch at the top of column two looks like it might still have some red in it, so I zoomed into the image in Final Cut Pro 7 and took a look. The top image is without the T1 filter, and the bottom image is with the T1 filter:

It looks like quite a difference, but eyes can be fooled. Let’s look at the corresponding vectorscope images, with some magnification applied:

The top vectorscope shows a definite skew towards red without the T1 filter. The bottom vectorscope shows that the T1 filter has restored black to perfect black.

The T1 filter is green, which requires white balancing when the filter is first attached, and there appears to be no issue with it skewing colors unfavorably. The left column shows a DSC chart shot with no filtration, and the right column shows a DSC chart shot through the T1 filter:

If you can see a difference then I’ll be forced to ask how strong a microscope you’re using. It’s a near perfect 99.99% match.

When I first wrote about the filter I referred to it as “T1” as that’s what it was labeled. I found out later that stood for “Test 1,” but the name stuck. Pre-order it from your local Tiffen dealer.

Production will begin October 1st, and it may take four weeks before all sizes are available. Standard sizes will include 72mm, 77mm, 82mm, 138mm, 4x4, 4x5.6 and 6.6x6x6. Other sizes will be available as a special order.

Tiffen let me keep the prototypes I’ve tested, and I’ve lent them out to friends to use on their EX1 and EX3 shoots. They’ve become quite desperate on the few occasions when I can’t get them a filter in time. I suspect Tiffen will be selling quite a lot of these. If you own or use Sony F35, EX1 or EX3 cameras, this filter is not an option: it’s a requirement.

Art Adams is a DP who is well red. His web site is at http://www.artadams.net.

Filed under: Cameras •Production •Tips •Training •

The Vertical Bar Graph Raw Exposure Meter

Filed under: Cameras •Post Production •Production •Tips •Training •

RED’s Blue Noise: Where It Went, and Other Color Anomalies You Should Know About

Here’s a 5600k illuminated chart from this test, shot with Build 20:

And here’s the same chart shot with the same camera, same software build, under 3000k light:

The charts look a little dim as I exposed using a meter set at RED’s recommended EI of 320. I did no color grading other than to white balance the images using the white picker in RedCine before exporting them as TIFF files. (Don’t look at the resolution trumpets as they are nowhere near accurate. I’ve zoomed in to the chart from a larger test frame.)

The thing I noticed immediately is that the greens don’t pop as much under tungsten light as they do under daylight. The green chip, as well as chips containing green such as cyan (blue plus green) and yellow/orange (red mixed with varying amounts of green), are desaturated and appear cooler than the same chips on the daylight chart.

Eyes are easily fooled, however, so let’s look at some parade waveforms. First, the daylight waveform:

And then the tungsten waveform:

See it yet? If not, here’s what I noticed after a bit of looking:

First, Build 20 under daylight:

Then, Build 20 under tungsten light. Here’s the difference:

When a color channel doesn’t detect its color in an image, it dips. That’s why there’s a downward notch in the blue channel under daylight: the blue channel doesn’t see any blue in the green chips, and it shouldn’t: there’s no blue there. Under tungsten light, however, the notch disappears. The tungsten waveform should show the same downward notch that the daylight waveform shows, and it doesn’t.

This means that the blue channel is seeing blue in color chips that don’t contain any blue.

I can’t positively determine what’s going on here, but I was able to replicate it—and that may give us a clue as to what’s happening here. Turn the page for an introduction to Final Cut Pro’s Channel Mixer…

Thursday, August 27, 2009

Filed under: Cameras •Production •

CameraMania!

Art Adams | 08/27

The Real Scoop on What the RED ONE and Canon 5D Are Really Good For

THE RED ONE

Recently I was contacted by a production company in New York City who were looking for a local DP. “We want to shoot 120 fps on the RED in front of green screen,” they said.

As I’m a relatively honest fellow, I felt I had to raise a technical point. “120 fps on the RED is fine. Green screen on the RED, no problem. BUT—you don’t want to mix them.” I explained why, and I told them that if they’d tell me about their workflow I’d find another method to get what they were looking for at or below their current budget.

“Excellent! We’ll get back to you soon,” they said, and then they hung up and dialed a line producer of my acquaintance. “Give us a list of DP’s in your area who know the RED.”

Jim Jannard is a master of marketing. He made billions of dollars selling sunglasses, and he’s a genius at making people want his products. Unfortunately he did too good a job with the RED, because I occasionally find myself in the position of telling a production company that the RED won’t meet their needs… and they believe the marketing rather than the DP. While the RED is certainly the best $17,500 camera ever, it is not the best all-around camera in existence.

WHAT IT’S GOOD FOR

The camera has very high resolution. It utilizes a 4k sensor, although as it uses a Bayer filter sensor the real resolution is only between 2.8k and 3.2k, depending on the quality of the lens used, due to loss of detail during the demosaicing process. That’s more than enough resolution for film output, and that detail has convinced many that simply owning this camera will allow them to open their own studio.

To some extent they may be right. RED has certainly brought film resolution HD to the masses. Whether the masses have the lenses and support equipment to actually use their RED cameras is another matter entirely. The important thing is that the promise is there.

The RED revolutionized another aspect of HD production: it was designed from the ground up to function “film style.” The captured images are intended to be color corrected later, adding a world of possibilities to the final look. I know two directors who love working this way as there is no engineering station to cart around during production and no waiting to tweak each shot with a paintbox. We just shoot. We work faster, actors remain fresher, and we can finalize the look of the end product in a proper viewing environment instead of under a tent of flags in the middle of a park in broad daylight.

When one considers that the RED has brought to the masses a film-like workflow and style of shooting, it truly is a revolution unto itself.

WHAT IT’S BAD FOR

The bottom line is that it’s still a $17,500 camera, and you don’t make a 4k film resolution camera at that price point without cutting a few corners.

The sensor is slow. RED says to rate it at ISO 320, but a lot of DP’s find that it works better at 160-200. It’s very hard to calculate an ISO rating for an electronic imaging system as ISO is based upon the toe of the film exposure curve, and HD doesn’t have a toe: HD gamma is a straight line down into the murk of noise. Still, setting a light meter between 160-250 seems to work for most people, depending on taste and comfort with noise.

The sensor only offers about nine stops of usable latitude before noise overwhelms the shadows. The RED often looks like it has more latitude because it’s easy to protect for the highlights and open up the mid-tones in post as long as you don’t open them up too far. The great benefit of the RED’s workflow is that those nine stops can be pushed around later in post, while ordinary cameras “bake in” the look to such an extent that you are largely locked into whatever look you create in the field.

The RED is not very sensitive to blue. It’s native white balance is 5000k, and under blueish 5600k light it captures a full nine stops of dynamic range. Under tungsten light, though, the blue channel is starved for exposure, as there just isn’t that much blue in tungsten light. Most HD cameras hide this weakness by boosting the blue channel gain, which is why the blue channel always looks noisy when viewed on a high quality studio monitor, although most cameras seem to be more sensitive to blue than the RED. Equally annoying is that the tungsten channel is overexposed under tungsten light: objects containing red will clip sooner under tungsten light, limiting the camera’s dynamic range to about 7.5 stops.

Adding blue to the lights or in front of the lens will ameliorate this problem somewhat, reducing blue channel noise and increasing red channel latitude, but at the expense of reduced overall exposure sensitivity.

The RED’s entry-level software is bug-ridden and inconsistent, at least at the moment. For example, the new RED ONE camera software build, no. 20, appears to largely eliminate blue channel noise under tungsten light as well as improve Rec 709 color rendition, but only if used with the current version of RedAlert. RedAlert is design to process one clip at a time, although the underlying engine can be used to process batched of clips using an included command-line tool called RedLine. RedLine, though, doesn’t allow for color correction.

RED’s free batch color correction tool, RedCine, has not yet been updated to work with the new build 20 colorimetry as it is maintained by a completely separate group from those who implemented the new colorimetry in the camera software and RedAlert. A new version of RedCine is due to be released soon, but meanwhile… you can get a taste of the new look, but you can’t use it very easily. If you can’t afford $100k+ for Scratch or SpeedGrade then your options are occasionally limited.

At high speeds the RED utilizes a smaller portion of the sensor in order to pull the image data off fast enough. As the effective size of the sensor decreases, though, resolution drops dramatically. The true resolution of 2k RED footage is probably closer to 1.5k, so at 1.9k (1920x1080) it’s not going to look terribly sharp. In a lot of situations that may be okay, and some post sharpening may solve the problem acceptably. It other cases, such as shooting in front of green screen, that softness may be a deal killer as it will make matte edges look a bit unreal.

The RED may be difficult to color correct if you desire bright, saturated colors. Anecdotal evidence suggests that RED footage will take a bit longer to tweak in the color correction. I can speak from personal experience when I say that some colors, such as purples and oranges, may not translate well without a lot of secondary work.

None of the viewing outputs are truly accurate. The monitoring output is 720p only, which makes critical viewing of a 4k image awkward at best. The image always looks better on a computer, though, so when in doubt it’s best to pull the hard drive off the camera and open a clip up in RedCine or RedAlert. A lot of picture oddities, such as weird color shifts and moire patterns, disappear when viewing the image on a computer display. Not being able to monitor accurately is a mixed blessing, as once you realize you can’t truly see what you’re doing, but that the image you’re capturing always looks better later, it’s easier to step away from the monitor and just shoot.

The RED is not a documentary camera. Its handheld ergonomics are appalling, and it’s very difficult to judge focus through the viewfinder. What might look sharp at the time may not look okay on a 40’ wide screen. It’s not a good camera for following an unpredictable subject handheld and under low light.

SUMMARY

The RED functions best in a traditional film-style shoot environment, with a camera assistant following focus, a fair bit of light, and a data-wrangling station nearby. This isn’t a bad thing at all: the fact that we can shoot HD film-style inexpensively is a blessing that the larger, more established camera companies weren’t going to give us any time soon. But the RED is still a $17,500 camera—so don’t expect it to do everything. What it does do, it does very well indeed—but it is not the last camera you will ever need or use.

And now, the Canon 5D…

Saturday, August 15, 2009

Filed under: Cameras •Tips •Training •

RED Build 20 Torture Tests

Art Adams | 08/15

Exposure latitude tests reveal perhaps too much about the RED ONE

Daylight Overexposure

Here I am at our baseline exposure reading of EI 320. This is what RED says the camera should be rated at, so we took them at their word with the goal of determining whether we thought they were right or not.

Remember, I zoomed into this chart in Final Cut Pro—so while it shows accurate color, please ignore the resolution trumpets as they are nowhere close to accurate.

Build 20’s Rec 709 pattern looks pretty good. It used to be stretched out on the green/magenta axis, resulting in greens and magentas being exaggerated. If there are two colors you don’t want to emphasize in flesh tone and clothing, it’s those two. (Green is worse.)

The colors on the red-magenta-blue side of the scope look more saturated than the yellow-green-cyan side. Orange, between red and yellow, looks the least saturated. Tertiary colors, between primaries and secondaries, are consistently less saturated.

The waveform shows a shift toward blue, as does the vectorscope above. This is most likely an error in RedCine’s “white picker” white balance tool, which doesn’t seem to be terribly accurate. At least any errors it created are consistent as I took the white balance it gave me for the baseline exposure and applied that number manually to all the other exposures.

My goal was to find where the white chips started to clip, and to see what channel clipped first. Since nothing significantly changed between this exposure and 1.5 stops overexposed—other than a consistent increase in saturation with each increase in exposure—we’re going to skip ahead.

Everything looks fairly normal here. The red, green and blue steps on the waveform are nice and crisp, showing an absence of noise and a nice sharp signal.

At this point the white chip on the DSC chart is 3.5 stops brighter than the 18% gray card.

This is where things start to become interesting. The RedSpace gamma curve is starting to compress the blue channel. Notice how the top of the “ladder” on the blue portion of the waveform is curved compared to the red and green channels. So far the underlying “raw” data is fine, but the RedSpace gamma curve knows that the blue channel is going to clip shortly and is trying to flatten it out to squeak out some extra detail.

At this point the white chips on the chart are 4 stops brighter than 18% gray.

This is pretty impressive. The blue channel is now severely compressed but the remnants of the stair-step pattern in the top of the blue signal show that there is still some detail there. The RedSpace gamma curve is working overtime, but it seems to be doing a good job. The green channel is starting to be compressed as well.

The vectorscope shows that cyan is beginning to fail, as the points on that side of the scope are collapsing inward, showing loss of saturation.

The white chip is now 4.5 stops brighter than 18% gray.

This is the end of the line for daylight overexposure. The flat lines on the waveform show that both blue and green have clipped, and red is heavily compressed and just short of clipping. This makes complete sense, as the amount of red in daylight is significantly less than either of the other two colors—the exact opposite of tungsten light, which has a lot of red and very little blue.

What’s really strange is that notch on the blue channel that extends above the point where the blue channel appears to be clipped. That represents the right side of the color chart where blue, purple and magenta sit. The blue-saturated colors seem not to be clipping, while the neutral color chips that contain all three colors appear clipped, but I’m not sure how they can be flat while the blue colors aren’t. Very weird. Perhaps someone smarter than I can explain that one.

The white chip is now 5 stops brighter than 18% gray.

Based on these tests, I see that RED build 20’s daylight overexposure latitude safely handles up to 3.5 stops of overexposure beyond 18% gray, and even extends to 4.5 stops of latitude without significant penalty. (This, of course, is based on exposure of a white object; overexposure of brightly colored objects might result in that channel clipping prematurely.)

Thursday, August 06, 2009

Filed under: Cameras •Post Production •Production •Tips •

Busy Person’s Guide to Color Correcting the Canon 5D

Art Adams | 08/06

In which I take some pretty footage and make it a lot prettier

We had no idea, when we shot this, that it would be the opening shot of the spot. I believe we shot this second to last, just before shooting Cary’s final line. (Cary was the son of the director/editor, Simon Sommerfeld.)

This is a Leica 35mm still lens, with a Canon mount adapter, probably around F8 to hold all the balloons in focus.

This is the original shot as it popped up in Final Cut Pro. The trees look like they’re being hit by a reflector but I’m pretty sure that’s just ambient bounce light off of sand. You have no idea how hard it is to make balloons follow instructions. Worse than kids filled with helium, although the kids’ voices under helium are a lot funnier.

We used a 35mm Leica still lens on this shot. It looks like I stopped down a bit to hold focus on all the balloons. I’m probably at F8, and so is the lens.

Lift/Gamma/Gain is the name of a module in Magic Bullet Looks that allows the user to tweak lift (pedestal, or black level), gamma (mid-tones) and gain (highlights, or overall signal strength). These three controls will achieve a significant amount of what needs to be done to the overall image. They all interact, so adjusting one requires readjusting the others. For example, boosting gain brings up both the mid-tones and blacks, although not as much as it does the highlights.

Adjustments to the overall image are “primary” color correction. Adjustments to a portion of the image or a range of color are “secondary” color correction. Nearly everything I do here is primary color correction. (Tertiary correction is when the director calls and asks me to change something.)

In this case I probably left the lift and gain alone and opened up the gamma slightly, brightening the mid-tones.

I’m a sucker for a good vignette. The Vignette module is like a circular grad, clear in the middle but darker at the edges. It’s a great tool for subtly directing the eye toward a specific place in the image by darkening everything around it. In this case it looks like I just darkened the corners a little.

Another tool for directing attention is the Edge Softness module. I like the Lensbaby selective-focus look, especially when the corners of the frame are filled with gack like bright tree branches. Sometimes the best way to make an image interesting is to mess it up a bit. (Or at least that’s my excuse while shooting.)

I boosted the saturation a bit here, although it’s hard to tell as I don’t think this JPEG can handle the increased color.

The last step in every shot in this piece is the addition of a Tiffen DFX 2 Bronze Glimmerglass filter. We wanted the look to be a touch soft and a little warm, and as most of the shots were either fairly neutral in white balance or were made neutral later, the Bronze Glimmerglass filter allowed me to add a precise and consistent amount of color to every shot.

We found that the full effect of the digital filter was too much, so I turned off the halation portion of the effect. Try doing that with a glass filter! (Actually I did, and I’m still paying it off. Quick tip: don’t use steel wool.)

——————————————————————————————————-

This is a Canon L-series 85mm autofocus lens set to manual focus. I originally set this shot up at F2.8 but only the hinge on the front swing was in focus. I decided to stop down to F8 to hold focus on the first swing, knowing that I could later blur the objects on the right side of frame.

Here’s the raw H.264 file. It’s a bit dark as I’m trying to hold color in the sky and not blow it out. It looks a little muddy but all the shadow information is there, ready to be tortured from the image.

I boosted gain a lot to make the bright side of the front swing pop. It’s got a nice hot edge on it that I really like. I kept the gamma down, in order to keep the left side of the front swing a little dark, and maintained the level of the blacks as they look pretty good.

If this shot was a prisoner of war it would have given up everything as soon as it saw me cleaning my teeth with a toothpick. It was that easy to open up the mid-tones.

