The Filmic ocio by @Troy_James_Sobotka has a false color view transform which seemed pretty useful. What would be needed to use this in ACES?
I believe Filmic is in sRGB/Rec709 linear working space, and the false color LMT uses FilmicLogEncoding as it processing space. What would be the correct way to use that LUT in an ACES config?
One thing I figured I should do is change the matrix which gives the luma, and switch it use luminance weights for ACEScg based on this post
- !<Look>
name: False_Colour
process_space: Filmic Log Encoding
transform: !<GroupTransform>
children:
- !<MatrixTransform> {matrix: [0.272229, 0.674082, 0.0536895, 0, 0.272229, 0.674082, 0.0536895, 0, 0.272229, 0.674082, 0.0536895, 0, 0, 0, 0, 1]}
- !<FileTransform> {src: Filmic_False_Colour.spi3d, interpolation: best}
The part I’m less sure about is the LMT process space:
- !<ColorSpace>
name: Filmic Log Encoding
family:
equalitygroup:
bitdepth: 32f
description: |
Log based filmic shaper with 16.5 stops of latitude, and 25 stops of dynamic range.
isdata: false
allocation: lg2
allocationvars: [-12.473931188, 12.526068812]
from_reference: !<GroupTransform>
children:
- !<AllocationTransform> {allocation: lg2, vars: [-12.473931188, 12.526068812]}
- !<FileTransform> {src: desat65cube.spi3d, interpolation: best}
- !<AllocationTransform> {allocation: uniform, vars: [0, 0.66]}
to_reference: !<AllocationTransform> {allocation: lg2, vars: [-12.473931188, 4.026068812], direction: inverse}
As it happens, I have just recently been experimenting with a false colour gizmo, with customisable colours and ranges (in stops). The current version uses the ordinary Nuke Colorspace node, which doesn’t include ACEScg. But I was thinking about changing the way it works so it just has a dropdown of common encoding combinations. I am happy to share it when it is done, if that would be useful to you.
That’s serendipity. Yes please!
hey @nick! Would love to have a look at your false colour tool, totally fine if its not considered ready!
Would you be fine to post it or send it over?
Wow, was it two years ago already.
I am happy to share the gizmo, with the caveat that it comes with no guarantees, as it is just something I was playing with for my own purposes, just to show how I thought a false colour overlay should work. This was because the ones built into monitors tend to be based on IRE, not stops, and the same IRE means different things with different encodings. I always thought monitors should detect the encoding of the signal from metadata, and set the false colour ranges accordingly.
Please email nick@antlerpost.com if you want the gizmo. I am happy to give it out at no charge, but I would like to know who is using it.
