No graduate saturation (example pic included)



PNG from RAW:

Look in the center how the blue around the yellow euro sign is very blue without gradually building to it. Turning saturation down does not help. HSV Equalizer same.

Is there a special tool to handle this?

(Ingo Weyrich) #2

Not without having an example file to reproduce. If possible, please upload a raw and the .pp3


Could not upload the full photo from the topic start since it is a portrait with someone in it.

I thought I found an alternative, but with the alternative the fix is just to filter out some blue. With the one from the topic start that does not work.

The problem is that the one from the topic start just has a smack of blue on it. Within that blue are no different blues. So the question is how to bring that out. I will try to find a working alternative meanwhile.

(Ingo Weyrich) #4

No chance without an example…


Try turning off the input colour profile, under Color tab, Color Management, set “No profile”. If colour accuracy is important, you need to find/create a well behaved colour profile for your camera.

(Glenn Butcher) #6

It would be informative to know what camera you shot this with. Even more informative would be the complete raw file, but I understand why you don’t want to post that.

It looks as if the camera-raw RGB image is being put into the PNG without gamut conversion to an appropriate output profile. The camera JPEG has benefit of such conversion from the in-camera JPEG engine.

+1 to @james suggestion. My colors got a whole lot better after I produced such for my D7000, just using a 24-patch ColorChecker Passport.


Yes that seems to work. The gradient is back! Colours are a bit flat, like the JPG but with some work it will look nice.

D7500. I will keep looking for a RAW that I can upload.

I am now trying the trick but just as with the D3300 I used before, with one of the files the colors are all wrong and with the other one the image starts out extremely light and flat, which requires huge adjustments in lightness and contrast. Is that normal?

(Glenn Butcher) #8

That raw would need similar colors to illustrate the difference.

You might try using the profiles that come with the Adobe DNG Converter. Also at RawPedia: I did this last weekend and found the D7000 Adobe Standard profile to be pretty close (well, I haven’t compared them side-by-side yet) to my calibrated camera profile.

Thing is, what your two images indicate is a gamut truncation, and resolving that will require tracing your image’s colorspace through the entire input-to-output mousetrap that is RawTherapee (or any other raw processor, for that matter), making sure you know with what profile you start, and where each colorspace conversion is happening. That in-camera JPEG engine becomes more appreciated, no? But, when you get to the bottom of this nit, you’ll understand color a whole lot better, or your care will be entrusted to the appropriate institution of mental health… :smiley:


Adobe profile does even better with creating gradient blue. On the reds though it is too strong. My model became a red smurf haha.

Maybe custom is the way to go yeah. Tough stuff.

(Morgan Hardwood) #10

Use an input profile which handles artificial light.

If you have access to a color target, then you can help improve support for the Nikon D7500 for everyone:


Those Color Targets are some expensive pieces. :open_mouth:

I wish I could mix the Nikon .icm with the Adobe .dcp, preferring the Adobe blue and the Nikon red.


If you use windows or mac, adobe provide a cost free dng profile editor, which should let you tweak the profile. Although I have not personally used it.

(Alberto) #13

A cheap alternative is to generate a dual-illuminant DCP profile by expoiting dpreview’s “studio scene comparison” tool. Just grab the raw files for the two illuminants for your camera from dpreview’s webpage, crop to include only the colour targets, and generate a DCP following the instructions on RawPedia. It won’t be 100% accurate maybe, but on my tests it gives pretty usable results…

(Morgan Hardwood) #14

Try these:

(Cranberry Carborundum) #15

To add on to this discussion, the “Adobe Standard” DCPs seem to cause gamut clipping in images similar to yours. I just had this specific problem and had to do a bit of fiddling around.

It seems that the Camera Neutral DCPs provide a better starting point when you really push those colors through Lab adjustments and LUTs.

(Morgan Hardwood) #16

RawTherapee’s DCPs prevent gamut clipping.

(Cranberry Carborundum) #17

My camera doesn’t have a DCP yet, and isn’t good enough to warrant purchasing a color target (crappy compact). For this case the Adobe-supplied camera neutral profiles provide much better colors than the embedded matrix.

I’m very sure the RT DCP would be good if there was one :slight_smile:

(Morgan Hardwood) #18

@CriticalConundrum but there is one, in fact two - I posted two variations above. The only reason I won’t ship them with RT is because they’re based on DPReview’s photos which are under an unknown licence. That doesn’t prevent you from using them for personal use.


Ok guys thanks for all the tips. I am getting quite confused now.
ICC, ICM, DNG, DCP, etc. Method 1, 2, 3, 4, 5. Is there a complete explanation on everything somewhere?

Reading now.

(Glenn Butcher) #20

DCamProf has a quite lengthy doc page, with a lot of discussion on what goes on inside both ICC and DNG profiles. Specific to this thread is:

A bit technical, but there’s enough descriptive prose to begin to understand some of what’s going on with deep blues.