Stunning Good Looks

Okay, I have to admit it: I now like this camera more than I thought I would.

RED ONE Build 15 EI Test

The idea behind this test was to underexpose a neutral reference and see at what point the RED’s noise becomes too much of an issue. Using Colorista in Final Cut Pro, I put 18% gray at 50 units on the waveform monitor for each exposure, effectively “printing up” each exposure to see how much “grain” (noise) showed up. Some of this is exaggerated by compression artifacts, but I think you can get a good feel for what’s going on.

I’m not sure how this would look on a film out, but to me there isn’t a ton of difference between noise levels at 320 and 640, although I think I’m more comfortable rating the camera at 320. I start seeing reduced contrast at EI 1280, which I really wouldn’t want to try on a paid job.

Compressed via FCP Compressor, H.264.

A while back, while browsing Reduser.net, I saw that someone had done an IR test and discovered that RED has a weakness when it comes to IR filtering. I did my own tests and here are the results:

First shot: RED Build 15, Redcode 36, processed in Red Alert (contrast +.7, brightness +3.0, Kelvin 5130k, all else at presets)
Filtration: Schneider Optics ND .30:

Second shot: same as above, but filtration increased to ND .9 plus ND 1.2--a total of seven stops of ND:

(left to right: PVC’s own HD guru Adam Wilt, and DIT/video engineer Jay Farrington of Chater Camera, San Francisco)

The theory is that when ND filters cut visible light to the point where the ratio between visible light and IR light is very low, the IR shows up much more dramatically than it would otherwise. Normally IR effects are overwhelmed by visible light.

I was a bit confused by this image. The darkest object in the shot--the filter pouch--is made of synthetic fibers and is bright magenta, making me think that the magenta cast in the second picture is an IR issue. But the picture has a magenta cast overall while not everything in the image is radiating the same amount of IR.

I’ve been interviewing Ira Tiffen, formerly of Tiffen, Inc. and a world-renowned expert in the creation and manufacturing of filters, for a magazine article on HD filtration that I’m writing for HD Video Pro magazine. I sent him the pictures above and asked for his expert opinion. This is what he had to say:

-------------------------------------------------------------

I am making certain assumptions which are commonly encountered. The camera has a built-in IR reduction filter. It is essentially clear to the eye, transmitting most of the light between 400nm and 680nm, which is the visible spectrum, minus the part between 680nm and 700nm. Once it reaches 680nm, the filter begins to reduce light. From 680nm to 700nm, there is a little red lost (which is why this type of filter has a very pale green cast to it), but further above 700nm there is a quickly diminishing transmission of light, all in the IR region. But there IS some transmission from, say 700nm to 730nm, causing a small ‘leak’ in IR. At some higher point, say 900nm, it may start to transmit more light again, but the camera isn’t sensitive to it.

We have to remember that the chip is generally going to have more sensitivity to light in the IR than film does. So a small ‘leak’ in light just above 700nm can be picked up as red by the camera, but with the ND 0.3 filter, at only one stop of light absorption, there is still an abundance of visible light to overpower any leak of near IR getting through the IR filter.

If the camera did not have the built-in IR filter, there would be a much greater amount of near IR light, enough to show a reddish cast on objects (usually black or dark gray) that tend to reflect more IR (remember that dark objects get hotter in the sun than lighter objects- this heat is the excess of reflected IR) and proportionately less visible light. This is why the effect appears even without ND filters when there is no IR filter in the camera.

Here’s what happened in the scene with the ND 2.1 (7 stops, 0.8% transmission) filter combination. All organic-dyed ND filters I have seen, including Schneider’s, begin transmitting an increasing amount of light from about 680nm and up. When there is no ND filter, or a ND filter not much denser than, say, ND 0.9, the amount of additional light in the far red (680nm-700nm) is proportionately too small to be seen among the much greater amount of visible light.

However, with the ND 2.1, the amount of visible light is brought down to less than 1% of the original. The small amount of near red is not affected as much because the filter dyes do not function well in that region. This is the key. So the small amount of far red that gets through is now much greater in proportion to the overall, including the visible, and so a reddish cast is created. Now Schneider ND’s tend to be cooler (read a bit bluer) which is an increase of light transmission just above 400nm, which acts to minimize the reddish cast that would otherwise show up throughout
the image.

In the ND 2.1 filtered image, there is an overall addition of blue in the areas that do not reflect much IR, like the bricks, the silver car, and most other areas. However, in the black filter case, and the gray jacket and sweatshirt, there is a greater amount of light in the far red and IR being reflected, so the reddish cast on those objects becomes enough to overpower the addition of blue that neutralized the red everywhere else.

To recap, without an IR filter the camera would suffer from an excess of red especially in objects that reflect a large amount of IR light. Even without a ND filter.

With just the IR filter, there will generally be a good balance between the visible and the near IR, so that color rendition is normal. The small amount of far red and near IR is overpowered by the large amount of visible light.

With an organically dyed ND filter that is not much denser than a ND 0.9, there is still enough visible light to counterbalance the excess amounts of far red and near IR.

