As I read stereo filmmaker’s blogs and read lists and group postings, the comments center around: What are the rules? What is the right camera lens spacing? How do you avoid breaking the stereo wind0w. Too much parallax? What is the right camera toe-in? If you just have the right formula, then perfect 3D can be accomplished. Everyone wants to know the right thing to do.
If something doesn’t fit that mindset, it is quickly dismissed. For example, it is impossible to simulate accommodation (the eye muscles focusing on objects at different distances) when all of the visual content is on a single plane (the screen in a 3D theater, or the plane of the printed surface on a multi perspective lenticular. Well, we seem to be able to see 3D just fine anyway–so it must not be important.
Common sense tells me that is wrong.
Since Whetstone’s observations in the 1800’s scientists have fixated on the notion that two eyes and binocular disparity is what seeing depth is all about. Common sense tells me that it is the most important component, but only because it facilitates the supporting capabilities to fully define seeing depth – which, in the overall scheme of things is about 70% of the process in my humble opinion.
Binocular disparity (a slightly different image seen from each eye) makes us aware that depth and space exist. Without that, the other supporting depth cues are considerably weaker. Just ask anyone who acquired stereopsis vision later in life. They will tell you that things like motion parallax did not give them a sense of depth – whereas someone having stereopsis vision all their life will see depth with one eye looking out the window of a moving car.
So it is with accommodation. I didn’t really “get” the importance of accommodation until one day I was walking directly under several electrical powerlines overhead. I looked up with the power lines going across my field of view from left to right where they simply appeared as horizontal lines overhead. I could determine that some power lines were higher (farther away) than others – but NOT because of binocular disparity. In this instance, there is little, if any, binocular disparity since they are horizontal lines with little perceived texture. Each eye is seeing a horizontal line. Indeed, if I draw a left eye set of horizontal lines and a right eye set of horizontal lines, the two images (with horizontal displacement) have NO depth when viewed with a stereoscope or similar viewing device.
It wasn’t motion parallax that made it possible to see depth. It was accommodation that clearly made it possible for me to see the various heights of the different power lines. The depth cue to me was equally powerful to that of motion parallax. And accommodation solved the problem of being able to see depth where binocular disparity provided no clues with regards to the height of the power lines.
So, it is no great leap to realize that where binocular disparity becomes weak as a sense of depth, other components of vision take over to fill in the gap. Seeing 3D is a system whereby the brain uses multiple senses to perceive and interpret a scene. When one or more of the senses conflict, and that conflict isn’t suppressed, it stands to reason that people might feel some level of discomfort no different from motion sickness or dizzyness. This HAPPENS when senses conflict and suppression of a conflicting sense is not suppressed.
Over time, most people – probably all people if so motivated – can overcome and suppress conflicting sensory input. And that process is adaptable. Take people at sea… they get their sea legs and then get their land legs. For some, it happens faster than for others.
Lets jump back to accommodation for a second. When we remove accommodation from our stereopsis vision, then it stands to reason that it isn’t a good idea to present imagery whereby accommodation would play a major role in seeing depth. For a 3D motion picture – that means AVOIDING large amounts of depth in a scene – or at least large amounts where there isn’t a specific area of attention.
Wait, that means it isn’t a good idea for things to stick way out of a 3D screen!
Yes, that’s right. Especially things with a horizontal orientation. And whenever there are scenes with a lot of depth then as many other depth cues should be implemented like motion parallax to help compensate for the supporting sense (accommodation) that is missing.
Once we start to understand that seeing with depth is a system of many perceptions, then we start to realize how to be better stereographers and 3D image artists.