But as I understand it (and I don’t claim to really understand film scanning) ADX is not a simple RGB primaries based encoding. Certainly the ADX to ACES conversion is more involved than a linearisation and a 3x3 matrix.
So that matrix would not be a direct ADX to XYZ matrix. It would be the matrix from the Relative Exposure values (created by going through lines 56 to 85 of the CTL) to XYZ.
Thanks Nick! I was aware that it was super straight forward. Just wanted to test to see what results I would get. I was advised by a few knowledgable people that the result were surprisingly good from a “normal scan”.