A 180 degree shutter says, "Cinema." Here's how it works.
By Mark Christiansen | April 03, 2009
Quick: what's the difference between shutter speed and frame rate? You might be surprised that there are even a few camera operators don't even know the answer to that one. Frame rate is the number of frames per second that are shot (or displayed). Shutter speed is the amount of time that the frame is actually exposed to light, and thus, the amount of motion blur contained in that shot.
This free script will add tons of value to your multiprocessor system
By Mark Christiansen | April 02, 2009
My only assumption with this tip is that you use Adobe After Effects CS4 (or even CS3) on a multi-processor machine, Mac or Windows. Beyond that:maybe you only have one main machine and often face the dilemma of wanting to render while continuing to workperhaps you monitor your system's performance carefully and have noticed that your After Effects renders don't always peg all of the processorspossibly you own or have owned a copy or copies of Gridiron Software's Nucleo Pro and have experienced the joy of background rendering in After Effects already. However you're not experiencing that joy in CS4, because Gridiron has been too busy with another little project to update it.It could even be that you are aware that you can kick off an After Effects render in a shell (Terminal on Mac, DOS on Windows), allowing you to render without the GUI, and thus keep working. If so, if you're like 99% of visual artists, you're not that fond of memorizing, typing or optimizing code. If any or all of these is true, get ready to buy Lloyd Alvarez a beer, because he offers the answer to all of these and more. Lloyd's site is home of many useful tools, another of which may appear in this space this month, but as my first true tip of the month I wish to promote his most infinitely valuable script. And I say infinitely valuable because BG Renderer is offered free, a 100% discount off the alternatives.
Not an April Fool's Day Joke! 29 tips in 30 days from Mark Christiansen in Production Values
By Mark Christiansen | April 02, 2009
My turn! In April I aim to post a tip a day, despite the distracting spring weather, my birthday, and NAB. If you also get distracted, celebrate your birthday or just forget to check in, I will also ping these daily from Flowseeker on Twitter.I'm doing this as a kind of penance, and to test a theory. I admit that when I think of posting here as a founder of Pro Video Coalition I tend to think of crafting a long and/or weighty article. something more along the lines of what I would have sent the magazines where I used to write with some of these people. I also have 29 topics already outlined, many of which have been kicking around, and I've even grouped them (in my own mind, at least) into 7 subtopics, one for each day of the week. Of course, I reserve the right to change my mind, and to offer a May Day Bonus if the mood strikes me (and I haven't collapsed by then).You may expect these tips all to be about After Effects, since my book focuses on compositing and designing effects in that software, and quite a few of these will at least involve After Effects, but many will go further afield. If you're interested in more tips and workflow having to do specifically with After Effects, check out the new April term at fxphd.com beginning April 13, where I'm teaching a class that uses my book as a starting-off point, and delves into some of the many areas that are either easier to show than write, or just further explores topics that are worth exploring further.Ah, but if I didn't leave a tip today, that really would be like April Fool's, wouldn't it? Today's tip is not really a tip at all - at least, not like the others to be posted - it's my strategy on how I plan to weather the current economic conditions.
Lose less life working with R3D source
By Mark Christiansen | March 19, 2009
Ascertain the target output format at the outset (if you control your own format destiny, skip to 2)Decide who, if anyone, controls the color intent at the outset (if no one has specified how the footage must look on set, skip to 3)Leave R3D and 4K behind as soon as possible (and no sooner)Take control of sharpness, noise & more during conversion; don't leave these to tools that can't work with themMake use of great free and cheap tools if you can't just rely on ScratchThe RED ONE camera is innovative technology that will only improve. Someday, perhaps the good people at RED will anticipate the workflows most often used to work with their footage and offer specifications that help ensure their users' success; for now, however, the following incontrovertale facts about RED often guarantee the need to make decisions more or less on your own:no tools exist to write an R3D file (although several can read them), nor does RED endorse any alternate standard for converting their footagethe color intent of an R3D file cannot be controlled (although it can be specified)It is no coincidence that the happiest RED post-production pipelines have been the ones in which a given studio is master of its own destiny when it comes to the color look and output format of its footage. Likewise, hapless studios have in some cases encountered actual crisis when confronted by the demands of the director or DP to match the footage to how it looked when shot, or when required to deliver a particular format to another facility (typically for the purposes of conforming and finishing, often on a system standardized around 10-bit log Cineon DPX files).Things can only get better. Meanwhile, here are some tips to keep from going crazy.