This spot was all about color, so I used the Color Grad module in Looks to bring some blue back to the sky. Actually, as I’ve discovered from playing around with skies in Photoshop, they are actually more cyan than blue.

The Vignette module takes the corners down a little in brightness while keeping the front swing roughly the same. By now the swing is developing quite an ego.

The Edge Softness module increases the feel of an out-of-focus background, further directing attention to the now swaggering swing.

And a little yellow-orange Bronze Glimmerglass warms it all up like hot cocoa at the Russian Front.

Turn the page for more…

Wednesday, August 05, 2009

Filed under: Cameras •Tips •Training •

Abel Cine offers New Mattebox System for Canon 5D Mark II

Art Adams | 08/05

Not just for cops anymore, custom donuts are mandatory for the Canon 5D

Tuesday, August 04, 2009

Filed under: Cameras •Production •Tips •

My First Shoot with the Canon 5D

Art Adams | 08/04

For a still camera it shoots pretty nice HD

My first experience with the 5D involved shooting a spec spot for director Simon Sommerfeld. Simon normally works as a DP, but when a client of his asked him to direct a short piece to help sell a TV ad campaign to the client’s board of directors, Simon jumped at the chance to use his 5D to tell the story. And it was my great good luck that he asked me to shoot it for him.

Initially we puzzled over the camera, trying to discern what settings belonged where in movie mode. We knew that there was a way to drill deep into the engineering menus and create a look but we didn’t know how to do it. Simon poured through user groups looking for tips, and he found some settings that seemed to yield a nice flat image suitable for capturing the widest range of contrast possible for post color correction. These same settings also kept detail enhancement low, which is always a benefit in HD: an overly-aggressive detail circuit can be very unflattering to flesh tones and can result in a “newsy” look. Detail can always be added later, if needed, but it can’t be removed.

(I’m trying to get more details out of Canon as to how one creates custom movie mode settings. Anecdotal evidence points to using Canon desktop software to create a look using a still image shot on the camera, and then saving the resulting look instructions to one of the camera’s built-in look memory slots.)

Judging exposure was both simple and complex. In movie mode the 5D offers no telemetry regarding exposure: no histogram, no zebras, nothing. (There is a hack that creates zebras in movie mode but it is not blessed by Canon.) It’s possible to see a histogram during playback, but without the benefit of clairvoyance that’s not a very helpful feature. What I discovered, though, is that the 5D’s LCD screen is actually quite accurate: it’s not perfect, but it is possible to judge exposure and contrast accurately. The best way to avoid clipping is to change the F-stop and watch for highlights that don’t change value, and then close down the exposure to taste. This procedure works fairly well but it does require post color correction.

Let me emphasize this: you should definitely plan on color correcting 5D footage. More on that later.

Focus is another matter. On the LCD screen, in movie mode, is a small box that can be moved around the frame using a joystick controller on the back of the camera. Placing it on an area of interest (usually a person’s eye) and hitting the magnify button yields what appears to be a pixel-for-pixel view of that part of the image. This is the only reliable way to set critical focus. It’s fairly simple to see when an object comes into focus, but not so simple to see whether focus tracks a moving object.

A fellow cinematographer gave me some advice: never shoot wider open than a F4, and shoot at F5.6 whenever possible—just to make sure focus holds, as I won’t know for sure it held until post. For the most part I followed that advice, although I occasionally snuck the aperture open to F2.8 for specific shots.

Turn the page for behind-the-scenes action…

Thursday, July 23, 2009

Filed under: Training •

NorCal Digital Cinema Society Meeting Tonight, 6pm

Art Adams | 07/23

For those who didn’t get the message via the Cinematography Mailing List or Facebook, the Northern California chapter of the Dig

Tuesday, June 30, 2009

Filed under: Cameras •Production •

Facebook Mobile spot, Shot on RED

Art Adams | 06/30

Just add talent, a stolen location, natural light and a RED ONE camera

Tuesday, June 23, 2009

Filed under: Cameras •Production •Tips •Training •

Black Glimmerglass and Hollywood Black Magic Filter Tests

Art Adams | 06/23

A camera assistant extraordinaire puts two new lines of diffusion filters through their paces

Tuesday, June 16, 2009

Filed under: Cameras •Production •Tips •Training •

Diffusion Confusion

Art Adams | 06/16

What you need to know about glass filters in the digital age

THE DIFFUSION TRIANGLE

Zupka relates that he was taught to think about diffusion filters as falling somewhere within a triangle whose points were (1) reduction of high frequency detail, or resolution, (2) contrast, and (3) halation. “These three things are what every diffusion filter affects one or more of,” he says. “A Fog or Low-Con affects mostly contrast and also some resolution. Classic Softs or Soft F/X primarily affect resolution, with little affect on contrast and no halation. Pro-Mists or Frosts affect contrast and resolution, with some halation depending on the color of the diffusion (white, black, gold, etc.).”

Schneider has three basic types of diffusion filters, which Zupka describes:

“The Classic Soft is a very, very clean diffusion. It only affects very high frequencies in detail, with no effect on contrast level. It just affects very fine details, like wrinkles. The filter contains internal lenses that refocus the image off the image plane. The open spaces between the lenses pass a sharp image, while the lenses overlay a defocused image on top of the sharp image. Different strengths merely increase the number of lenses.

“Black Frosts contain black particles that diffract light around the particle, as well as scatter light a bit. Our particles are porous, like volcanic rock. Being able to see through the particle keeps blacks and dark tones from getting mushy, while diffraction and scattering contribute to the diffusion effect. This is different from Tiffen’s Black Pro-Mist, whose black particles are crystals with facets that scatter more light and reduce contrast.

“Our Digicon filters are made of very, very tiny black particles whose effect is to reduce highlights while increasing black levels. The midtones are not affected. This can add up to two stops of latitude to a typical HD camera, compressing highlights and lifting shadows to protect detail at exposure extremes. An example of how this filter works entails watching a waveform monitor as the filter is inserted in the matte box: the highlights come down 10 ire and the shadows come up 10 ire. The color rendition is completely neutral.”

On page 3: diffusion in front of the lens or behind the computer?

Wednesday, June 10, 2009

Filed under: Cameras •Pre-Production •Tips •

Using the ND .90 Filter as a Preproduction Tool

Art Adams | 06/10

Spot problem fabrics with an ND filter and a flashlight!

Friday, June 05, 2009

Filed under: Cameras •Production •

Far red on the EX1/EX3/F35: “It’s a feature, not a bug!”

Art Adams | 06/05

An official response from Sony tells us that far red is more about helping than hindering

Thursday, June 04, 2009

Filed under:

Look for me at CineGear

Art Adams | 06/04

Wednesday, June 03, 2009

Filed under: Cameras •Pre-Production •Production •

The Last Far Red Filter You’ll Ever Need?

Art Adams | 06/03

Tiffen’s latest test prototype kills far red dead on both the EX1/EX3 and the F35

Let’s start with what we saw on the EX1. Here’s both a standard black-and-white chip chart hung alongside a MacBeth, mounted against a background of black duvetine and draped with some random fabric that reflects both IR and far red.

In this unfiltered frame you can see some unintended red in the black fabric draped across the bottom right of the frame.

This frame shows what happens when we add the Tiffen prototype far red filter. It’s interesting to notice that the colors containing red on the MacBeth chart are dramatically less saturated in this image than in the first one. I think you can make a case that the unfiltered image shows reds to be a little over the top, but I’d be curious to see which version of the chart you prefer: saturated reds, or desaturated reds.

The other possibility is that the colors that reflect red are also reflecting far red, to which the camera is oversensitive. When I shot this same test with a DSC chart, and overlaid the filtered results over the unfiltered results, there was no color shift at all. See those results here, on page 3 of this article. It could be that the dyes in the DSC chart are more far red-neutral than those in the MacBeth chart.

The following images show something very curious going on. Even though we are adding internal ND, there’s no increase in far red contamination. (The two different notations of ND indicate which filter is being used in two different terminologies.)

When adding ND to any other camera, the far red or IR contamination typically becomes worse because you’re blocking visible light while allowing more far red/IR to pass. In this case, though, that doesn’t happen when using the camera’s internal ND filters: the far red contamination is the same no matter what internal filter is used. This implies that Sony has built some sort of far red filtration into the ND, although why they weren’t able to take care of all of it is a bit of a mystery. It could be that they are emphasizing the camera’s ability to reproduce a wide variety of reds at the expense of ensuring that every black material seen by the camera remains black. I’ve got an email in to Sony asking for more details, and I hope to have something to report soon.