With a much denser ND filter, the overall amount of visible light is reduced enough in proportion to the amount of far red and near IR getting through that there can be a reddish cast through the whole image. With a cooler ND, with some addition of blue, the color can appear normal for objects that do not reflect excessive IR. However, for those objects that do, the amount of IR is enough to still cause a reddish cast because it is then proportionately greater in amount than either the visible light or the increase in blue that worked to counteract it for parts of the scene that do not reflect much IR.

-Ira Tiffen

-------------------------------------------------------------

I’m extremely thankful to be able to publish this incredibly in-depth explanation of what was going on in the images above. It just shows that no matter how long one works in the industry one can never say, “Now I know it all.” There’s always more. I just love that.

“Now I’ve seen it all” is a completely different matter.

Ira Tiffen is currently working on a book about filtration. He has a publisher but no firm release date. Let’s hope it comes out in the next year or two. I, for one, am dying to read it.

I’m told by Bob Zupka, of Schneider Optics, that they have a filter, called the 486, that was designed for early digital still cameras that didn’t have any effective IR filtering and were extremely sensitive to IR. This filter has a pale pink cast, much like a UV/haze filter, and cuts the spectrum like a knife above 680nm. The cast is easily eliminated by white balancing or basic color correction. Expect to see some HD-size 486 samples at the Schneider Optics booth at NAB. I’m hoping to test one with Adam Wilt shortly thereafter. Stay tuned!

RED ONE Build 15 Flesh Tone Latitude Tests (tungsten)

In my quest to figure out just how much overexposure latitude I can get out of a RED under tungsten light, I decided to do a test with a flesh tone reference and a RED with build 15 loaded. It turned out to be almost a completely different camera.

Previously I was only able to see two stops of overexposure latitude on a Kodak 18% gray card under tungsten light. That was using build 14. This build seems to yield 3 stops of latitude before clipping flesh tone, which already has more red in it than a gray card. Also, instead of the red channel clipping and causing highlights to turn cyan until the other channels clipped, I see no indication of that at all now. The clipped highlight is very clean and holds color a lot longer than I expected.

On top of that, the DRX highlight recovery tool in Red Alert now seems to cause more problems than it solves. When I used DRX on these clips they turned green, the exact opposite of what they’re supposed to do--which is to take red channel clips that have turned cyan and use the other two unclipped color channels to rebuild detail in the clipped channel, while using the white balance meta data to blend all of that into a neutral highlight.

I don’t know what they did to keep the highlights so clean, but I’d be willing to take this camera into a real world situation now. The only thing that bugs me is that the red channel, which used to be the cleanest color channel by far, now seems very noisy. Maybe that’s a side effect of whatever was done to clean up the highlights.

Hi all- I apologize, but I’ve been swamped with work of one sort or another, and when I tried to sit down and start sorting through the latest RED test footage today I couldn’t finish it. I took a nap instead. I must be getting old!

I’ve got three days of shooting this week, and hopefully I’ll be able to pull the footage together and post it by the end of the week. If I were to speculate wildly about the results of the latest round of RED testing, I’d say the following:

The RED definitely has an infrared problem, as previously seen on Reduser.net. We put 7 stops of ND on the lens and saw very severe magenta color shifts on black cloth due to IR contamination. Apparently if the ratio of visible light to infrared becomes too low the RED sees the IR quite easily. I understand that at least one filter manufacturer is currently working on the problem.

We tested build 15, which seems very different to build 14. On build 14 we saw severe problems with red clipping under tungsten light that turned highlights cyan and required highlight recovery work in REDCine and REDAlert to bring back highlight detail. On build 15 it seems that a lot has been done internally to eliminate this problem: the camera seems to hold highlights vastly better under tungsten light, and highlight recovery actually seems to make things slightly worse. There’s obviously something different going on in the camera with this build.

I think there’s some additional processing going on in there, particularly with the red channel. It’s interesting to note that build 14 saw the red channel being the quietest, with green second and blue the noisiest. Now green is the quietest (although still fairly noisy), with red being noisier and blue about the same. Something seems to have happened to the red gain. Also, preliminary histogram examination shows red rolling off around clip instead of hard clipping, which is interesting.

It’s a completely different camera now. I hope to post some build 14 and 15 comparisons and more data later in the week. Stay tuned.

Thanks, as always, to camera guru Adam Wilt for his help and photos. Also thanks to DP Alan Hereford, DP-in-training Ted Allen, and video engineer/co-owner Jay Farrington of Chater Camera (chatercamera.com).

I’ve also got a new showreel coming together, so I suspect I’ll have a lot to say about that process at some point.

I’ve now had independent confirmation from a couple of people who watched my RED build 14 exposure tests and have had great luck treating their RED as a daylight film stock. Apparently clipped highlights are significantly less of a problem and noise in the shadows drops way down. It’s a much smoother look.

If you haven’t had a chance to test your RED before shooting to see how well you can recover clipped highlights, it might be a good idea to go HMI or daylight if at all possible. I’m hoping to test recovering clipped highlights and blue filtration by the end of the week, so stay tuned.

Meanwhile, look back a couple of posts in this blog if you need to learn how to recover clipped highlights in Red Alert, and remember that the same control in Red Cine, called “highlight,” does work--but you need to “jiggle” or double-click another control before the new value takes effect. That is slated to be corrected in the next build.