New site shows why The Curious Case of Benjamin Button is the favorite for the VFX Oscar
By Mark Christiansen | February 07, 2009
Perhaps in part due to the impending SAG strike and the 30 or so films that have been put on hold as a result, 2008 was a somewhat thin year for tentpole visual effects features; on the VFX Show podcast where I'm a co-host, we had weeks at a time, particularly from late-summer until the holidays, in which very little was released for us to review. Despite that, the work that did show up was strong and the three contenders for the Visual Effects Oscar are all deserving of the award: The Dark Knight for visceral believable action and convincing comic-book additions, Iron Man for going appropriately over the top while obeying the laws of physics and optics.
But the winner should clearly be The Curious Case of Benjamin Buton, because that film did something no film before it has done, and the way in which it was made will resound for years to come. Button is truly the first film to leap across the uncanny valley and present us with a virtual leading man. You may or may not have liked this film - some people noticed its uncanny resemblance to Forrest Gump, and I am on record not having bought its fundamental premise - but just the fact that my quibble was with the premise, not the execution of a reverse-aging Brad Pitt, is one more sign that Digital Domain utterly pulled off a feat that has been anticipated for decades, a grail of sorts.
And that's not all. A freind recently pointed me to a beautiful site dedicated to the visual effects for the film, which lavishly presents behind the scenes breakdowns of the film's effects. Among the other many achievements of this film shown on the site are the many locations recreated with extensive matte painting work, including a full aerial recreation of post-war Paris (a city that hasn't changed so much since then, but which strictly prohibits flying at the altitude the director wanted due to post-9/11 restrctions); the many scenes on board a tugboat, including a battle sequence, in which I guarantee you'll be surprised to see what was shot on a stage and what was completely digitally fabricated; and last but not least the work of Lola, a studio often left off the credits entirely, to de-age the film's stars. This type of work is in nearly every other case kept secret; only with a film whose premise is the passage of time can their astonish transformations be brought fully to light.
Some people were profoundly moved by this film, others like me found ourselves wondering if it had been made primarily as a platform for the effects, but few will be seriously surprised when The Curious Case of Benjamin Button brings home the Academy gold later this month.
Red Giant Warp: Get More out of MochaAE
By Mark Christiansen | February 05, 2009
After Effects CS4 truly went from worst to first in one specific area: corner pin tracking. The old method of tracking and applying four individual points was so miserable I recall an entire class of my Academy of Art students unable to replace a billboard on the first try.CS4, however, includes MochaAE, the planar tracker that analyzes regions instead of points. Any unified object can be a "plane," even some whose contents are not perfectly co-planar; the output that is sent to After Effects is a Corner Pin track which can be pasted right to a layer.And that's where the ancient Corner Pin effect in After Effects falls short. Specifically:There's no way to import the MochaAE data via Corner Pin - the best method is literally to open a text file and paste it inThe layer receiving the Corner Pin track must fit perfectly into its four corners alreadyThere is no scale or offset control if the corners don't align perfectlyScaling up via a warp is not always pretty, especially because...Corner Pin animation data does not generate motion blurEach of these points is amply addressed by Red Giant's Warp plug-in, which is the key to geting the most out of MochaAE. This first video example shows the following benefits of the effect; in it, I take moving footage from a handheld shot of one screen and apply it to another, which would be a real pain without Warp.That's a straightforward way to use Warp's bells and whistles, but there's another approach you might have missed that comes in handy if you ever want to offset a track result within the same shot, and the camera is moving. By applying the data to both the From and To pins and cropping and offsetting the result, you can pull off a trick like this one, an instant set extension in which I grow a skyscraper.Without Warp, and despite that MochaAE can easily generate these tracks, these results would not only be more difficult to create, the quality of the output would be inferior, making this plugin a solid must-have for heavy users of MochaAE. Throw in the fact that you also get two other effects to generate reflections and shadows respectively (one of which was called out in a previous article), and the $199 price becomes all that much easier to take.
