Originally Posted by krichter1
Hey Claw have you read any confirmation post/info what 1005 actually did for the 8500?
I'm also really surprised that JVC hasn't released another FW update to fix GammaD unless they feel it's not their
The whole gamma D thing seems to be poorly set up. Given a screen size, screen gain, aperture, and buib settings, the maximum on screen output level is defined. With that defined, (IMHO), you should be able to set what output level corresponds to the 50% input level, the output level that tone making should start, and perhaps some control for how quickly it moves to full scale white.
For example, say you have the RS600 and you set high lamp, manual aperture such that you have 1500 lumens. With a 1.3 gain, 120" 16:9 screen. You should get 156 nits for peak brightness.
Now supose you decide to you want the 50% video input to correspond to 48 nits, instead of 100 Nits as specified in ST2084. Why would you want to do this? First because 48 nits peak brightness is the same level standard movies are graded for in the theater, and second, because it gives some room for the HDR effect (in this case 3:1, from 48 to 150 nits to tone map highlights).
So I envision controls where you would set the peak brightness in nits, followed by the 50% brightness (say from 40-150 nits), and a control which changes the shape of the curve from the 50% level to the 100% level.
Alternatively (with the same screen), one could select 100 Nits for the 50% input level and only have 50 nits to tone map from 100 to 150 nits.
Also alternatively, one could lower the manual iris setting such that only 1000 lumen are projected, resulting in only 100 Nits peak output (but with much higher on/off contrast). Then setting the 50% level to 70 Nits would effectively be equivalent to SDR (with inputs from 50%-100% mapped to 70 nits-100 nits).
With these controls, the user can trade off HDR/brightness for contrast and black level performance the JVCs are known for.
I'm sure there are some issues with what I am proposing, but I believe there should be a way to set this up better than the current method.