Difference between Open-Source and Commerical Transforms

[RikshaDriver]

What is the difference between these open source Protune Transforms and the commercial GP-Tune Transform?

For reference and clarity, the open-source transforms are based off the NATIVE to sRGB color matrix data that has been revealed in an old iteration of the now-defunct GoPro Studio software. Whilst this is specific to the HERO5 Black, the color Matrix has been defined by GoPro themselves and all other open-source NATIVE matrices have been calculated using this data.

The GP-Tune Transform uses metadata from each compatible GoPro camera to construct a distinct NATIVE color correction matrix unique to that camera. In my opinion, this calculation provides far more accurate results. If we use the HERO5 Black as an example with a NATIVE to Rec.709 (sRGB) color transform, the two corrected images have noticeable differences.

protune-transforms (HERO5 Black)	GP-Tune Transform (HERO5 Black)

The GP-Tune Transform also supports the WIDE white balancing parameter which is default on the GP-Log profile and also available on GoPro Labs firmware.

Where white-balance is set to non-NATIVE (i.e. Auto, 5500K or some other WB), the recorded clips are limited to a Rec.709 Color Gamut. For this scenario, the open-source and commercial transforms are largely the same with some exceptions.

As of Release 2.0, the GP-Tune Transform includes additional parameters such as the ability to define White Balancing WRGB parameters, enable Gamut Mapping and Highlight roll-off. Two new Custom ACES Input Transforms are also included with user-defined values.

Both transforms will be maintained separately. The protune-transforms will rely primarily on the data obtained from GoPro Studio whereas the GP-Tune Transform will use the independent color matrix calculations and features.

For the Free version of DaVinci Resolve or if using other NLE/VFX such as Assimilate Scratch, Vegas Pro etc. which support the OpenFX standard, refer to the GP-Tune OFX Plugin instead.