Economies of Scope
By Mark Christiansen | November 19, 2008
Six months ago I posted an article here with 10 reasons (and a bonus speculation, which now seems way off) why Scarlet would "change the game in 2009." At that time, Scarlet was RED's "entry level" camera, explicitly designed to capture indy filmmakers and "soccer moms" alike. When I posted the article, Scarlet had received far less press than the "bigger is better" Epic announcement; RED was announcing one camera that would eclipse the RED One, another which would significantly underprice all but the lowest end of the fixed-lens video camera market.The game has indeed been changed - by RED themselves, and only halfway to the finish line. The media seems to have a fondness recently for the phrase "we were all wrong" given world events of late, and now that the middle of the Scarlet line is now powerful enough to obsolete the RED One, it's worth looking at whether there's anything to get excited about at the prosumer end of the spectrum.Let's take a quick look at my top ten from back then:1) Sensor size - clearly this will be the salient characteristic of the initial RED line forevermore; with its latest announcement RED makes it clear that the Mysterium chip can scale to previously unheard of resolutions, and Jim Jannard has offered insight that resolution-wise, RED uses rules none of us was even considering.2) Price - the phrase 3K for $3K may not be gone but it is certainly forgotten. $3K now buys you a "brain" - no viewfinder, no grip, no recording device - in short all of the "sold separately" limitations Mike Curtis pointed out apply equally well to this humblest entry as they do to the truly configurable REDs.3) High-speed framerates - RED clearly maintains an advantage here, coupled with...4) Redcode which also delivers what looks to be vastly improved dynamic range. This is the part of the announcement that took me the longest to notice: 13 stops, which may mean you can shoot without clipping even on a bright contrasty location and retain color in the low midrange.An aside before I move on - in the previous article, I mentioned Redcode RAW and soccer mom in the same point. There you have the use that has been abandoned; instead of giving the DSLR-like option of shooting low-end compressed for users who don't need RAW all the time, RED remains firmly The Post-Production Camera - NOT for use by the novice until a drastic revamp of the pipeline occurs.5) RGB via HDMI - It's hard to believe I considered this worthy of the top ten, but at that time it did seem notable that RED would provide an easy way to pipe video straight out of the camera. Again, I no longer see this as a priority - until it is.6) Plug and play - yeah, well, no. This is tbe big downer to the crowd who was hoping for the $3K camera that would actually eclipse everything else at that price; the price has effectively gone way up, because while $3K may still end up buying you 3K, you won't have the viewfinder to see it, the device to record or transfer it, or even the grip to hold it without adding extra dough. Or so it seems.7) Fixed lens - what was an acceptable and even enticing trade-off - a fixed lens with optics that Jim Jannard would accept - now just looks like you're a chump if you end up with the one and only RED that doesn't have the ability to mount just about any lens you would ever want, leaving you to fumble with your Brevis or Letus for customization. Likewise, the 2/3 sensor to match this lens now sounds like it will cause you to miss the proverbial boat, not of image fidelity but of cinematic quality, especially compared with the almost-good-enough DSLR's now learning to do this stuff.8) Customization - good lord, this one now belongs at number one with a bullet.9) Controls. Who can say? The controls for the previous camera certainly looked enticingly ground-breaking, and Jim has not lost his affinity for posting enticing renderings of physical interfaces. Just because all we get to see right now is a brain in a box and some lovely renderings (see above) doesn't mean that Scarlet won't innovate this area significantly, especially when you think of actually shooting a movie with a DSLR.10) Fearless market position. Well, yes. Some things never change because they are fundamental. It's notable that RED is willing to roll out these sweeping changes months ahead of even the earliest arrivals. It's almost as if Jim Jannard, who clearly isn't in this just to make a buck, is challenging the Japanese camera