Applying 1D .cube LUT to image

Software G'MIC
,
1

I have been unable to apply a 10-bit 1D .cube LUT file to an image using G’MIC. I am prevented from uploading the LUT file, but suffice to say it is a properly formatted 1D LUT that has been verified to work in Photoshop. I am able to load and view the LUT just fine, but when I actually go to apply the LUT to a 32-bit floating point TIFF image, I get the following error:

gmic lut.cube,0 div[0] 255 test.tif +map_clut[1] [0]

[gmic]./ Start G'MIC interpreter (v.3.5.3).
[gmic]./ Input CLUT from file 'lut.cube' (1 image 1024x1x1x3).
[gmic]./ Divide image [0] by 255.
[gmic]./ Input all frames of TIFF file 'test.tif' at position 1 (1 image 1000x1000x1x4).
[gmic]./ Map color LUT [0] on image [1].
[gmic] *** Error in ./ *** Command 'map_clut': Specified CLUT [0] has invalid dimensions (1024,1,1,3).

The odd thing that is confusing me is that in the documentation for “map_clut” an example is given that inverts the colors in an image, and this command does work as expected:

gmic uniform_distribution {2^6},3 mirror[-1] x test.tif +map_clut[1] [0]

[gmic]./ Start G'MIC interpreter (v.3.5.3).
[gmic]./ Input set of 64 uniformly distributed 3-d points in [0,1]^3.
[gmic]./ Mirror image [0] along the 'x'-axis.
[gmic]./ Input all frames of TIFF file 'test.tif' at position 1 (1 image 1000x1000x1x4).
[gmic]./ Map color LUT [0] on image [1].

[gmic]./ Display images [0,1,2] = '[empty], test.tif, test_c1.tif'.
[0] = '[empty]':
  size = (64,1,1,3) [768 b of float32].
  data = (0,0.333333,0.666667,1,
[...]

What am I doing wrong?

2

Nothing wrong.
The fact is that map_clut assumes the CLUT is a 3D CLUT.
I’ve never encounter a 1D CLUT as a .cube file to be honest.
If you have one that you would share with me, it would help me to make it better supported by command map_clut.

3

It’s actually not my own LUT, it can be found here as a part of a project for accurately inverting color negative film.

I suppose my confusion arose from the fact that the documentation for “input_cube” lists an argument “_convert_1d_cluts_to_3d={ 0:No | 1:Yes }”

4

(post deleted by author)