LMTs Part 4: How do they work and how are they made? (continued...)

Tags: #<Tag:0x00007f34c95d5d40> #<Tag:0x00007f34c95d5c00>

Clearly I’m very late to this thread, but I had some time off today and thought it would be interesting to implement this as pure Nuke nodes for those who are interested in playing around with it.

I’ve chucked a Nuke version of @sdyer 's LMT.Academy.Analytic_3.ctl up here on github here:

My install of ctlrender on Mojave is currently broken, so if anyone could render out a version of SonyF35.StillLife.exr through Scott’s code it would be much appreciated. But I think this should match.

I can’t open your Nuke project, since even after I unpick the differences between your OCIO config and mine, I don’t have a Nuke 12 license, so get an error from the new shuffle node you are using.

But here are examples of the test image run through ctlrender with the sRGB and Rec. 709 ODTs respectively:


Rec. 709

1 Like

That’s super helpful, Thanks Nick.

Turns out I had completely skim read the code and didn’t implement Rotate_C_in_H, and just used another Rotate_H_in_H. That’s been corrected and updated on the repo. It now matches your examples cleanly.

I’ve also swapped the new Nuke 12.1 Shuffle for the old one. The OCIO config is the Nuke 12.1 bundled ACES 1.1 config.