manufacturers to come up with viable competition - and thereby compete with their own obsolete product lines as well.Still, lovers of cameras like the Sony EX-1 have lost a dream with this announcement - the dream of heading down to the Apple store (my now ludicrous sounding +1 point) and for less than ten grand, picking up the entry-level post-production camera and a system built to finish film-quality images.So the RED is dead to consumers. Long live the RED! I for one welcome our Mysterium overlords; they have given ample room to the Sonys, Canons and even Nikons of the world to answer the challenge for the prosumer while remaining well in reach as an easily rentable, astounding, game-changing imaging device. My +1 guess that RED would need the Apple Store to distribute it is no longer viable; RED no longer has to solve this problem, because for the time being it is no longer most interested in getting tens of thousands of cameras into the hands of non-professionals, a game to the bottom of which these other companies race one another. RED appears smarter than that.Now if they'd just standardize the workflow.
Billions were fooled, but was any real harm done?
By Mark Christiansen | August 13, 2008
If the 21st Century, as I think it very well may, becomes known as the era of Things Are Not As They Seem (if the acronym TANATS catches on, you heard it here first), maybe we'll look at one seemingly harmless moment in 2008 as a watershed.
Perhaps you were one of the billions worldwide who witnessed the Opening Ceremonies of the 2008 Olympics and was fooled by an aerial sequence of 29 giant pyrotechnic footprints leading to Beijing National Stadium, of which only one - the very last - actually occurred as depicted. The preceding 28, representing the Olympic events preceding this one, relied on our old friends the particle simulation, motion control and compositing (not to mention a glance at the Farmer's Almanac for the likeliest weather conditions on the night of the event).
This of course is one more sense in which they would have been really screwed if it had rained that night (as it did all weekend) - and maybe you weren't aware that Chinese officials are even trying to control that.
Not that there aren't other precedents of misrepresentation in the name of controlled spectacle: these organizers evidently wouldn't let the girl who sang the patriotic "Ode to the Motherland" appear as the singer, apparently because among 1.3 billion Chinese there are no sufficiently cute seven year olds with a set of pipes. It's encourage to note the complaints among the Chinese about the message this sends to gifted singers with ordinary looks.
That substitution may lead to some serious need for therapy later in life - by comparison, the fireworks stunt seems harmless enough on the face of it. Sure, one could criticize that:even the nation that invented fireworks couldn't pull this off (and it may be significant to note that they really did create firework footprints that night, they just didn't photograph them, evidently due to the hazards of combining aerial photography and pyrotechnics) vfx work is reported to have taken a year, which seems to me bad P.R. for the nascent Chinese visual effects industry (yes I'm joking, but if you land a year's budget for a sequence like that, call me) even though we're used to computer graphics on our screens all the time, everywhere, this show was just one more big spectacle, so what's a little alleged live TV fakery (hello David Copperfield!) However, is there any doubt that in this Era of TANATS, something like this will cross the line? It's easy to dismiss claims that the moon landing or Zapruder film were composited - we humans simply weren't very good at that sort of thing in the 1960's. That's a pretty thin argument nowadays, when images continue to shape our lives despite how used we are to their fabrication. My kids routinely ask if fantastic images - including the real ones - are real or fake, and have done so since preschool.
So maybe the question is when will be the first time billions of people are fooled by a fake live transmission and it actually matters to what we think about justice, right and wrong, good or evil?
That sounds grandiose. But this being the Olympics, one need look no further than the athletes themselves, and the question as to whether they have faked their performance with chemistry, to glimpse what a mess is created when we hope the rules of the past will get us through the reality of today, and tomorrow.
