Technicolor CineStyle vs. Picture Styles
- Thread starter jmlau
- Start date
FilmBrute
Member
My understanding is that any profile that captures a high DR will have to make sacrifices. The more stops of DR you try to fit into an IRE range, the worse it is for color b/c you are basically spreading out the bit depth over a wider dynamic range. By looking at a waveform you can get a feeling for how this relationship works.
If you need to get the skin colors right, shoot Flaat10 or strait up Portait. Both will make better use of the 0-255 values for that particular application.
Samuel H
Major Contributor
The problem with CineStyle is not that it spreads itself across a range that is too wide for the codec. It's that, at the same time, it leaves a huge chunk of the codec's color space basically untouched: it doesn't go below 16, and it reserves everything above 200 for just a single stop of light, and it is so desaturated that it leaves most of the UV channels untouched too. When you get that image and spread it across a more reasonable range of colors, every IQ issue gets amplified.
And also: Portrait with contrast=-1 has a slight S-curve, but it can be useful; Portrait with contrast=-4 is wiggly and should be avoided; flaat has the same colors, and a much nicer light-response function to work with in post:
Portrait -4
Portrait -1
Flaat_10:
And also: Portrait with contrast=-1 has a slight S-curve, but it can be useful; Portrait with contrast=-4 is wiggly and should be avoided; flaat has the same colors, and a much nicer light-response function to work with in post:
Portrait -4
Portrait -1
Flaat_10: