I am running darktable-3.5.0+1170~gb428b0d7f-win64.exe. I’m guessing it’s available in a Unix build.
Next on my agenda …
7 Likes
@hannoschwalm Hanno, though, as an RT dev and user, I really appreciate your efforts to port the goodies of RT to dt 
If you need support to port capture sharpening to dt, just ping me 
10 Likes
I will ping (Still fixing a minor opencl rcd issue …) AND more important, i don’t yet know what might be the “best dt way” to make the contrast mask available in a general and UI-friendly way.
1 Like
Thanks a lot to all of you for the interesting and very satisfactory discussion on my question.
Vittorio
Rawtherapee seems sharper to me to.
Here is a comparison, using the raw from this thread:
20200718_0062-dt_01.CR2.xmp (40.8 KB)
darktable 3.4
20200718_0062-rt.CR2.pp3 (11.3 KB)
rawtherapee 5.8
In both versions I have disabled everything except white balance (as shot in camera), and crop, so you can view 100% on screen. Both used Amaze demosaic. Both used linear pro photo as working space. Both saved to jpg, 100% quality, no resize.
Rawtherapee is a little sharper, though the difference is quite subtle. Is this simply due to different implementation of Amaze algorithm?
And now using the OP’s image in this thread. Settings as above. Again, difference is subtle, but rawtherapee looks a little sharper to my eye - edges a little more defined.
DSC00662-dt.ARW.xmp (6.9 KB)
darktable 3.4
DSC00662-rt.ARW.pp3 (11.3 KB)
rawtherapee 5.8
You got rid of the blotchy’s…I will have to look at your xmp…I could get the DT image very sharp but as in the OP…DT had some large dark blotches that interfered with visual sharpness that were not in the RT edit…
EDIT…ah this is a different sculpture…try the one in the OP …it seems trickier due to shadows and that real coarse grain or blotches…
fwiw i think ingo’s deconvolution works very well on this image. it seems to have quite an aggressive optical low pass filter. here are my versions with capture sharpening enabled and disabled (had to dial everything to 11…):
1 Like
Ya I think from the OP it may be more of a contrast issue. Toggling back and forth the DT image has more contrast but this amplifies the large dark blotches…the RT image seem brighter or less contrasted and so maybe appears sharper because its tonally “smoother” in the provided images???..
I don’t think so. The RT result from OP has Capture Sharpening enabled. As @hanatos pointed out, on this image that works very well
Left RT without sharpening, right with Capture Sharpening at default (automatic) settings
1 Like
it looks fantastic on this one. now i wish my camera was so blurry that i could be happy to restore so much detail every time i press that button
which settings did you use there? what’s sigma and how many iterations? i’m suspecting this image is a candidate for the larger filter sizes…
1 Like
great thanks. radius here is really sigma, right? so my sigma was even larger. good to know. didn’t implement the larger filter sizes and would like to avoid that
1 Like
right
I think the CS is great but in the OP look at this area…DT
vs
RT
I don’t think this is the CS
The RT image presented here seems more “blown out” than the DT image but this does make it look smoother and likely even better when sharpened…
1 Like
Showing results of different crops/magnifications/(maybe even lens corrections) is confusing.
Also the OP RT version had more than CS enabled (usm sharpening, wavelets), while the result of my last post had only CS enabled
1 Like
In my experience for the majority of images a 5x5 kernel (which is used in RT up to sigma 0.84) is enough.
There are rare cases (for example shots at >= F22) where a sigma > 0.84 is needed.
I was agreeing with you on the CS for sure and I think it is just his screen shot because when I load his PP3 it looks way less exposed…more like yours so it could just be a web thing
OP image with their sidecar
This doesn’t match to me with what they posted originally…maybe just my eyes…
To be honest I didn’t try the pp3 sidecar of OP, I only looked at its settings using a text editor 