NOTE: No less a filmmaker than Errol Morris has posted a thorough and thoughtful blog entry on a related topic.
As the standard is devalued, the world undertakes a slow-motion search for an alternative. What can be done for QuickTime?
By Mark Christiansen | July 12, 2008
How bad are things for the US Dollar these days? So bad that, as reported by the BBC and mentioned recently on This Week in Tech, Gisele Bundchen no longer accepts modeling pay in dollars, nor apparently do many high-end boutiques in the capital of U.S. commerce, New York City. European travel is effectively twice as expensive as it was just a few years ago simply because of the exchange rate. So it may come as a surprise how familiar the situation of the world economy in regards to the dollar is if you're a video professional using QuickTime. I'm not actually joking.
Don't overlook the impact of RED's entry-level camera, even for pros like you.
By Mark Christiansen | May 07, 2008
RED will not debut Scarlet, its entry-level camera, until early next year, which means that those who are already hype-weary with all things RED are already moving on to a different part of the site. Nontheless, there are solid reasons that Scarlet will change the entire landscape of low-budget digital video, assuming RED can get enough of them into the hands of the public (more about that at the end). Scarlet's impact will be somewhere between that of the Canon 10D when it debuted and that of the iPhone. Here's why.1) 3K native sensor. It's easy to lose sight of how major a step forward 3,000 pixels of horizontal resolution is for a digital video camera when that camera is debuted alongside an existing 4K camera and a 5K camera due at the same time. So let's try this with the hype language used by the digital still camera manufacturers: 8.5 megapixels. Per frame. And this is not cheating by calling a 1280 native sensor HD (yes I'm talking to you, HVX-200). 3k means you could cut this image down 35% to HD. You could sneak it onto an IMAX screen without anyone seeing pixilation. It will be years before Sony or Panasonic have anything like this at the price. Speaking of which...
Great strides are being made in 3D technology, which is great news for the few projects that need it.
By Mark Christiansen | April 22, 2008
Day One of NAB was 3D day in the Content Theater, and at the end of that day, I participated in a podcast discussion with Ron Brinkmann and Mike Seymour all about stereo imaging, which seems to have been a major official trend at NAB this year. At the end of the on-air discussion we did an informal poll and found that half the folks in the room felt that 3D display would be a bona fide new artistic medium within the next few years, and the other half - with whom I'm aligned - felt that stereoscopy remains the fringe curiosity it has always been, at least in regards to narrative filmmaking. U23D as an experience is hard to top - I even overheard other moviegoers express how preferable seeing that movie in IMAX was to actually attending a U2 show - but the legitimacy of 3D display for a mainstream Hollywood film is only being debated, it seems to me, because James Cameron has put the issue on the table.My thought is that even if Avatar does revolutionize the use of 3D in tentpole Hollywood blockbusters, few (if any) other filmmakers will be able (or willing) to match, let alone top it. Also, as long as headgear is required to view images in 3D - and the polarized specs I received for U23D take the cake - we humans are really no further along technologically than when we started down this road (although the imaging technology itself for the price point is improving by leaps and bounds thanks to IMAX and high-frame-rate digital displays).Until it's a hologram, no goofy glasses required, any move toward 3D is going to look like what it's always been - something cinemas grab onto when sales are under downward pressure. 3D can be fun - 10 year olds love the 3D films I worked on for Robert Rodriguez - and James Cameron will no doubt provide an amazing experience, with or without the extra dimension - but 3D on the projected screen is is a boon to immediacy and sensation, not emotion and story. U23D in IMAX is a fantastic spectacle - better in many ways than actually attending a U2 show - but if we see more of these in cinemas, it can only mean that substantial stories have headed for the more intimate home theater.
