Originally Posted by LTD_666
Could you send me by PM the technical information to use the Arne's software and whatever else that you consider relevant in relation with this software?
Thanks in advance
As explained in the post above, I haven't heard back from Arve yet so I'm flying blind.
I'll release more definitive info when I'm sure of what each parameter is doing and how to use it.
For now, here is what I *think*, but I could be completely wrong, I've had very little time to figure this out. Please take the below with a bucket of salt:
Max Brightness: target in nits, lower = brighter (steeper curve) but less saturation, higher = dimmer but more saturation, I suggest as before between 600-1100nits
Soft clip start: where to start the roll-off (in nits). It has to be lower than the above, I'd say at least a couple hundred nits. I use 800 for a 1000-1100 target, but 400 for a 600 target.
Hard clip brightness: where you want to clip (I initially thought 4000nits but I get some clipping on 4000nits so I'm using 10000nits now). This is one of the parameters I'm not sure about. With my own curves, I could adjust the roll-off point manually to clip where I want to. I could still do this, but I need to understand better how the roll-off parameters work to achieve that without any manual adjustment.
Clip end slope (0-1): 0 = straight from soft clip start to hard clip brightness, 1 = most curved shape between these two points. Any value between 0 and 1 are possible to adjust the roll-off shape of the curve, I'm using 0.3 for now but this should be adjusted depending on the other parameters.
Soft clip method: 0=soft, 1=hard, I'm using 0.
Soft clip gamma (0-1): not sure about that one, maybe Arve knows. I use 1 for now, I haven't had the time to figure it out. It might be the parameter I need to adjust to get the roll-off I want, i.e. using a hard clip brightness of 4000 yet resolving up to 4000nits without having a silly shaped curve.
All the curves I uploaded are generated by the software, I didn't edit them at all, which is why it's still a work in progress. The only difference is that they are translated from whatever resolution they use internally (256 points I think) to the 11 points of the custom gamma files when the JVC Autocal software exports them as text files. You can edit them with the JVC Autocal gamma feature, but then you lose some of the original resolution, as I assume it will recalculate the target in between the point you change and the point before it and after it.
We need more info from Arve (or whoever designed this wonderful piece of software) ideally.
Have fun with the above, and please let us know if you can confirm this or have better ideas about what each of these does.
Still, I'm quite pleased with the experimental results. Not ideal yet, but I think already better than what we had as we're not approximating the targets with the screen offset trick on a workflow designed for flat panels, we're using actual targets. That makes a big difference, plus of course it's incredibly faster.