Meanwhile, it appears that this filter—when finally released—may be the only far red filter ever needed for this camera. No expensive sets of IR ND filters for EX1/EX3 owners: one screw-in filter and you’re done. This makes perfect sense, considering that a full set of IR ND’s is beyond the reach of many EX1/EX3 owners who don’t want to spend 30-40% of the cost of their camera again in filters.

There is going to be at least one more prototype of this filter manufactured before the final version is released, so stay tuned.

On to the F35…

Tuesday, June 02, 2009

Filed under: Cameras •Post Production •Production •Visual Effects •

Porsche Spot VFX Breakdown

Art Adams | 06/02

Director Ian McCamey shows how many layers it took to create one shot in our “Porsche: Magnet” spot

Click to play audio / video »

Friday, May 22, 2009

Filed under: Cameras •Production •Tips •

Why Does Blue Hurt So Much?

Art Adams | 05/22

The 80A filter eats up more than twice the light of an 85. What gives?

Tuesday, May 19, 2009

Filed under: Cameras •Production •Tips •Training •

Tiffen tests new no-ND IR filter for EX1/EX3/F35

Art Adams | 05/19

Crisp clean blacks, and only a half-stop light loss, make this new filter a winner

In this article I wrote about Tiffen’s IR ND filters, which use a special dye to absorb far red without using a hot mirror. On the previous page we saw how hot mirrors reflect unwanted wavelengths of light, but those filters are the exception: every other filter works by absorbing light.

Filters work on the same principle that gives us color vision. When full-spectrum light hits a colored object, the part of the spectrum that IS NOT the object’s color is absorbed and disappears. The spectrum that does contain the color of the object is reflected, and that’s the color that our eyes see. In the filter world, light that is the color of the filter is passed and the other colors are absorbed. For example, in the case of an 85B filter, the orange cast of the filter passes mostly warm light (oranges and reds) and absorbs most of the blue, changing the balance of the light from 5600k to 3200k. It doesn’t absorb all the blue—the filter would have to be a much deeper color in order to do that—but it absorbs enough to reduce the amount of blue to a level that tungsten film would expect to see under tungsten light.

What Tiffen did was to find a colored dye that absorbs the far red part of the spectrum, between about 680nm to where the on-sensor hot mirror cuts at 700nm. Since it’s a dye filter the color doesn’t change when viewed off-angle, unlike dichroic filters, so it will never vignette. It can be used with any focal length lens without any issues at all. Until recently, however, Tiffen could only make that dye work in conjunction with neutral density filters. The dye is a greenish color, and the amount of green needed changed depending on the strength of the ND. If you used a heavy ND filter, the green dye had to be heavier to block more far red. If you used a lighter ND filter, less green was needed because less far red was being passed.

The problem is that the Sony F35, EX1 and EX3 could, under certain conditions, see far red WITHOUT any ND filtration at all. And Tiffen didn’t have a far red filter without built-in ND, so if you were shooting under tungsten light you’d have to take at least a one-stop light loss (by using a Tiffen IR ND .30 filter) if far red reared its ugly head.

Until now. Tiffen sent me a prototype filter that only costs a half-stop of light loss and seems to work very well indeed.

Here’s a DSC color chart placed next to my normal far red/IR reference: the DSC chart’s black polyester bag. As you can see, the bag looks a bit maroon on the EX1—but in reality it’s a deep black. The same sensor dye that gives us reds to die for is also making this black fabric look anything but black.

This is the same chart and bag but with the prototype Tiffen dye filter. (It was labeled “T1,” for “test one,” so while I labeled it that in the tests the filter’s actual name will be different when it is released in a month or so.) The black bag is now black again, and I’ve only lost 1/2 stop of light. I did have to white balance through the filter to remove the green cast, but once I did that all the colors seem to be dead on accurate.

Here’s a close-up of the bag with no filtration and with the blacks lifted to between 20 and 30 IRE:

Blech. That’s not a good look. But if we add the Tiffen prototype dye filter, it looks like this:

Yeah, that works. The bag is a nice rich black again.

And now let’s compare the Tiffen prototype to the gold standard, the Schneider Tru-Cut 680:

The Tiffen prototype filter cuts IR at least as well as the Schneider does. That’s very impressive, considering that the Schneider is a dichroic filter and the Tiffen prototype uses a light green dye.

There are a couple of important differences:

Good news: The Schneider Tru-Cut 680 can be stacked with any amount of ND.

Bad news: It vignettes cyan on lenses wider than “normal” perspective (for a 35mm film-sized sensor, about 50mm).

Good news: The Tiffen prototype cuts far red without any vignetting and at only a half-stop light loss.

Bad news: It can’t be combined with any ND, because the amount of green dye in the filter has to increase with the strength of the ND.

The next big question is this: does the Tiffen prototype cut any light that we want to keep? Does it cut the red end of the spectrum too much? Let’s take a look on page 3…

Tuesday, May 12, 2009

Filed under: Cameras •Production •Tips •Training •

Optically White Balance the RED with Schneider CTB Filters

Art Adams | 05/12

Get the red out—and put some blue back in!

Once again, here’s the histogram and gray card under tungsten light and without any filtration:

Red is way ahead of the other channels, meaning that it will clip sooner under tungsten light than in daylight if the camera shoots either a bright neutral tone or a saturated red object.

Let’s take a look at the first, and lightest, CTB filter:

SCHNEIDER 1/8 CTB FILTER

MY METER SAYS: OPEN UP 1/3 STOP

The 1/8 CTB filter is a very light blue, but we can already see the effect it is having on the camera’s white balance: the red peak has moved considerably to the left, showing less exposure in that channel, while green moved a little bit and blue barely moved at all. In theory a blue filter shouldn’t affect the blue channel much because it is passing blue and absorbing the other colors.

By tightening up this grouping we’ve given ourselves a little bit more overexposure latitude, as the red channel will be slightly less likely to clip when shooting bright white or red objects.

The next darker filter is:

SCHNEIDER 1/4 CTB FILTER

MY METER SAYS: OPEN UP 1/2 STOP

The 1/4 CTB filter is absorbing a lot more red. Blue and green are almost in alignment, and the overall grouping is a lot tighter. This filter will allow for even greater overexposure latitude, keeping red from clipping a lot sooner than the other colors and reducing noise in the blue channel due to better blue exposure when the T-stop is opened to compensate for light loss.

Once again, notice that blue has barely moved at all. The changes are most significant in the red and green channels.

The heaviest filter is:

SCHNEIDER 1/2 CTB FILTER

MY METER SAYS: OPEN UP 2/3 STOP

This is an excellent grouping. Blue and green are essentially in balance, and red is only about a half stop ahead of the other channels. While red will still clip a half stop sooner than blue or green, that’s considerably better than clipping one-and-a-half stops sooner—plus the blue channel is getting as much exposure as the green channel, which means it will be a lot less noisy when the exposure is increased to compensate for the reduction in red and green light.

A Wratten 80A filter would bring the peaks into complete alignment but at the cost of another stop of light. I almost never have that much leeway when shooting under tungsten light, and my past experiments with the 80D filter (which falls between the 1/4 and 1/2 CTB filters in strength) show that half correction to daylight is plenty of compensation under most circumstances.

Schneider CTB filters are available from, among other places, Filmtools.

Art Adams is a DP who leads a colorful life. His website is at www.artadams.net.

Thursday, May 07, 2009

Filed under:

Whatever Happened to Ira Tiffen?

Art Adams | 05/07

He’s still working with glass… but not how you think!

Tuesday, May 05, 2009

Filed under: Cameras •Hardware •Production •

Hands on with the Gearnex Gear Head

Art Adams | 05/05

We take the wheels for a spin

Thursday, April 30, 2009

Filed under: Cameras •Post Production •Production •Tips •Visual Effects •

Tiffen DFX 2 Digital Filters

Art Adams | 04/30

Take your footage to the next level with this simple post option

There aren’t a lot of controls in the typical Tiffen DFX filter, which is fine by me. The filters themselves are specific enough that you don’t need too many controls: the optical properties of each filter type are built in to the filter, so instead of constructing a look you are only finessing the strength, color, or where it’s applied.

I picked the “Glow” filter as a generic example. Here’s the original image:

This looks pretty nice, but it could look better. There are some small skin blemishes that could use some cleanup and the skin texture could be smoother. This kind of adjustment isn’t a “make or break” tweak, but it will definitely take this shot to a level beyond what I could capture with an inexpensive camera on a low budget.