By Mark Christiansen | April 20, 2008
According to a highly reliable source, Apple emphatically denied in an NAB press meeting earlier this week any truth to the rumor that the Pro Apps are for sale.Now, even if Apple sold Final Cut Studio and its brethren, I hardly imagine the buyer (and who might that be? Avid?) would do much to mess with its success, at least in the short term - but here is a company with a firm policy of not responding to rumors that is apparently issuing an emphatic denial and, well, one can't help but think of analogous situations that have occurred in the recent past.One thing that is certain, from speaking with a former Apple employee, is that the development team, once housed inside 1 Infinite Loop and very much on the radar, is now located far from the action. If you don't think that matters, keep in mind that this is a company whose important decisions are all made by one guy. In that building. It may simply be that Pro Apps were important to Apple back when they needed to show that the Mac was not an inferior platform for high-performance, high-profile entertainment work, but that job has now been done, the app is mature and in no danger of leaving the platform, and the company has found other areas to, ahem, shake up. Like the multi-billion dollar mobile telecommunications industry, or the multi-billion dollar entertainment industry.Anyhow that's all just speculation. Beyond the denial, all I know for sure is that a) Final Cut Server looks great and b) attending NAB after losing my iPhone in the powder at Sugar Bowl was like living in L.A. without a car.
At the show ruled by suits and dilberts, the rebels get their turn
By Mark Christiansen | April 19, 2008
With big exhibitions like The NAB Show falling out of favor, and some disappearing altogether in the 21st century, NAB struck back this year by offering more educational fare than in years past, featuring keynotes and panels of experts from the industry, as well as day-long classes. On Wednesday was "A Million Dollar Look on a Thousand Dollar Budget," a keynote and panel on getting cinematic production values out of equipment you may already have sitting around your studio.The session was kicked off with the Legend of Zelda fan trailer "linked" here (nerdy pun for gamers paying attention) which appeared on April Fool's Day, followed by a keynote by Stu Maschwitz of The Orphanage (and author of a fantastic blog) and then a panel featuring Dave Basulto of Clarity Pictures, Alex Lindsay from Pixel Corps, D.P. Taylor Wigton (447 Productions) and moderated by Brian Valente from Redrock Micro.
Network Television Finally Reaches the Web Legally
By Mark Christiansen | March 12, 2008
As I type this, Hulu is temporarily closed as it is prepared for public roll-out on Wednesday.What, you might ask if you're too busy creating images to watch them on a beta website, is Hulu? As best I can tell, it is an attempt by network television - specifically NBC, although other networks are apparently invited to join (and ABC and CBS have not as of yet) - to take control of how its content - its shows - can be viewed online. Not offline - this is not IPTV, where you can order up a show to watch on your 50 inch plasma - instead, it's a direct response to YouTube as the online home of every type of video. Instead of continuing only to demand the removal of network shows from other services, Hulu attempts to be the place to go to get them, and it makes its money back via advertising. In some ways it's a brand-new model, and inevitably in other ways it looks an awful lot like the old model.
adobecards.com: sign of things to come or a splashy tease?
By Mark Christiansen | March 06, 2008
Now if only you could create something like this entirely with Adobe software.Or does the joker indeed get the last laugh?
Parody illuminates two contrasting styles
By Mark Christiansen | March 04, 2008
A friendly reminder: if you want your motion graphics work to be shared far and wide, humor and satire can be a great way to go.Note: the AEList apparently beat me to the punch on this one... s'okay, wanted to test video embeds here anyhow...
For once a great performance is not overlooked simply because hardly anyone saw it.
