I'm still not buying it, you talk as though you can process over 60 images per second. So how about you run 60 different images per second past your eye and tell me what some of them are......
You don't seem to have completely read or grasped what I said. Our neural system in general doesn't "process" information in the same way as a basic video processor would. It's very good at perceiving CONTRASTS (which means differences in general, be it spatially or temporally). But the structures that can detect contrasts very fast can not do so accurately. Conversely, perceiving information accurately is pretty slow. With our visual system, we CAN perceive very fast motion, but we would be totally unable to describe it precisely. Our brain doesn't even "process" all images, it processes their changes.
Explaining that in a simplified way, that would be a bit like equivalent-time sampling. Trading accuracy for speed, or speed for accuracy. Obviously something is lost along the way, but it can still provide some useful information in some contexts.
Our auditory system works much with the same principles. The hair cells in our cochlea aren't able by themselves to process sound with as fine frequency/temporal changes detection as we actually can perceive. Higher-level neural relays allow to detect contrasts to drastically enhance the overall performance. To illustrate some of that, even people with "absolute hearing" cannot possibly detect a 1Hz difference in a pure sine wave presented alone. Now if you present two such sounds with a 1Hz difference only (or even less) at the same time, or just one after the other, but very close temporally, many people can perceive the difference. We don't "process" things individually.
Since the faster the differences (contrasts) are, and the less details of them we get, it's probably completely possible to design a video processor based on some kind of psycho-visual algorithms, so that it could generate a high frame rate without actually having to generate complete frames for each frame, a bit like psycho-acoustic compression, but for our visual system. I don't know all the MPEG algorithms for video well, so that may be what they do (or at least some of them). But I don't know of any graphics card that would do this on the fly for gaming purposes. It would probably avoid the need to generate full frames while still being able to provide improved motion perception for the users. Just a thought. That may already have been patented a while ago... I think some TV sets that have 100Hz/120Hz refresh rates (or even double that) already have such processors to improve motion perception. They don't all seem to work very well though, but that's worth mentioning.
Also consider what I said after that part in my previous post:
Although again the difference would be noticeable by many people (we are not all perceiving things quite the same way) when they are focusing their attention on the video itself, if they are actively playing a game, I ventured that their attention is on action and that they will probably NOT perceive a difference (apart again from pure latency, which is something else), at least not consciously, as conscious perception is not multi-tasking very well for most of us. That may still improve the speed of our reflex actions, but that would remain to be proven scientifically in a study in which we would test various frame rates but make the latency identical, which would not be trivial.
Home
Help
Search
Login
Register
