Light23 wrote:High dynamic range IS awesome.
By giving each projector a different gamma setting I wonder if that would make my 2d monitor HDR'ish since it uses 2 projectors?
http://www.firstshowing.net/2015/dolby- ... tion-demo/
I know, I know, the Sony is probably doing 8 bits.
But HDR can work with 8 bits. The limited range of traditional video, 16-235, or full range 0-255.
[Another upside to having such a great contrast ratio is it increases the perceived resolution.]
A few years ago Cnet's Geoffrey Morrison did a TV face-off with trained TV reviewers and untrained participants with Pioneer's Kuro plasma (768p) against several 1080p LCDs and plasmas. Not one person noticed the Kuro wasn't 1080p. In fact, most lauded it for its detail. Why? Its contrast ratio was so much better than on the other TVs that it appeared to have better resolution. The difference between light and dark is resolution. If that difference is more pronounced, as it is on high-contrast ratio displays, they will have more apparent resolution.
Haha, that's funny, I was considering buying two of these pico projectors to do just that, run them in parallel to do pseudo HDR.
The problem I ran into is that there is no lens, hence no lens shift, so aligning the images will be impossible unless you're willing to offset the top and bottom projected image in software.
Which would work for letterbox movies actually, you run one MPC-HC window at 3840x1080 and write a shader that displays the low dynamic range on the left and the higher dynamic range on the right, with the appropriate tone mapping obviously (and some falloff / blending between the two so there isn't a sharp discontinuity) and of course the appropriate vertical "software lens shift" to each of the left/right subframes independently. Writing such a shader would be pretty fun and pretty easy (I wrote one to correct chromatic aberration from anamorphic lenses and shared it on ShaderToy a while back).
The other problem for this is the fact that the "high dynamic range" subframe would be still pretty dim, so that one you would have to seriously boost the lumens of at least one of your projectors (like, by a factor of ten or more). Of course if you keep the projected image size very low as you appear to be doing with your home made screen setup, I think you can definitely do it.
Also, don't forget, 8-bit range for the low end and 8-bit range of DR for the high end results in a 9-bit effective DR display. You would obviously concentrate the lower 8-bit to handle, say, from 0-100 nits in the HDR signal, and the higher end to handle 100-1000. You could also just clip some of the high end DR to reduce the banding further, for instance let's say your original HDR movie rip (in 10-bit, HDR10 format) indeed goes from 0-1000 nits in range. So let's say you clip the top 50%, so your range of valid values becomes 0-500 nits. Then suddenly, you can do something like 0-50 nits is the low dynamic range projector's duty, then 50-500 is the job of the high DR's projector duty. I believe HDR done this way (or something like it), using two 8-bit projectors resulting in 9 effective bits of DR total, could do 0-500 nits without banding artifacts.
Agreed with everything else you wrote. Contrast is king for detail perception, in fact it's why HDR is so important, it increases the apparent resolution greatly. It's all due in large part to the fact that humans are far more sensitive to changes in luma than chroma, and that's why we have 120 million rods (luminance detectors) in our eyes and only 17 million cones (color detectors). Not only that but the rods individually are far more sensitive than each cone, to minute changes. This is the result of millions if not billions of years of evolution to max out our survivability in various light conditions to detect shapes of predators or prey, etc.
People who disdain HDR's importance betray their lack of understanding of biology and I assume probably don't even believe in evolution. Or science in general, most likely. I've had this same argument countless times over at AVS forum and I can almost always tell when someone's going to be against HDR by how they respond when I bring up the biological / evolutionary reasons behind the fact that we have 17 times more rods than cones in our eyes. This is not by chance. It's because we are well adapted to living and surviving on a planet where during the day the sun shines down billions of nits while at night it could be a handful. The ability to handle all this glorious dynamic range is why we are here, at all. Without HDR vision we would be toast.
HDR TVs simply take advantage of our HDR vision. Why hobble our vision system to 0-100 nits pathetic range in rec 709? It's absurd that anyone would even argue against it. Watching an SDR TV is the audio equivalent of watching a movie with the subwoofer turned off through tiny satellite speakers. (literally, both scenarios deal with the effects of massively crushed dynamic range)
Let me know if you need help writing the MPC-HC shader that does the image offsets to align the images together. Of course this is for movies only. If you want an HDR gaming TV you will need to write custom code, perhaps through raspberry PI which is dirt cheap and has HDMI inputs and outputs. You would still lose a few lines of addressable vertical resolution but it could be done without scaling the input (just an offset in UV coordinates at pixel resolution increments = no scaling)
