Stereo3D 'Beauty Shot' tests

Dear Mr. Bata,

I just wanted to thank you for posting these beauty shot tests. I think i've learned a lot by watching them.

In particular, the longer lenses, as expected, flattened out the model so much that the only depth, was between her and the background (I assume she was shot in front of a greenscreen, and composited?).

The sense of "being there" really came alive, at 1 meter from the subject, when she feels fully dimensional. The model is fine-featured, so her face still looked great with that much roundness (I wonder how Penelope Cruz would look in the same situation?). However, I noticed that as your IA got larger, there was a good deal more cross-talk. I wonder how much of that was due to youtube's player, or the fact that I was watching it in anaglyphic? I'll have to try the sbs play in my 3D monitor, see how that looks. I have the feeling I'll be returning to these videos repeatedly, for further study.

Again, thank you for posting this. If you came to some conclusions from these tests, I'd love to hear them.

best,

matt faw

Hi Karel,

Thanks for sharing the subway video. For some reason, on YouTube, it only displays the beginning of your text. When I hit the "show more", the panel opens up, but only reveals a blank page. So I can't tell what the various iterations of the shot represent. Would you be willing to copy/paste that text here, so I (and others) can get the point you're trying to get across? I very much appreciate when 3D pros like yourself share your experience; I look forward to seeing the video within context.

As for my own 3D experience, I'm a strange mixture of just-getting-started and deeply-committed. That's because I've started shooting my first 3D feature, even though the technology is still pretty new to me. It's a documentary on the science of consciousness, and I had to take advantage of a biannual international conference in April to start shooting interviews. The interviews were all shot in front of greenscreen, so I have a good deal of leeway, in terms of creating my stereo final shots. Now, in parallel, I'm trying to build a story out of the interviews, while simultaneously shooting as many stereo tests, reading and viewing as much 3D material, and learning as many post tricks as possible, so that when I do actually get a cut, I'll hopefully be up to fleshing the movie out, without giving any short shrift to the stereo, which I am deeply committed to, as a viewer, and as a filmmaker. Probably not the most logical path into S3D, but it is very exciting!

Thanks again for making your experiments and experiences public, so the rest of us can learn from your work. I look forward to seeing more context on the subway video, so I can learn about these errors, that I am very likely making, myself.

Excellent! Thanks, Karel, for the youtube trick. Nice illustration. I checked the video out first, in anaglyph, on YT, but I'll download it in 1080 to check out on my passive monitor, as well. Like your beauty shot tests earlier, this one is worth coming back to, several times.

I love the depth in the shot. I've done a bit of shooting in the subways, here in L.A., because it's such a great subject, but I'm working with fixed lens cameras (GoPro and NX3D1), and the IA is pretty close in both, so the depth isn't nearly so impressive. Would you say what your lens length and IA is?

Re: greenscreen, I am convinced that it was the right way to shoot these interviews. Not only does it provide flexibility in post, but it also allows me to insert my interviewees into the visual metaphors, so they're not just disembodied voices and talking heads.

As for matching, I did a very quick and rough short film, as preparation for the feature. That included a lot of greenscreen under very flat lighting (outdoors in shade), which integrated very well into indoors 3D Studio Max environments. I actually found that the 3D-ness helped sell the composite, because the actor seemed more "there" instead of "cut-out". Maybe it also helped that I wasn't trying for too much naturalism. My (2 min.) short:
https://vimeo.com/39072542 side-by-side
https://vimeo.com/39068925 anaglyph

With the feature, I want to integrate the greenscreen footage a lot more with plates of exteriors, because real-world texture is so much more interesting in stereo, than most of what virtual environments will provide. So I'm trying to explore PFTrack and other possible post solutions that will allow me to composite onto moving plates, have objects in the foreground occlude my composited subject, etc. Also, I'm learning a lot more about pulling keys in After Effects, so the final should be a lot cleaner than some of the rough keys I pulled for the short. But, as you say, it's still a tricky thing to match lighting, especially from my interiors to exterior plates. Thankfully, the very subject of the documentary (consciousness) allows me some room to play, without as much need for perfectly naturalistic imagery. That's actually a pretty good place for me to learn; make the 3D as interesting as possible, but not be too worried that I won't be able to pull off naturalism.

Thanks again for the videos, and for the conversation!

best,

matt faw