16
[Max] General Discussion / Re: ACES Color Space
« on: 2023-10-17, 18:47:59 »
Not sure what your example is showing but I don't think it's what busseynova or I are referring to. If you use ACES (at 1.0, say, since that's the default), and render an EXR, load that into Photoshop and try underexposing it - you can't - whites go grey. Now if you do the same without using ACES (or any other kind of highlight compression) you can successfully underexpose the render in Photoshop and loads of data is retained. I understand (from chats with Tom/Corona guys) why this is happening, and it's not a bug, but it's still a workflow issue/hindrance. The "gold standard" would be to have ACES (or ACES-like) tone mapping but still retain the same (or similar) level of dynamic range in the rendered result.
The best example to test this on is a simple scene using corona sun and sky, with the sky slightly overexposed in the base render by say 1 stop. If your EXR has proper dynamic range then you can very easily recover that in post and expose down to reveal the sun showing clearly and the sky becomes blue etc. as you would expect from for example a CR2/DNG RAW photo. When using ACES this ability is removed, as you'll just darken the base render without revealing any more data, and that's the issue.
The best example to test this on is a simple scene using corona sun and sky, with the sky slightly overexposed in the base render by say 1 stop. If your EXR has proper dynamic range then you can very easily recover that in post and expose down to reveal the sun showing clearly and the sky becomes blue etc. as you would expect from for example a CR2/DNG RAW photo. When using ACES this ability is removed, as you'll just darken the base render without revealing any more data, and that's the issue.