Skip to main content
R
Rufus2015
Known Participant
March 28, 2018
Answered

Absolute and Relative, again.

  • March 28, 2018
  • 1 reply
  • 3667 views

I made up a digital copy of an x-rite CCP, using RGB values.

I then removed the brightest white patch, by making it transparent, and de-saturated the colours until they were all within gamut of my printer/paper combination.

The colour patches are separated by a transparent strip.

Now I printed this using Absolute and Relative rendering intent.

I expected both prints to be the same, as no colours, black or white were out of gamut.

The absolute rendering looks good, given I don't have a viewing booth.

But the relative has made the lighter colours 'muddier' and shifted the "Black" and the lightest Grey, reducing the contrast.

Q:Is it the transparent part causing the problem, or is something else going on?

This topic has been closed for replies.
Correct answer NB, colourmanagement

Hi Rufus

The Adobe Color Print Utility would print with no transform, but, then you don't get to choose RelCol or AbsCol which seems to be what you are after.

In theory your method of "working in printer colourspace" - ( either that or just assign printer ICC to the file before printing)  that might work, we call it a null transform.

Or, say your original's Adobe RGB, then select the same colourspace (Adobe RGB) as printer profile. That should work too.

But we could help better if we knew why you are trying to do this?

Also you write:

"I made up a digital copy of an x-rite CCP, using RGB values.

I then removed the brightest white patch, by making it transparent, and de-saturated the colours until they were all within gamut of my printer/paper combination."

whats a CCP?

Did you use gamut warning to check that all colours were within destination gamut?’

I don't think thats a definitive indication of "fit"

Although it's commonly thought to fulfil that need, what gamut warning actually reveals is areas of an original which will move by a certain delta E when being converted to the destination profile. So something "just" out of gamut would not show up.

Also - gamut warning, in my experience, is not at all good at showing dark or light "out of gamut" colours - it only seems to show colours around mid-range. And example, dark suntanned skin, maybe oiled for the beach, gamut warning is OK but convert to printer ICC and the areas approaching shadow may lose colour, so instead of dark brown they are printed as grey. With a 32D gamut mapper like thew wonderful Colorthink its easy to see whats happening here, the area is moving UP inn the colourspace (lighter) rather than towards the centre (less saturated)

There are 4 ways to check whether its in gamut that I can think of right now, there may be others:

1: Use Colorthink from Chromix, map the image pixels in 3D Lab against the destination colourspace gamut boundary

(Colorthink can even create vectors to show the start and end point for each mapped pixel going thorough an ICC-ICC transform)

2: Look at the Photoshop levels histogram before and after, any clipping? this tends to illustrate "out of gamut" properly.

3: (this only works for colour patches not images which would be impractical) - convert to the output colourspace and compare all the patch values, tedious I know. makes Colorthink look extremely good value.

4: you already thought of this one, create the patches in printer colourspace.

I hope this helps

if so, please do mark my reply as "helpful" and if you're OK now, please mark it as "correct" below, so others who have similar issues can see the solution

thanks

neil barstow, colourmanagement

1 reply

davescm
Community Expert
davescmCommunity Expert
Community Expert
March 28, 2018

Hi

Relative Colometric intent adjusts for the white point of the media, absolute colormetric does not. So a difference in those near white colors is expected.

Render Intent - Color Management Systems - Technical Guides

Dave

    R
    Rufus2015Author
    Known Participant
    March 29, 2018

    Hi Dave,

    Yes,  know that, I was trying to produce a chart that would not be altered by rendering intent.

    So, is it possible to work in the printer's colour space, so no colour adjustment takes place when printing?

    R
    Rufus2015Author
    Known Participant
    March 29, 2018

    Hi Neil,

    We spoke on the phone a year or so ago, so hello again.

    I copy art work for local artists, which I manage to get reasonably OK, but at the cost of lots of test prints.

    I thought that if I could print a chart, so that I have a digital copy and a paper copy whose colours are very similar or identical,

    then photographing the chart to create a profile which could then be applied to my other photographs and get close to a decent copy.

    The CCP is x-rites colour checker passport, but some of the colours are out of gamut, so the printed version has shifted colours.

    This all feels like swimming in treacle, having to fight against the different colour gamuts.

    I thought that producing a colour chart that was within the printer/paper gamut would make this easier, but it all seems to be near impossible.

    I am trying to eliminate as many variables a possible, but the rendering intent colour shifts are making this harder.


    I use Lightroom for my work, and that does not have the absolute rendering intent, so I thought my modified colour chart would allow me to use the relative intent as if it were absolute, but this is not the case, hence the thought that using the printer colour space would help, but this also seems impossible.

    Terms & ConditionsAccessibility statement
    Using the community Terms of use Privacy Cookie preferences Do not sell or share my personal information ad choicesAdChoices