Here’s what happens when I add the “glow” filter without any tweaking:

This is a bit heavier than I’m going for, but it gives me a coarse idea of what the filter will do. Let’s do a little tweaking with the filter controls:

The checkbox next to “Glow” enables or disables the filter’s effect. This is what I like the most about packages like DFX 2 or Magic Bullet Looks: you don’t need to leave Final Cut Pro to do your color grade the way you do in Apple’s Color. Everything is done in the timeline, which (after rendering) allows me to not only immediately see the new look in the context of the edit and the audio, but also allows me to quickly turn the effect off, or adjust it, if it doesn’t work as intended. Shots rarely stand alone and must be viewed, and graded, in context with the shots that occur around them. A professional colorist might be able to grade a completed project in one pass, but I tend to tinker and play with looks and it helps me to do this without leaving FCP.

Filters can be stacked and it’s simple to turn them on and off to see what combinations may work for the scene. Every change results in rendering time before the clip can be seen in motion, but results can be seen in still form immediately.

The “View” option determines what we see while tweaking the filter controls. The three options are “output”, which shows the image with the filter applied; “original”, which shows us the original image, or “selection”, which drops us into a black-and-white mode that helps us isolate specific parts of the image where the filter will be applied. More on this shortly.

“Blend” is a fairly powerful tool. The “add” option simply adds the bright glow effect to the image without much discrimination. This setting would work well if there weren’t many bright highlights in the shot already, as “Glow” would create them. IN this case, as I already have highlights in the image, the alternate option, “screen”, preserves the highlights and offers a higher degree of subtlety. For this image “screen” is the blending mode that we’ll use for the final look.

The “Brightness”, “Horizontal Blur” and “Vertical Blur” sliders that appear beneath the “Blend” pulldown menu only affect the areas to which “Glow” has been applied. That can be the entire image, or it can be portions of the image that we’ll select below. “Gang blur” syncs the horizontal and vertical blur sliders so they travel together, although if you’re going for an anamorphic look you could uncheck that box and increase the amount of horizontal blur without affecting the vertical blur.

“Color” determines what color the “Glow” filter highlights will be. That’s pretty handy. If you want the highlights to glow white, as they would when using white diffusion, then you’d leave this alone. If you wanted the look of a flesh-colored nylon stocking behind the lens you might want to change this. We’ll do that below.

The six options below “Color Correct” apply to the entire image. Sometimes applying a filter overall will dramatically change the image in some way, so this section allows you to do an overall compensation for the filter’s effect. For example, if you wanted to apply “Glow” to the entire image but the blacks became very milky then you’d be able to compensate overall by manipulating “contrast”, “gamma” and “brightness” here.

“Selection” is probably the most useful tool here. “Position” and “range”, however, aren’t the most intuitive labels for these controls, so let me try to explain what they do in better terms:

When using the “position” tool we need to change the “View” pulldown menu at the top to “selection”. DFX shows us this black-and-white image:

“Position” divides the image into bands of brightness, and by moving the “Position” slider I can select a range of highlights, mid-tones or shadows. The bright areas in the selection preview are what “Glow” will be applied to, while the dark areas will be ignored. Right now “Glow” will be applied to nearly everything in the shot to some degree because there’s very little that’s truly black (“unselected”) in this image. I want to isolate the flesh tone highlights, primarily on the legs, so I’m going to move the “position” slider until those highlights are the brightest things in the image:

You can see that I’ve still got too much of the image selected, but I’ve made the flesh tone highlights on the legs as bright as they can be. I’ve used “position” to determine which slice of luminance values I want to affect the most, and now I’ll slide “range” down to reduce that slice to just what I need:

I’ve retained the bright areas of flesh tone as selections while eliminating most of the folder and clothing. The martini glass will still get a little bit of glow but I’m good with that.

Here’s what the controls look like afterward:

There are a couple of things to notice here:

The color of the glow has changed. I used the eyedropper tool to sample the darker flesh tones and add that color to the filter effect. This gives the flesh tone highlights a bronzed effect.

I’ve changed the “Blend” option to “screen”, so instead of adding glowing highlights to already bright highlights and taking them way past clipping, the glow effect will be blended in a way that preserves the highlights that already exist without blowing them out.

You can see the new “position” and “range” settings. “Position” is set a bit lower than it was initially, and “range” is set dramatically lower in order to isolate a narrow swathe of brightness levels. (It might help to thing of brightness as a pie, with “position” selecting a slice and “range” determining how big that slice is.) I didn’t use the “blur” slider as that feathers the selected area and I wasn’t really interested in doing that. I’m being very specific as to what portions of the image I want to affect.

“Force 16-bit processing” is a neat trick: it takes your footage, which in this case is 8-bit DVCProHD, into a 16-bit color space for manipulation. After the effect has been applied the result is transferred back into 8-bit space on your timeline. This can help smooth over some odd artifacts that can creep in when manipulating 8-bit data, such as banding due to rounding errors. If you have a fast enough machine that the render times don’t become prohibitive, I recommend turning this on.

“Mix” affects the opacity of the effect. In this case I want it full up, so that’s about where it is. Adjusting this slider changes the filter’s transparency, with 0 totally eliminating the effect and 100 being completely opaque.

Here’s the result:

It’s a subtle effect, but then the best ones usually are. I don’t want to telegraph that I used special filtration, I just want the audience to respond to the look. I think you’ll agree that this is a silkier, smoother look than the original. Here’s the difference:

One thing that I’m not showing, due to my making the control panel as big as possible, is that every aspect of this filter can be key framed. Every adjustment can be set to change over the course of the shot. You can’t do that very easily with a glass filter.

Tiffen DFX 2 contains digital versions of popular Tiffen diffusion filters as well as a host of other effects that could prove handy. One of my favorites is “Halo”, which emulates the “diffusion in the telecine path” look. For those of you who weren’t around in the 80’s, putting diffusion in the telecine was quite popular: by putting a diffusion filter in the optical path while scanning color negative film the diffusion effect was reversed, so instead of causing highlights to glow the shadows bled into the highlight areas. It’s a very interesting effect. Give it a try in Photoshop by inverting an image, applying the “diffuse glow” filter, and then inverting it back to normal.

I’ve got the Final Cut Pro DFX plugin, but you can get it in a stand-alone version as well as in plugin form for Photoshop, Photoshop Elements, After Effects, Autodesk Combustion, and several flavors of Avid.

Art Adams is a DP who positively glows about post color correction. His web site is at www.artadams.net.

Wednesday, April 15, 2009

Filed under: Lighting •Production •Tips •

Rosco Shows New IR Filter (and other goodies) at NAB

Art Adams | 04/15

LitePads, IR filters, RoscoView make Rosco booth a destination

Tuesday, April 14, 2009

Filed under: Cameras •Production •Tips •

Gear Head for the RED

Art Adams | 04/14

Sacramento company GearNex brings smooth moves to your favorite HD camera

Monday, March 30, 2009

Filed under: Cameras •Production •Visual Effects •

Porsche F35 Spec Spot Revealed

Art Adams | 03/30

PVC readers get an early look at my finished F35 spec spot

Friday, March 27, 2009

Filed under: Cameras •Production •Tips •

IR Filter Cheat Sheet

Art Adams | 03/27

You asked for it, you got it: the non-brain-exploding “just what I need to know” IR filter cheat sheet.

Wednesday, March 25, 2009

Filed under: Lighting •Production •Tips •Training •

Mysteries of Color and Light

Art Adams | 03/25

What I learned after a year of developing the Kelvin Tile LED light, plus some other handy tips and tricks of light and color

A tungsten filament emits light across the entire visible spectrum. It’s spectrum is weighted toward the red end, which is why it is a warm light, but it contains every wavelength, from one side of the visible spectrum across to the other, in an unbroken chain. With that entire range available, any visible wavelength that an object can reflect is available to be reflected. This is not true of all LED lights, or indeed most HMI and flourescent lights.

The Kelvin Tile solved this problem by using six different colored LEDs together to create as full a spectrum as possible. By changing the brightness level of different LEDs we were able to skew the color temperature to any value between 2000k and 10,000k with very good color reproduction. We jumped through hoops trying to nail specific color targets in ways that looked appropriate on both film and HD, because while there are lots of color mixtures that look right to the eye there are much, much fewer that look right to an objective chemical or digital medium.

We learned a couple of things during all this testing:

There is a significant difference between light you look at and light you illuminate with. If you’re looking at an LED that contains red, green and blue dyes then you can mix any color your eye can see out of those three colors. But you can’t light objects effectively with that LED package because dyes emit light only on very narrow wavelengths, and if you try to light something that’s yellow with narrow spectrum red, green and blue LED lights you’re out of luck. It won’t work, because yellow falls between their spectra. A yellow object won’t reflect yellow light if yellow light isn’t present to begin with. White light is no guaranty of a full spectrum. The eye is easily fooled.

This has to do with additive versus subtractive color. When you look at a light you are seeing additive color, because as you add colored lights together they get brighter, and equal amounts of the primaries will create white light. We you look at an object you are seeing subtractive color, because when white light hits an object all the colors that aren’t contained in that object are absorbed, or subtracted. The remaining light is reflected and creates the color we see with our eyes. If you don’t illuminate the object with light that contains all the spectra necessary to reflect its color, then you aren’t going to see its proper color.

We discovered that using an emissive light color meter (it looks like a very large spotmeter that provides CIE XY color coordinates when aimed at a colored light source) tells you nothing about how that light will look when photographed. It’s possible to find a color of light that is “3200k” on a CIE chart and then use an emissive light color meter to make the LED’s emit that color of light. We tried that and ended up with light that looked good by eye but that photographed very green. Plus, getting the color of the light correct says nothing about the spectrum of that colored light. The key is that you have to create that color using the largest spectrum possible for maximum accuracy in color reproduction. And, lastly, getting the color right requires looking at targets being lit by the light, not by looking at the light itself.

For some reason that we never quite figured out, the Kelvin Tile saturates colors more than tungsten or HMI’s do. They “pop” a little more. It’s a very interesting effect, and it looks great on flesh tone.

We were never able to get the magenta target on any of our color charts to look right. After a bit of research we discovered that this was because there is no wavelength of light that is actually magenta in color. In light, magenta appears when there is an absence of green. If we lit a white wall with the Kelvin Tile set to 3200k or 5600k, and then dialed out the green LED, the wall turned magenta. That’s why there are minus green gels, but not plus magenta gels: you’re subtracting green, not adding magenta.

Litepanels has released an adjustable color temperature LED unit that apparently mixes 5600k phosphor LEDs with 3200k phosphor LEDs. Any color temperature in between can be attained by mixing the two color temps together. It’s a clever design, although I’m curious how full spectrum each of those phosphor LEDs are. (There are no spectral charts on their web site that I can find.)

On the next page: what’s the practical impact of knowing all this stuff?

Thursday, March 12, 2009

Filed under: Business •Web Video •

The Quicktime Conundrum, Part 2: Solved by our Readers

Art Adams | 03/12

My readers submit additional tips for flawless H.264 Quicktimes

Thursday, March 12, 2009

Filed under: Business •Web Video •

The Quicktime Conundrum

Art Adams | 03/12

How to get around Quicktime’s H.264 gamma bug

My first, and biggest, problem was figuring out how to export my video out of Final Cut Pro without seeing boosted gamma. In his article Chris Meyer recommends using X.264, an open source implementation of the H.264 codec that doesn’t play games with that hidden Quicktime gamma tag. I tried it out, exporting video out of Final Cut Pro via File>Export>Quicktime Conversion, and also through a preset I created in Compressor—and both showed the awful 1.8 gamma shift. Was X.264 really the answer? Was Compressor the real problem?

I exported a clip in ProRes and discovered that, when brought back into Final Cut Pro, the clip looked fine, but when viewed outside of Final Cut Pro in Quicktime Player it showed the same gamma problem. I needed an intermediate codec that retained as much image detail as possible but did not display a gamma shift. A bit of Google searching showed that others had wrestled with this issue and had settled on the Animation codec and the Photo-JPEG codec. Both retained proper gamma when exported from Final Cut Pro, which meant that further processing through X.264 via Quicktime Pro should maintain proper gamma.

And it did. Clips exported via the Animation and Photo-JPEG codecs held the proper gamma and retained more than enough detail to be re-compressed for the web. But that wasn’t the end of the story. (When it comes to post there’s always more to learn, isn’t there?)

The Animation codec is supposed to be very, very good: it’s considered a “lossless” codec, which means it will yield very large file sizes but at very high quality. As none of the projects I put on my web site are longer than 60 seconds, intermediate file size wasn’t a problem. (The largest intermediate file was 1.23 gigs for a 58 second clip.) I did have some problems with exported DVCProHD footage showing lots of jagged edges on diagonal lines, but changing the output codec to Photo-JPEG from Animation solved that problem.

The final step, encoding via X.264, was a little tricky. Read on…

Thursday, March 05, 2009

Filed under: Cameras •Lighting •Production •

Anatomy of a Porsche Spec Spot

Art Adams | 03/05

We push the F35 as far as it can go—and then some!

Joining me in crime were DIT Louis Block, camera assistant Rod Williams, and fellow DP Alan Hereford, who stopped by on his way home from a job to lend his support. Also helping were Ian’s friend Kevin Bailey, who served as VFX supervisor and starred as the driver, and a cast of a dozen actors and extras.

Left to right: director Ian McCamey, actor Patrizio, me, Louis Block and Jay Farrington.

The trick was trying to light 30’ of sidewalk with whatever lighting/grip was available at a camera rental house. Initially I tried using an 800w PAR aimed at the car just to see where the exposure fell and to see if I could pull off a very clean and neat single source look. While placing the hot spot of the PAR on the car lit it up to a reasonable exposure level, and the fall-off from the sides of the beam gave me consistent exposure down the length of the sidewalk, it looked nasty and lit. It was technically correct but artistically incorrect. So I changed tactics and decided to focus simply on making the shot look pretty and then figure out how to expose it properly once it looked good by eye.

Our AD relays Ian’s instructions to the actors via iPhone.

The result was two Kino Flo 4’x4 lights hidden against neighboring office roll-up doors, standing vertically and hidden by door jambs; the 800w PAR bouncing off two 4’x4’ bounce cards near the camera; and a open-face tungsten 1k hidden around the far end of the building lighting cars in the background. The Kinos were bulbed for daylight while the Par was gelled with double CTO, and I asked Louis to split the camera’s color between the two so we had a nice warm/cool mix. We also exploited a practical fixture that was mounted too high on the building for us to block: it was daylight-balanced and its light came from the same side of the frame as the Kino Flos so it blended right in. There were several daylight-balanced practicals hung over doors that we couldn’t turn off so we diffused them with one of my favorite diffusion materials ever: typing paper. The hard practicals became soft, subtle glows that added a sense of depth as the talent walked past them.

The idea was that the Kinos would provide some nice modeling from within the shot and light the area around the “club exterior”, while the warm Par provided a smooth base fill. (Soft fill light placed near the camera makes a lot of things look good in a hurry.) As people came closer to the camera the Kinos dropped off and the warm Par become a very pretty frontal key. As we were going to be shooting multiple passes for visual effects I wanted to light once and have that setup work for everything—not easy to do, but in this case hiding the Kinos nearly in plain sight worked perfectly.

The Kino Flos were eggcrated not so much to direct light but to cut down spill to the sides. The people on the sidewalk were much closer to the Kinos than the car was, and the light level on the car determined the exposure, so the eggcrates helped cut side spill onto the people and kept them from overexposing unless they stepped directly in front of the light.

Louis Block works his painterly magic.

Once the lighting was in place, thanks to Alan and Rod’s help, I turned to Louis to save me. We ended up pushing the F35’s gain to +6db and bringing up both the black stretch and the gammas to open up the shadows. We also changed the shutter to 216 degrees, or 1/40th/second, a mode I use quite often as this gives me an extra 1/3 stop or so and is a rock-solid flicker-free shutter angle. Boosting the gamma, black stretch and gain added a fair bit of noise to the image but it wasn’t a horrible amount, and we knew this spot would only ever appear on the small screen. (Ian is cutting the spot and he plans to work on the noise a bit in post.) Given the choice between a poorly lit scene with adequate exposure or a beautifully lit scene with a bit of underexposure, we opted for beauty and a little bit of noise.

The blacks and gammas have been boosted in this frame to show the noise that was lurking in the shadows.

We shot in standard gamma as the scene was fairly low contrast. Our shooting stop was T2, and the EI reading on the side of the camera said something like 650 or 670. When I went out to take an incident reading of the car with my meter set at ASA 640, I read T1 2/3 toward the Kino Flos and “E” toward the big bounce by the camera. The action looked much brighter on the monitor than it did to my eye. Just for kicks I changed the ASA rating on my meter, stored a couple of measurements and showed them to Louis and Rod while holding my finger over the ASA readout. I showed them T2.8 1/2 toward the Kinos and T1 1/3 toward the fill… and then moved my finger to show them that I’d set the meter’s ASA to 3200. They were impressed.

Now on to the shoot itself…

Wednesday, February 25, 2009

Filed under: Cameras •Production •Tips •Training •

EX1/EX3 IR Filter Shoot-Out

Art Adams | 02/25

The EX1/EX3 built-in Hot Mirrors suck. Or do they? One way to find out…

FIRST UP: SCHNEIDER OPTICS

The first filter I tried was the Schneider Tru-Cut 680. This filter has consistently worked on every camera I’ve tried it on:

It works perfectly, but unfortunately the dichroic coating on this filter is fairly heavy and causes vignetting at short focal lengths. Here’s the same filter on the EX1 at the short end of the zoom:

The cyan vignette is the result of looking through the dichroic coating at too sharp an angle. The Tru-Cut 680 is an excellent all-around IR/far red cut filter except for this one failing.

The Schneider ND 1.2/Tru-Cut 720 combo removes about half the red contamination on both cameras:

And the Schneider Tru-Cut 750 had no effect at all:

I was still stuck on my initial hunch that the EX1 and EX3 had either very poor or very absent built-in Hot Mirrors, but something about this test was feeling familiar. Still, I stuck to my assumptions as I tested the rest of the filter families on the next pages…

Monday, February 23, 2009

Filed under: CS4 •Post Production •Production •Training •

Fake Tilt-Shift Cinematography

Art Adams | 02/23

I’ve seen this in stills and it’s very convincing. Next week I’m doing to do it in motion.

Wednesday, February 18, 2009

Filed under: Cameras •Production •Tips •Training •

F35 IR Filter Shoot-Out

Art Adams | 02/18

We test IR filters from three major manufacturers to see which does what on the F35

IS THERE ANYTHING TO WORRY ABOUT?

We took a look to see if there’s anything to worry about when using ND filtration on the F35. Once again, here’s the camera with no filtration:

Here’s the same shot with a Schneider ND .30 filter:

At this point we saw a tinge of red contamination in the black DSC chart bag, which was quite surprising when using so little ND. It doesn’t read quite as strongly in this JPEG but it was very apparent on our monitor.

We swapped the ND .30 out for an ND .60:

Now we can definitely see the chart bag take on a reddish hue that wasn’t visible to our eyes. In the interest of time we jumped right to the heaviest possible filter pack we could use under our test conditions, the Schneider ND .9 and ND 1.2:

The overall red cast is normal. It’s extremely hard to make a neutral density filter that is truly neutral in color, and the more ND one uses the more obvious the color shift becomes. Due to the F35’s ability to see a much broader spectrum of color than other cameras, it is extremely sensitive to small shifts in hue that wouldn’t be much of an issue on other cameras. For this reason the F35 is equipped with a number of digital white balance slots to compensate for hue shifts in different filter packs.

The true test is when we white balance this image and see what color the chart bag really is:

And there it is: significant IR or far red contamination in the DSC chart bag. This image is much more subdued than what we saw on a CRT, where the black polyester of the chart bag was a rich maroon.

So far there aren’t any surprises here. Let’s move on to the filter tests and see if we can figure out how good the sensor’s built-in IR cut filter really is…

Tuesday, February 10, 2009

Filed under: Cameras •Production •Tips •Training •

RED Hot Mirror Shoot-Out

Art Adams | 02/10

We test Hot Mirror filters from three major manufacturers to see which does what on the RED

THE SETUP

First, here’s the scene we shot for all of our tests:

In the center is a DSC Chroma-Du-Monde chart. On the left side of the frame is a black nylon jacket, which we used as our far red/infrared target. On the right side is a bright red kimono, used to test whether the filters cut to far into the visible spectrum. As you can see from the slate, we’re using a RED ONE with software build 17, and while we technically rated it at EI 320 we really “exposed to the right” until the red channel was clipping in Rec 709 but not clipped when viewed in raw mode. (Raw mode shows what’s really happening, and I’ll frequently open up the F-stop until a color channel clips when viewed in RedSpace or Red 709 and then check that exposure in raw to make sure the channels are, in fact, not clipping.)

We shot this test under tungsten light in order to maintain a consistent look. Although far red/IR effects are most often seen in daylight when using lots of neutral density filtration, tungsten light contains just as much infrared and remains consistent over time.

We used an 85mm Arri/Zeiss UltraPrime, at the stops noted on the slate. It’s important to know that all of the filters we tested use a dichroic coating to reduce far red/IR contamination, and when viewed at an extreme angle they change colors, usually becoming cyan from the camera’s perspective. As an additional test we viewed each filter through a 16mm UltraPrime and looked for color vignetting around the edges. That particular lens comes with its own “darken the corners in telecine” look as you can see here:

When you see that degree of darkness in the flat field filter tests, please understand that this is due to the lens losing brightness around the edges of the frame and not due to the filter. The filters cause any color shifts that you might see around the edges of the frame.  (The flat field was created by covering the matte box with Lee 216 White Diffusion and bouncing a light off a piece of foam core three feet in front of it.)

Let’s proceed to the first round of tests, on the next page…

Monday, February 02, 2009

Filed under: Cameras •Post Production •Production •Tips •Training •

The Not-So-Technical Guide to S-Log and Log Gamma Curves

Art Adams | 02/02

What you need to know about log curves—with hardly any math at all

THE WASTE OF COUNTING BITS

Linear gamma images, in their raw form, are unwatchable: most of the image will appear very dark because the vast majority of luminance values in an average scene are going to fall into one quarter of the bits. Out of 16,384 steps per channel, flesh tone is probably going to fall around 2048 or so, which is pretty far down. Cameras apply what Sony calls a “coarse” gamma correction of .45 to this linear image in order to brighten up the shadows and mid-tones and make it viewable on any Rec 709-compliant TV monitor.

Side note: CRT monitors have a natural gamma response of 2.2, which is the inverse of .45. It was long ago decided that cameras would encode linear images with a .45 gamma curve so that the pictures would appear normal on CRTs. LCD monitors do not have this natural 2.2 gamma response so they are designed to artificially mimic it in order to stay compliant with existing standards.

The problem is that linear gamma is somewhat the opposite of logarithmic gamma, which is how our eyes see. Without getting into hardcore math, this is how logarithmic gamma works:

Imagine one lit candle in a dark room. If we light fifteen additional candles, one at a time, the sequence of total candles lit looks like this:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

By the time the sixteenth candle is lit you might assume that the room would appear to our eyes to be sixteen times as bright as it did with one candle lit. But because we see in logarithmic values, the room actually only seems four times as bright as it did when the first candle was lit.

A logarithmic change means that the apparent change in brightness between one candle and two is the same to our eyes as between two candles and four. Adding four more candles, for a total of eight, looks like the same brightness increase as increasing the number of lit candles to sixteen.

The top line is the number of candles, and the bottom line shows the perceptual steps where our eyes see the SAME apparent increase in brightness:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1 2   4       8                     16

And there’s the difference between linear and log gamma:

The sensor (linear gamma) counts ALL the candles, even though our eyes gradually stop seeing small increases in brightness due to the lighting of individual candles because each additional candle is perceptually too small to count.

Our eyes (log gamma) look at brightness in terms of how bright the first step is (adding a second lit candle to the first) and how much change needs to occur to see that step again (adding four candles to four, adding eight candles to eight, etc.).

Here’s another way to look at this scenario. In order for the eye to start at one brightness level and then see four additional steps of equal brightness:

1+1+1+1+1

We need to light the following numbers of candles:

1 2 4 8 16

It’s important to notice that the logarithmic steps get bigger as the room appears brighter, because that’s how we perceive reality. Our eyes are very good at detecting small steps in dark and mid-tones but the brightness steps get much bigger when we look at bright objects. This is the opposite of how a sensor sees, because linear gamma allocates the most, and smallest, steps to highlights and the least, and biggest, steps to mid-tones and shadows.

On to the main course on page three…

Sunday, February 01, 2009

Filed under:

No pain? NO GAIN!

Art Adams | 02/01

What adding and subtracting gain really does, and what you need to know before you use it

Thursday, January 22, 2009

Filed under: *VIDEO* •Cameras •Production •Training •

No 85 Necessary on the F35!

Art Adams | 01/22

Take the digital 5600k plunge—your camera will thank you!

The Sony F35 is probably the quietest camera in the world, which is why I really had to exaggerate to show this process in action. I shot a DSC color chart twice:

Under 5600k daylight, using D5600 mode with no filtration:

And then, under 5600k daylight, in tungsten mode with an 85 filter and all other settings the same:

(The chart shot with the 85 filter looks a little blue because the 85 only corrects from 5600k to 3400k. Full correction to 3200k requires an 85B filter.)

In Final Cut Pro I zoomed in 1000% so that the black chip filled the frame. Here are the resulting waveforms:

First, D5600 with no filter:

Next, tungsten mode with an 85:

The thickness in the blue channel under tungsten light is noise resulting from gain. In film terms,it’s almost as if the blue channel, and the blue channel alone, has been “pushed” a little bit. (These show that the daylight chart was a little warm, and the 85 chart was a little cool, but this was how they read on white balance preset in both modes with no additional painting.)

