Timeline for Output tonemapping to PNG produces gray even for superwhites
Current License: CC BY-SA 4.0
9 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Aug 6, 2020 at 14:28 | comment | added | GaryO | Thanks @susu -- I wasn't sure that "Filmic Blender" was the same as Filmic built into recent Blender, especially since the page you refer to only mentions the "Filmic Log" transform and its looks, and not "Filmic". But the chart in the false-color section is really useful. It gives a sense of the overall curve shape, at least for grayscale values. Also it seems that the PNGs that Blender produces with Filmic do have the sRGB ICC tag which is good. | |
Aug 5, 2020 at 17:48 | comment | added | susu | all the info here: github.com/sobotka/filmic-blender | |
Aug 5, 2020 at 15:58 | vote | accept | GaryO | ||
Aug 5, 2020 at 15:58 | comment | added | GaryO | Thanks -- I understand color science pretty well, but this is a useful post for people. The key point in your post, for me, is that the Filmic transform assumes scene-linear 16.19 is the reference white point. Do you have a reference for that? I want my final result video to be all white (just barely), so I need to know what values to create in the Cycles render. I assumed it would be 1.0 but clearly that's wrong, at least for Filmic. Is that transform (maybe 3d LUT?) documented anywhere? It may be best for me to simply render EXRs and do all my tone mapping in post. | |
Aug 5, 2020 at 9:22 | comment | added | jachym michal | I really enjoy how you explain things. I was trying to understand these things for a while now, and with your answer it all finally clicked together :). | |
Aug 5, 2020 at 6:15 | history | edited | susu | CC BY-SA 4.0 |
added 44 characters in body
|
Aug 5, 2020 at 6:09 | history | edited | susu | CC BY-SA 4.0 |
added 818 characters in body
|
Aug 5, 2020 at 5:33 | history | edited | susu | CC BY-SA 4.0 |
added 172 characters in body
|
Aug 5, 2020 at 2:58 | history | answered | susu | CC BY-SA 4.0 |