Page 2 of 2

Re: Stretch gets lost in 1.5.369

Posted: Sun Oct 27, 2019 4:30 am
by devonshire
Straight to the Color module, just taking defaults - and this is what comes up.
Doesn't come close to the more refined result that Wipe's [color] option was able to produce... on my monitor, it's practically flourescent.
06_Color_Initial.JPG
06_Color_Initial.JPG (367.03 KiB) Viewed 4556 times

Re: Stretch gets lost in 1.5.369

Posted: Sun Oct 27, 2019 4:39 am
by devonshire
I have Color module s-shots for Legacy and Hubble as well, but this probably illustrates the problem sufficiently.
Wipe [color] was able to produce a very reasonable starting point, but that just disappears, and the Color module...well, these are going to take a lot of tweaking. Why?

Thanks for your help with this1

- Bob

[Edit: FWIW, the colors of all examples look a lot more subdued on my Thinkpad's panel than on the Benq monitor I use for processing.]

Re: Stretch gets lost in 1.5.369

Posted: Sun Oct 27, 2019 6:21 am
by admin
Hi Bob,

Thanks for uploading those screenshots. Here too, I cannot see anything wrong or constituting aberrant behavior.
On my color-calibrated monitor, the result from Wipe is very, very far from a balanced or acceptable rendition, whereas the initial Color module result is starting to show important aspects correctly.
Stars in the Wipe module result are devoid of color, whereas in the Color module different color temperatures are visible, with blue stars notably present.
Areas strong in H-alpha emissions (such as Melotte 15 in the center) show strong, deep reds without channel clipping. Pushing blue a bit more, you should start seeing subtle shades of purple where the blue light from hot young O and B-class stars mixes in with weaker Hb and O-III emissions. In the result from Wipe, they are any one's guess what color Melotte 15 is, as it is entirely washed out. The muddy brown/red color cast in the Wipe module is entirely a coincidence rather than the result of any procedure resembling color correction.

Even if you do not prefer the Color Constancy rendering responsible for the color recovery of the center shock waves, this mode is instrumental in making sure you got your balance correct. You can switch to Legacy mode afterwards and reduce the Bright Saturation to mimic the washing out if you wish. If you are sure the result is too saturated as a whole, you can, of course, easily reduce the saturation with the Saturation slider. Please be aware that consumer-grade screens tend to greatly oversaturate colors to make them "pop". After color calibration with, for example, a cost-effective x-rite Colormunki Smile, you will find correct tones are much more subdued.

Lastly, if you wish to use the Wipe result as a starting point, you should be able to use the Legacy preset, set Saturation to 100%, Bright and Dark stauration to Full, and all Bias parameters to 1.0.

Re: Stretch gets lost in 1.5.369

Posted: Sun Oct 27, 2019 3:55 pm
by devonshire
Ivo,

Thanks, as always for the help!

For my part I rather liked, and was focusing on, the texture of the nebulosity in the wipe [color] rendition, vs. same in the Color module - but yes, the star colors are clearly not there.

Anyway... I've put a Colormunki on order and will take another run at this w/your advice after it's arrived and my monitor's calibrated.

Thanks again,

- Bob

Re: Stretch gets lost in 1.5.369

Posted: Tue Oct 29, 2019 12:13 am
by admin
Let us know how you go Bob. The default AutoDev stretch is indeed different to the Wipe stretch, so that will impact the difference in presentation as well.