What’s interesting is that the red channel is much quieter in tungsten mode but shows a little noise in D5600 mode. (The expanded width of the red line indicates noise.) Under D5600 red and blue show about the same amount of noise, although blue is considerably quieter than when using an 85 filter. In the test with the 85 filter the red channel is very, very clean—although the blue channel is a lot noisier.

My opinion is that the increased red noise under D5600 isn’t anywhere near as much of a problem as the increase in blue noise with an 85 filter. Remember, though, I had to zoom in 1000% just to see this little bit of noise. The Sony F35 is the quietest camera I’ve ever seen, in both tungsten and daylight, and my preliminary tests show that I can push it to +9db before seeing any objectionable noise. And even then it’s quieter than any other camera I’ve seen at +9db.

What this should tell you is that you don’t need to carry an 85 filter EVER when using this camera. You won’t need to add any color correction to shoot in daylight, and take the resulting exposure hit, when shooting low light day interiors or at magic hour.

So next time you’re shooting in 5600k light with an F35, leave the filter case alone and instead go to the first OPERATIONS menu page and turn the “D5600” option to “on”. And make sure production knows that you’ve saved them a $10 filter rental.

Wednesday, January 21, 2009

Filed under:

Correction: Schneider Optics Sings the CTB’s

Art Adams | 01/21

In which I actually do the math to see how Schneider’s new CTB’s measure up to Wratten filters

Tuesday, January 20, 2009

Filed under: *VIDEO* •Cameras •Post Production •Production •Training •

Schneider Optics Sings the Blues

Art Adams | 01/20

Schneider’s new 1/2, 1/4 and 1/8 CTB filters have “RED” and “Phantom” written all over them

Tuesday, January 20, 2009

Filed under: *VIDEO* •Cameras •Production •Training •

Return of IR

Art Adams | 01/20

Three manufacturers are striving to create the perfect filter, and we’re going to help them

Thursday, January 15, 2009

Filed under: Cameras •Production •Training •

The Sony F35 and ND Filters

Art Adams | 01/15

Thought your ND filters were really neutral? Guess again—this camera can SEE the differences

Wednesday, January 14, 2009

Filed under: Cameras •Training •

The Not-So-Technical Guide to the Sony F35

Art Adams | 01/14

It’s not so hard, once you know all the secrets…

SLICED BREAD, MOVE OVER!

The F35 is amazing. Which it should be, as it costs around $200k. It is very much designed to be used film-style, utilizing a Super 35mm-sized sensor and no filter wheel. The built-in gamma curves will work to the satisfaction of nearly everyone, and the options for additional curves are attractive both in terms of cost and added benefit.

The most interesting part of the camera is the sensor. Silicon sensors only detect brightness, not color, so to derive a color from brightness only each photosite has a small colored filter over it that is either red, green or blue. By detecting how much light comes through the filter a processor can determine how much of red, green or blue is being detected.

Cameras with prism blocks can detect red, green and blue for every point in the image because the image is split three ways, through a prism, and directed to three different sensors, each with a red, green and blue filter in front of it. Each sensor sees a full 1920x1080 image, filtered by its individual color, and those three separately-colored images are combined into one image where each point has a real red, green and blue value.

Single-sensor cameras have a potential flaw in that they can’t see three colors in one spot at the same time, because each photosite, or light sensitive point on the sensor, can only be filtered with one color at a time. One common solution to this problem is the Bayer pattern sensor, which samples clusters that contain two green photosites for every blue and red one. (The RED is an example of a camera with a Bayer pattern sensor, as is the Vision Research Phantom HD, Silicon Imaging SI-2K and Arri D20/D21 cameras.)

Bayer pattern sensors work fairly well but suffer from lower resolution in the blue and red channels, as there are twice as many green photosites as there are for red and blue. Bayer pattern data also requires a lot of processing power in post: as each photosite samples only one color, a complex algorithm examines surrounding photosites and makes an educated guess as to what the other two color values should be for that one point on the sensor.

Keep in mind that a photosite is actually the physical point on the sensor at which light is gathered, which is not the same as a pixel. Pixel is shorthand for “picture element”, which is really a point sample of one or more photosites.

In the case of a Bayer pattern, for example, one pixel representing a green photosite will be assigned additional values for blue and red based on the values of the red and blue photosites surrounding it. The algorithm basically guesses at the missing values in order to make that green photosite appear as a full RGB pixel. That means storing less information during shooting because each photosite only has one value (either R, G or B), but it also means processing each photosite into a full-color RGB pixel in post, and that can take a lot of computer processing time.

The F35 RGB-striped sensor design is quite ingenious. It is arranged 2160 photosites vertically and 5760 photosites horizontally—or twice the vertical resolution and three times the horizontal resolution of a normal 1920x1080 HD raster. The photosite columns alternate colors: one column sees only red, the next sees only green, and the last sees only blue, and that sequence repeats horizontally across the surface of the chip.

Each pixel is created out of six photosites: two rows of red, green and blue are sampled vertically, and then those values are shifted horizontally to fall on top of each other. Imagine that these are the first two rows and three columns of photosites on the sensor:

R G B

R G B

First the two rows are added together vertically:

RR GG BB

Then the results are shifted on top of each other, so that R and G move right to stack on top of B. That becomes one pixel with a red, green and blue value derived from roughly the same spot on the chip:

RGB

Adding the signals from the two vertical photosites reduces noise significantly by doubling the color’s signal strength. Noise is random, which means that two random noise values added together never equals double the noise, so the color values are given a boost above the noise to create a very clean signal.

This is probably the most efficient and accurate method of sampling RGB pixel values across a single sensor. The usual three-chip prism block, with three sensors arranged around a prism block, provides three true RGB values for each pixel but suffers from two major anomalies:

(1)white shading errors, which are attributable to the use of an RGB prism block and result in green/magenta fringing around highlights, and

(2)increased depth-of-field due to using sensors that are roughly the size of a Super 16mm film frame.

The F35’s single sensor allows the use of 35mm film lenses and provides for 35mm reduced depth-of-field, which on the surface are the most obvious benefits of using this camera.

The problem is that the sensor is too good for our legacy broadcasting standards. The solution starts on the next page…

Saturday, November 22, 2008

Filed under:

Even Rembrandt Sang the Blues

Art Adams | 11/22

Great painters had a lot in common with modern cinematographers

Friday, November 21, 2008

Filed under: Lighting •

Lighting Advice for Budding DPs

Art Adams | 11/21

Wherein I realize I’m finally wise enough to give lighting advice to others

Sunday, November 16, 2008

Filed under: Cameras •Editing •Lighting •Post Production •Production •Training •

My First Shoot with the Sony F35

Art Adams | 11/16

In which a series of tests becomes the fastest spec spot shoot in history

Friday, November 14, 2008

Filed under: Cameras •Lighting •Pre-Production •Production •

The RED Outdoors

Art Adams | 11/14

This spot would have been hell if I hadn’t used the RED

Thursday, November 13, 2008

Filed under: Cameras •Lighting •Post Production •Production •Training •

Lighting Simply for the RED

Art Adams | 11/13

Two PARs, a couple of bounce cards and some grid cloth make this spot shine

This is the first shot in the piece, and the last shot we executed. During location scouting I noticed that the main living room window faced due west, and figured that I could use that to my advantage as the day wore on. I made sure that we scheduled our biggest shot for this room, the final shot of the piece where the actor heads toward the front door, first—so that we didn’t have to block out a lot of unwanted sunlight.

For this shot, done late in the day, I used the sunlight to my advantage: we hung an 8x8 grid cloth across the front of the window and used the sunlight to create a big, soft but directional source. A silver card provided the final touch by putting a highlight into the left side of the chair arm. The background was lit by ambient light.

I like grid cloth a lot. It’s very soft and has almost no specular quality to it, which means that it yields soft light with no hard light mixed in. An example of a soft diffuser that lets some hard light through is poly silk: the fabric itself lights up and creates a large source but if you look through it toward the light you can see a substantial hot spot where the light is. That hot spot will create a hard shadow.

Sometimes I like a bit of specular light. I frequently use a 2x3 or 4x4 silk next to the lens, filling actors from the key side, because the specular light will create a nice hit in the actor’s eyes. But in this case I wanted a nice smooth solid source. Why? I’m not sure. It just felt right.

If you like shots containing plants, you’ll love the next page…

(Page 2 of 6 pages for this article  <  1 2 3 4 >  Last »)

Back to Stunning Good Looks »   

                   

 

You must be registered to comment. This is an effort to reduce spam. Please REGISTER HERE.