I suspect this is likely some gamut mapping voodoo being done as happens in a couple of other modules… but it appears it can be pretty significant… I was working with an image from this thread…
I had the image looking as such with a very very minimal stack of modules… out of gamut if you check wrt srgb but not the working profile but still I like something in that ballpark as the look for a sunflower…
I know DT developers do an amazing job on the math and color science but I would rather it didn’t sometimes… I might want to open the module to make a small correction or play with the saturation/gamut but if this is a common thing then you are going to find yourself correcting a new version of the image just by opening the module…
Maybe this is just one of those Red/Yellow flower images and it won’t be common.
I recall seeing this before but it seemed to be related to whether the guided filter was activated in the module when you applied it… this is or was not the case for this image…
I guess it will be just like any new DT module …with time the nuances will grow on me
thanks I will check again…did I upload the right xmp…
Only these modules…
Edit so it seems that it was due to a tweak in the green input channel that I had made at some point…
I tried on this image also a Canon image but I will try another raw type. and it could be out of gamut but I did not see anything turning on the color eq but when I went in and matched that setting from the sunflower image I saw a similar behaviour… what looks like some gamut magic…
I have not had time to experiment much with the new color eq module, but I’m excited about it :).
I know in the past I’ve seen a case where the Open-CL and non Open-CL code yielded different results (unintentionally), so I thought maybe that might be what is going on here. But for me it didn’t seem to cause any noticeable differences.
I’ve processed that image of mine over again a few times from scratch since the time since I took it and eventually the one here from scratch using Sigmoid as the tone mapper after it was available, so it’s a different starting point for comparison than what Todd shared above, but I’m sharing the color EQ before and after below just in case someone notices something I don’t.
Thanks for sharing…I think its just gamut mapping kicking in… so it will just happen in certain situations…or maybe my display profile is a bit of a bad actor too. I recently ran a calibration and I was experimenting with BPC settings… in any case you can see it pretty clearly when it happens…
Here is another image…its a playraw but I can’t remember which one… I will hunt the link
This is what it looks like with only exposure added…
I think the tweak induced very slightly negative blue values and so it’s being corrected… There is a very slight correction without that change in the channel mixer and the blue values are positive…it would just be harder to show but it can be seen toggling it on and off and with some slight changes in the picker values…
