How to Prevent Stereo Fatigue in 3D Features?

[Matt Faw]

Dear all,

I love watching 3D films, but I also find that the whole stereo effect starts to fade, about 30-45 min. into the film. At some point, I have to blink a lot, and remind myself to look for the 3D, in order for the effect to wake up in me again. Then, it starts to fade again.

Now, as I'm trying to plan the edit and depth budget for my own 3D feature, I am keenly aware of this fatigue effect, and want to find ways to pace the stereo, and periodically reinvigorate the audience.

Thankfully, since my documentary doesn't have to stick to any naturalistic rules of continuity, all possible techniques are on the table. I'd love to hear y'all's thoughts about ways of dealing with stereo fatigue, and how to keep the audience's eyes fresh.

[Pavel Houda]

You might be experiencing vergence/accomodation related fatigue. One way to minimize the problem is to keep the subject of attention at the screen depth. Larger screens help as well. There is a bit of work on viewing technologies that allow more correct convergence/focusing harmony, but there is no common solution to this issue yet that I know.

[David M. Cole]

I realize that you are probably talking about the theater experience - so - sorry for hijacking...

One of the biggest problems at the moment for home 3DTV is the eye fatigue that results from frame-compatible transmission (side-side or top-bottom). The anamorphic decimation of the image, combined with compression decimation, is a brutal combo. The result is that objects edges - where they occlude what is behind them - are blurred. Occlusions are critically important in depth perception. Any objects with parallax (negative to positive) that have blurred edges will cause occlusion vs parallax rivalry. The result is a nasty discomfort that sneaks up on you after 10 min or more (typically).

One reason this problem remains unchecked in the industry (besides the obvious advantages to cramming 3D down the existing 2D pipe) is that a LOT of executives never watch more that 5 min 3D demos. They have no idea what long-term viewing feels like.

The theatrical problem has been well discussed recently (Cameron et al). Even if the 3D is perfectly shot, 24 fps causes way too much motion blur.... and we're back to occlusion vs parallax rivalry.

[Matt Faw]

Thanks, Pavel and David, for your replies.

I posted this on Creative Cow's forum, as well, and the response I got there was that stereo fatigue just meant there wasn't enough negative parallax to keep it interesting. I'm not sure I completely agree with that comment, but I do think that negative parallax can be some of the pepper in the stew, which keeps it from getting bland. Poke them in the eye, once in a while.

That respondent also talked about an feature that's shooting now, in which they shoot all their big action scenes in stereo, but the rest of the film will be projected in 2D. I think there may be real merit to this idea, if indeed the 2D does "reset" the audience's eyes. The audience may actually walk out of the film being more impressed by the depth, than if the whole film had been in stereo, if they don't lose the 3D effect, a third of the way into the movie. In particular, what a shame it is to have the big final action scenes seem flat, not because of lack of craftsmanship going into the stereo, but because of a physiological fact about our eyes + brains.

I guess one of the big questions is: does (nearly) EVERYONE experience stereo fatigue like this? Everyone I've talked to about it does, but that's been a pretty small sample. If it is (near-) universal, then I think it suggests that we as filmmakers should really focus on what works in the big picture, which may include getting rid of depth occasionally, if it helps the overall impression.

[Matt Faw]

I also wonder if there aren't some non-intuitive ways of refreshing the audience's eyes?

Things like flashing lights, moments of intense binocular rivalry, black leader, alternating periods of deep negative vs. positive parallax? Just speculating, of course, but I don't think that we should rule out unconventional ways of keeping the audience clued into the stereo.

I'm hoping more people join in this thread, because I think this is an issue that effects all of us, who are interested in producing longer form works. We can lovingly craft the best stereo on the planet, but if the audience's eyes have already shut down to depth, then it's all for naught.

[Adam Stanislav]

Matt, I’m with Pavel. Keep the object of interest at the screen level most of the times.

In real life, we can look at any object and our eyes will converge as necessary. In 3D film (and video and perhaps even photography) if the object appears behind the screen or in front of it, the eyes are forced to converge at a distance that the object appears to be even though physically the object (or should I say the image of the object) is always on the screen. This is unnatural, as far as the physical action of the eyes is concerned, and causes fatigue. If, however, the object of interest (typically your actors but can be anything that determines where the action is) is at the screen level, the eyes converge at the screen level, just where the physical image is, which is natural and, therefore, causes no fatigue.

You can make an occasional exception but most of the time you want the eyes to converge on the screen. Also, when you do make that exception, never cut from one conversion distance to another, but make the change as smooth as possible. Again, in real life we usually do not jump from one distance to another, we change the distance gradually because our eyes have to move physically and that takes some time. Not a long time, but still it happens gradually. Cutting from one conversion distance to another without a smooth transition is a cardinal sin of 3D filmmaking.

Keep it natural.

[Leonard Levy]

Sounds like the responses here are confusing 2 different kinds of fatigue. The original post refers not to eyeball fatigue as in "this 3D hurts my eyes," but rather fatigue as in " this 3D is so boring I don't even notice how cool it is anymore."
The first just means paying attention to all the rules that stereographers always harp on.

Re: the second, In some ways it can be good since it means you've become immersed in the story. But I do understand the issue. I have no experience with features but as a viewer I notice that if the depth budget is varied with the content so that some scenes play with less depth, some with more, and some more or less in neg or pos space then you should be able to hold interest better.

My favorite film for this was Harold and Kumar's Christmas where Paul Taylor varied the depth in different kinds of scenes to fit the story. Of course it helps when your protagonists are getting high or dropping hallucinogens every now and then.

[Matt Faw]

Thank you, Adam and Leonard, for your responses.

My 9-5 job is editing for a production company, and so after hours of looking at a 2D screen, I go out for lunch and the real-world 3D jumps out at me. Everything seems in focus, my eyes feel like a twin-lens camera, and all the leaves in the trees pop out in vivid detail and depth. The effect is extremely powerful (and pleasurable) for about 10 minutes, and then fades, at least until I remember to appreciate it again. So, my experience is that stereo fatigue happens even in the non-cinematic world, as my eyes habituate to the depth. This is of course, independent of technique or abnormal convergence. As far as I can tell, it is the contrast from my previous 2D viewing that elicits such a powerful response from my lunchtime eyes.

Originally Posted by Leonard Levy

My favorite film for this was Harold and Kumar's Christmas where Paul Taylor varied the depth in different kinds of scenes to fit the story. Of course it helps when your protagonists are getting high or dropping hallucinogens every now and then.

Perfect! What I'm creating is a documentary on consciousness, so I have total room to go into hallucinogenic territory. I have no need, whatsoever, to follow a naturalistic depth script, so I can play, play, play. One experimental strategy in consciousness studies is to work with radical binocular rivalry (utterly different images to each eye, and/or one eye flashing, while the other is subliminal, etc.), so I can (with great carefulness) break even the most cardinal rules of stereo filmmaking, at least for short periods of time. Of course, I do want to make sure that there is method to my madness, and that my experiments keep the audience engaged, rather than alienated.

I'll check out H&K's Christmas on Blu-Ray and see what Taylor did.