Output/jpeg problem

RT5.5 appimage

The display doesn’t match the jpeg output. The monitor is calibrated

Note the toning (intended)

Note the output as a straight B&W.

What am I doing wrong?

I’m trying to do a B&W with a slightly warmer highlights.

Thanks

Syv

I’m going to take a guess, are you saving the JPEG as 1-channel monochrome? If so, you’re losing your ‘color’; in spite of the image being ‘black and white’, you need the RGB channels to convey the toning.

1 Like

Can you upload the raw and resultant jpg ?

@foto And the .pp3 please

Here are the files:

cr2/raw: https://drive.google.com/open?id=1zg1oI-CPqmo0h8CNA9xIyEMBmTl06eSf

jpeg try1: https://drive.google.com/open?id=1cxxduPcUNRpdLwNr0Z6ujw2dmyOxPxV_
pp3: try1: https://drive.google.com/open?id=19TNQKp4xccep-AKk1elo0hdFe-q5JQic

jpeg try2: https://drive.google.com/open?id=1SkkhnhUjH8rPJR3j4zl93yP3BpO5J_2X
pp3 try2: https://drive.google.com/open?id=1CPoTcizmT01i4hJ96IeYwZX8u4S-CdTe

Thanks

2 questions then:

  1. How do I do a sepia b&w?
  2. My main problem is why does the RT editor shows a toned image but the 8bit rgb output is strictly b&w. They should match.

Thanks

Syv

Black and white is a misnomer. It is really grey scale. Often 3 channels (RGB) of the same grey image. Sepia is not black and white because the values of each channel aren’t the same.

1 Like

I’m trying to have a B&W image with slightly warmed highlights. What do you suggest for me to do?

Thanks

Syv

tl; dr; desaturate the image, then do a red channel curve and lift the left point a bit. Save this as a three-channel RGB image. (I just looked up tl; dr; and decided it was pertinent… :smile: )

Okay, I just tried this in RawTherapee, but I have to learn some things before I get there. I’m going to illustrate in rawproc, my hack editor, and petition someone more RT-knowledgeable to tell us how to do such there.

This is one of my favorite monochrome conversions. I start with the color image, developed to look nice. I then apply the grayscale tool, which desaturates the channels to make them equal, R=G=B for each pixel. RT has such a tool. Note that after I do this, it’s still a RGB three-channel image, R=G=B. Notice the histogram; it’s just the blue channel, because the other two are behind it exactly:

Then, I add a curve, and set it to the red channel. Then, I drag the lower-left point up just a scooch, in this case up 6 on the y axis. Notice the histogram now, the red channel is now shifted right just a bit, and the image now has what I’ll call a “poor man’s sepia” tone:

The point is, a sepia toned image is a color image. All Sepia.

1 Like

@foto
Test with W10, last dev branch
use of CIECAM2/Viewing conditions/temperature for toning
I can reproduce. As the output is not the same as what is displayed, and it is not a softproofing, for me it seeems a bug.

For that I should use color toning/color correction region

I pushed a fix

This compels one to use the CIECAM module for monochrome toning, no?

Otherwise, there’s no way to desaturate a color image and just leave it as R=G=B colo, and then mess with tone … ??

No, it just makes preview and saved file consistent when using b&w and ciecam02 in combination.
Before the fix you got R = G = B in saved file and the toned output in preview when using b&w in combination with ciecam02. Now you get the toned output also in saved file.
You still can use the color toning module or the Lab a and b curves for toning after b&w as before.

2 Likes

Doesn’t really work. I have tried various condition:

  • sliders
  • lab color correction grid
  • color balance shadow/mid/high

Each one give me a different color (which it should) but when I change the values within each method, it does change the preview but the jpeg output is identical within the coloring method. (I didn’t try extreme values)

Thanks

Syv