Because there is nothing to adjust. The curve is designed to produce a smooth transition between black, grey and white levels, and that’s what it does. But setting black = 0% was nicer for the interpolation, as it produced less “tension” on the curve. Black > 0% is mathematically more challenging, because you put more constrains on the interpolation. It’s kind of like trying to fit a thick plastic sheet in a box, you need to squeeze and force it. Just try black = 0.001% and see how it helps getting rid of the undershoot.
This isn’t about me being able to use filmic, but about the same settings in 2 dt versions showing very different results. @anon41087856 explained the science part of that and I do agree that master needs a somewhat different approach and getting used to.
And while replying to you Aurelien also answered the visualization issue. I do get the rationale behind that explanation. I’ll go and experiment with some different settings and see if making one of those an auto applied default makes sense.
You mean loading the same XMP in 2 different versions does not produce the same result ? That shouldn’t be the case. Only the default params have changed.
The 3.2.1 xmp was loaded via lighttable->history stack-> load … the 3.3.0 version was done manually to make sure I got the exact same numbers (see the video for the results).
BTW:
Do you mean setting target black luminance to 0.010 (!) in the filmic display tab? Values lower then 0.010 aren’t excepted. This does solve (all but) my visualization" issue… Thanks.