Comparison to negfix8?

[tomkyle]

There is already a positive converter: aZ99wro's negfix8 (Chris Hunt's GitHub version) which is mentioned quite often. Its results seem darker on my images somehow, maybe it is using another inversion method, or caused by missing ICC profiles.

I'll have to understand and retrace what exactly happens there. It may turn out that positive script is surfluous, but I see possibilities for output quality improvements as well.

Marginally, negfix8 seems to be not maintained any more, and I am not sure about the code quality patterns it is following.