By Mark Christiansen | February 26, 2008
Visual Effect Academy Awards™ are not much different from any other category in at least one respect: great performances in films that underperform at the box office tend to be overlooked. I and many others thought that Transformers had this year's visual effects Oscar™ all sewn up not only because the work was amazing - not just the amazingly complex 3D animation but some really fantastic compositing. Pirates of the Caribbean 3 (on which this author contributed a few shots) was clearly not going to win as that would break an Oscar taboo: the repeat winner (since Pirates 2 took a statue only last year). And yet, nearly as much of a long-shot seemed to be The Golden Compass simply because the film was a flop, and Hollywood is allergic to losing money (despite many examples to the contrary) - this despite that many in the visual effects community believe it contained the most ground-breaking work, raising the bar for complex interactions between computer generated creatures (realistic looking daemons, the animals representing the soul/anima of the human characters) and recreating grand scenes of steam-punk London and Oxford and grand vistas of the Arctic. Not since What Dreams May Come has a vfx film lost money at the box office and taken the statue.Perhaps Hollywood's love of giving the prize to anyone but ILM - who along with the 49ers were the bay area force no one could beat in the 80's and early 90's - trumped the box office vote. However it happened, a great visual effects film (albeit a failed re-telling of one of the best novels of the past decade) won.
Adobe's foray into the hardware market, or just a cool peek into the near future?
By Mark Christiansen | February 13, 2008
Check out this video in which Adobe engineers debut some much rumored hardware designed to interface with a future version of Photoshop (and, perhaps, other of our favorite Adobe apps?) to provide persective and depth of field in post.Not only is it a stunning demo, such a lens might not require mechanical controls for changing focus or aperture; the lenses are the fixed focus type found in your point and shoot camera.It's possible to recreate depth of field in post, or do without it entirely, Gregg Toland-style (but since we all know that DOF is crucial to a cinematic look, that ability may be more useful for 3D artists who need that kind of image fidelity).Adobe has vastly increased its investment in research and development so this is hardly the last innovation we're likely to see, particularly in cases where the proof of concept comes from academia.
David Fincher and Gus Van Sant take contrasting approaches to recreating the oddest decade of a unique city.
By Mark Christiansen | February 05, 2008
Every day that I cycle from my home into the Presidio I pass through the intersection of Washington and Cherry streets, site of a murder that is the centerpiece of David Fincher's film Zodiac, a film that fabulously recreated the San Francisco of the 1970's. Last weekend I cruised Castro Street (not what you're thinking when you read that) to witness its own fabulous 1970's makeover for Gus Van Sant's biopic Milk, about San Francisco's most famous murder of that era.There is great irony here: the corner of Washington and Cherry has the timeless quality that goes with old-money upper class neighborhoods, yet Fincher chose not to shoot there at all, instead painstaking recreating the neighborhood as 3D matte paintings and shooting the taxi driver murder in front of a green screen. These allowed them to add period details that perhaps no non-local filmgoer could notice, such as that the street trees would have been 35 years younger - this for a scene that takes place at night.Castro Street, meanwhile, is as different from its 1970's self as any vibrant commercial tourist destination would be, and yet for Milk, the filmmakers are going back in time using set decoration: redoing the storefronts that have changed hands (you have to hand it to Rossi's Deli for appearing virtually unchanged in 30 years) and repainting the Castro Theater, the cinema that is the neighborhood's visual centerpiece. Hilariously, they have taken the level of detail right down to the real estate listings (in the window depicted below).In an era when even a romantic comedy has 2-300 digital visual effects shots, what's up? I'd like to hear your suggestions in the comments, but I think it's mostly a question of taste, or even what you might call comfort zone. Yes, the director of Se7en and Fight Club can afford have an artist or two spend a year of their lives working on one effects shot, a time-lapse of the TransAmerica Pyramid being constructed, even though it's tangential to the storyline at best, so you could follow the money and simply say that Zodiac was a bigger budget film. But locations aren't cheap, especially in high-end coastal cities.So this is really a study in contrasts. Fincher makes an investigative drama and can't help but insert almost-unreal effects and even motion graphics into the story (at one point the letters from the Zodiac killer occur as three dimensional projections all around the offices of the Chronicle). Van Sant, on the other hand, is an old-school actor-centered independent filmmaker, just the kind who really hates green screens and handing over key shots to digital artists. I don't doubt there will be effects shots in Milk, but it looks like the heavy lifting is happening in "pre," not post.