An experimental set of primaries aiming for the ideal oxymoron of a camera-agnostic camera native space.
Each primary is chosen either directly from or fitted to a dataset of virtual primaries extrapolated from known camera matrices. This is very much brute-force.
The source files are of the form:
dignat-DATASET-FIT_METHOD.py
Old don't use.
DigNat to XYZ D65:
0.95862252 -0.05840237 0.05023578
0.38338622 0.89185891 -0.27524513
0.04954812 -0.46386142 1.50337105
XYZ D65 to DigNat:
1.02266986 0.05437236 -0.02421817
-0.49738313 1.21281632 0.2386691
-0.1871715 0.37241947 0.73961072