Linguists would have had a field day at The Foundry’s stereo 3d Nuke Masterclass at Vue, Leicester Square today, observing the fascinating alien language used when referencing the how’s and why’s of all things stereo 3d.
The phrases being banded around in the packed cinema, perhaps predictably full of mid-20s to 30-something males, were beyond comprehension to all but the relatively few DoPs, camera operators, techie geeks, vfx artists and producers who’ve already tried to tackle a full-scale stereo 3d project from acquisition through post to delivery.
Try these for size… “disparity field”, “interaxial separation”, “monoscopic depth cues” and “interocular distance”. Not to mention the ‘basic’ stuff such as “positive and negative parallax” and “convergence points”. This stuff isn’t for the faint hearted.
The Foundry’s Simon Robinson started off proceedings by trying to explain a lot of these terms, with the aid of some handy diagrams. Personally, I was still pretty confused despite his valiant efforts, but he was preaching to the converted, with many audience members already fairly well versed in stereo 3d geek speak.
The first proper talk of the day was by Andy Miln, director of stereo 3d specialist, Inition 3d, about how best to shoot in stereo 3d. His detailed examples elicited loads of complex questions from the 3d-educated audience, addressing the finer details of using different cameras, mirror rigs (which apparently reduce those aforementioned “interaxial separations”), doing focus pulls and things like that.
Then Framestore’s Theodor Groeneboom, who worked on Avatar, took us through using Nuke to do stereo 3d post production. He made the point that it doesn’t matter how well prepared you are on the shoot and how many times you’ve tried to eliminate potential problems, a tonne of post production cleaning up work WILL be required to bash the stereo effect into shape and make it all comfortable to watch.
The main culprits requiring post work are colour mismatches and vertical misalignments between the left and right eyes and unmatched lens reflections. All are fixable using quite fiddly post techniques, but the message (and this shouldn’t come as much of a surprise) is that stereo 3d is heavily demanding of time and effort in both production and post.
After a morning spent at this Masterclass, the best advice for doing stereo 3d would be to surround yourself with people who’ve had proper, in depth prior experience of stereo production.
Not only will they help ensure everything runs much more smoothly and many stumbling blocks are avoided, they’ll also provide a much needed translation services for mere mortals who don’t yet know their interaxial from their interocular or their disparity field from their monoscopic depth cues.