Comment Re:Is it "too real"? (Score 1) 607
I think the one key thing that you do not expect is how aperture effects DOF and exposure. So you need a certain amount of light to enter the back of the camera. I think the way you think is very technical, think about it like you need enough to beat the S/N for a decent photo. Can you buy that? Well when you have the shutter open for a short period of time you are letting less light in. So to compensate you can use more light. Well that's hard because you already are bathing the set with a lot of light. But let's assume you could get more light. There is also the issue of darker materials reflect less light than lighter ones and you need to make sure that the white and other such light colored objects in the scene do not wash-out completely while you have still detail in the darker areas. In fact with a lot of light you can actually see the lens elements themselves in the image you capture. So it's hard, you need just the right levels. Images that are overexposed look dull and lifeless as well as washed-out. The more overexposed that they are, the harder to correct, and if over-saturated in areas, they really cannot be fixed later. Photographers call this a blown highlight. Again in your way of thinking, all those values are pegged at 16K or whatever even though there was a whole range of them beyond. Incidentally CCD response also happens to not be linear. But there are more fundamental problems.
See the another way you could let more light into the back of the camera would be to use a larger aperture. This means that the shutter opens to a larger diameter. So first the problem is one that you can throw money at to solve to a certain extent at least. For practical reasons lens that have a larger aperture have other deficiencies. One very common one is that they have a much more limited zoom. To some extent better made lens can solve that, but there are some limits. Like you want a big CCD back there right, so you can average to beat some of the noise. Oh and you say you want 3CCD (which further reduces light to each CCD BTW). Ad yes you do want to be able to pull the shot (zoom), so soon you could make an awesome camera, but it would rival the Hubble in terms of size. (I exaggerate a bit, but you get the idea.) Still it's hard, they tend to be less quality lens for the same price as well, like 5 elements instead of 7, more aberration, etc. But beyond those practical aspects, there is the most fundamental problem, and that is depth of field. When you have a small aperture, it's like a pin-hole camera. Things close by and those things relatively nearby will seem in focus. But use a wide aperture and only those things relatively close to the focal point of your lens will be in focus. It's actually sort of neat in say single subject portrait photography since anything makes an interesting sort of random looking background and you only need to focus on a single face, but it's not good for most films where you have many subjects you want in focus as well as having the background maybe out of focus but with enough detail to still make out more or less what is there instead of plasma looking globs of color. That's an optics things, really can't be solved computationally or anything like that. Would need radical new lens technology, one that splits light most likely, getting you back to your original problem. Yes there are high speed videos, lots of them, but next time you watch them pay attention if you can see anything other than the popping balloon or what not in focus and how saturated the or overblown just generally balanced well the images look. Usually it's some scientific context and you can make-out what process is occurring, but it does not look good in terms of cinematography, like the shadows look dead or really noisy.