Santa Claus and the 3-D Disconnect

by Aris Silzars

3-D is coming! 3-D is coming! We're finally going to have all of our movies in 3-D! At least so goes the self-serving promotional excitement among some display-industry forecasters as well as entertainment-industry gurus. But didn't we go through all of this once before about 50 years ago? What's different this time?

Is the quality of the images so much better? Is polarization technology now so superior? Yes, the projected-image quality is better, and no, polarization technology has not changed much at all. This led to my prediction, which I offered in a column in Information Display in June 2006 that "Truly realistic 3-D such as needed to create a believable virtual-reality experience is at least 25 years away for single viewers and perhaps 50 years away for multiple viewers – and that is at the unlimited cost level."

However, there is one very interesting change that does create a new opportunity for stereoscopic 3-D. The advent of computer-generated images and entire movies that no longer resemble the Disney-like cartoon movies of 40 and 50 years ago is a major driving force for exploring 3-D effects. As computer-generated images become more and more realistic, adding a 3-D effect perhaps can enhance our viewing experiences even further.

But will truly realistic 3-D movies become the standard fare for viewing, rather than a novelty feature that shows up in only a few "I've suspended my reality" movie experiences? Well, that's where we run into trouble, much of which is due to the complexity of the human visual system.

At the most recent meeting of the Pacific Northwest Chapter of SID, Brian Schowengerdt from the University of Washington presented on the research that he and his group are conducting into what it will take to produce a 3-D image that begins to replicate how our visual system interacts with the world around us. Unless we understand and properly imitate this image-acquisition process, there are problems created that go way beyond a less-than-fully-realistic visual experience.

For good reasons – some dating back to our caveman days – our visual systems have developed to tell us a lot more than just what our immediate environment looks like. For example, our visual system works with two of our other senses – touch and hearing – to tell us what is moving and whether we are part of that motion or not. In a complicated process, it also tells us what danger may be lurking close by or further away. This is accomplished not only by the stereoscopic effect of seeing with two eyes, but also by what is in focus and what is out of focus, and how that corresponds to what our eye muscles are doing to position our eyes at the correct angle to focus on a given object. Of course, our head position and any slight movements we make add to the data stream that our brains are continually processing. The problem is that if our sensory system senses a conflict, it goes into a conflict-resolution mode – typically by responding in ways that we don't find very enjoyable. If the conflict is slight and persistent, it may react mildly by giving us a headache. If the conflict is sudden, it reacts by making us nauseous.

I asked Brian why should we get nauseous if we have a visual disconnect? His explanation is that this was built into us during our early evolutionary days when signs of dizziness meant we had eaten something poisonous and, therefore, our bodies were set up to eliminate the poison. So, as far as our visual sensory system is concerned, a 3-D movie that only presents a stereoscopic view is not going to be acceptable for anything more than a novelty experience. It cannot "look right" because the focal planes are all at infinity and, as objects move in and out of the scene, our eyes get confused by the lack of focus changes and eye-directional accommodation errors. For most of us, the disconnect is not bad enough to make us nauseous, but a headache is certainly not out of the question, and the feeling that we were in a "spacey" and unrealistic environment will be a typically minimal reaction. I have also heard this called a "dollhouse" effect or a diorama-like effect.

If the conventional approaches to stereoscopic 3-D are unlikely to be successful, is anything on the horizon that is likely to make my seemingly pessimistic prediction come true? The first inklings of future success may be provided by the work of Schowengerdt's group at the University of Washington on a laser-based system that projects an image onto the retina. For each part of the image, the correct focal distance is generated, and the viewer sees the scene with the objects in their correct focal planes. (This work is described in more detail in a paper published in the Journal of the SID 14/2, 135–143 (2006)]. Therefore, we may finally be on the path to truly realistic 3-D that will be adequate at least for single viewers.

Realistic 3-D that can be enjoyed by a large audience would indeed be a wonderful Christmas present from Santa to the display community. However, it seems that we will have to set our sights at a more modest level and first get to where we can present realistic 3-D images to single viewers. With the clamor for evermore exciting video games, this may not be such a bad place to start. Gamers will appreciate the single-viewer experience – especially if it does not make them sick or give them headaches after many hours of intense use.

There are exciting opportunities in 3-D and the creation of realistic images. We just need to get beyond the notion that high-quality stereoscopy is all it takes to make this happen. That's actually a really good thing because it will allow us to take advantage of innovations that go well beyond what we conceived more than 50 years ago. Could we in the display community ask for any better Christmas present from Santa this year?

If you have thoughts on this topic or others, or if you would like to submit something that could embellish this space on a topic that relates to the business of displays, please contact me at silzars@attglobal.net or at www.arissilzars.com. •