Problems measuring camera SNR with darktable and Raw Therapee

Someone has to write the tip and pester the devs (or submit an issue or pull request). :stuck_out_tongue_winking_eye:

1 Like

I just pushed a fix to camconst.json

@nathanm darktable has ā€œmatch greensā€ that should do the same as Green Equilibration in RT if Iā€™m not mistaken.

image

I support the idea of a tooltip. In fact, now that am more aware of it: there are quite some more sliders that donā€™t have tooltips. Perhaps itā€™s even better to start a GitHub issue to cover all tooltips simultaneously (and put this one on top :wink: )

2 Likes

OK, but I donā€™t understand how to interpret the comment about AMaZE not being able to handle 2 greens, but VG4 does.

Am I better off (in general) using AMaZE with green equilibrium = 1, or using VNG4 where presumably one does not have to set that feature?

VNG4 is not good for high frequency details but handles different greens well out of the box. With the fix I mentioned above you should get better Amaze results for images from your camera because then green equil. is automatically applied to Amaze demosaiced images as well.

There are only a few cameras which need that setting. For example a lot of Olympus and some Canon cameras and also some other camera/lens combinationsā€¦

OK, I tried RT 5.8-2178 dev with green equilibrium = 1, and it fixes the problem!

This uses the settings from case 4 above - which include the lens corrections. The other cases seem good too.

Thanks so much to everybody who helped resolve this!

3 Likes

:partying_face:

1 Like

@nathanm This issue is a very good example, why we usually request a raw file which shows the issue. Would you have submitted the raw file with your first post, we and you would have saved four days of searching :wink:

Well, not entirely, as you could not know about the new preprocessing camera wb setting in RT dev buildsā€¦

Anyway, great that it is solved now :+1: