RSS/Atom feed Twitter
Site is read-only, email is disabled

Colour management and the proof module

This discussion is connected to the gimp-developer-list.gnome.org mailing list which is provided by the GIMP developers and not related to gimpusers.com.

This is a read-only list on gimpusers.com so this discussion thread is read-only, too.

4 of 4 messages available
Toggle history

Please log in to manage your subscriptions.

Colour management and the proof module Alastair M. Robinson 05 Aug 02:14
  Colour management and the proof module Sven Neumann 05 Aug 15:00
   Colour management and the proof module Simon Budig 05 Aug 18:54
    Colour management and the proof module Alastair M. Robinson 05 Aug 19:32
Alastair M. Robinson
2004-08-05 02:14:59 UTC (over 19 years ago)

Colour management and the proof module

Hi All,

I've been playing with the colour proof display module, since it seemed like a good place to start playing with colour management.

As it stands (in gimp-2.1.2), the proof module only allows a single profile (for the monitor) to be selected. Profiles are only useful in pairs, so it builds a transform from lcms's built-in sRGB profile to the monitor's profile, which is then used in the display filter.

I've played around a bit and improved the filter to the point where source, destination and proofing colour profiles can be selected individually, and am now getting some impressive results from my newly-profiled scanner and a rough profile generated for my Laptop's LCD, which like many older TFT displays has a rather odd colour balance.

Now I have a couple of questions:

Is there any way of retrieving an image's parasites from within its display filters (i.e. can the filter reference the image), or am I limited to global parasites?

Has any thought been given to setting up global display filters so that the user doesn't have to select them for every newly-opened image?

Finally, someone (Sven?) mentioned something about widgets (or more precisely, the colour selectors) being able to make use of the display filters; how far has this progressed?

All the best, --
Alastair M. Robinson

Sven Neumann
2004-08-05 15:00:57 UTC (over 19 years ago)

Colour management and the proof module

Hi,

"Alastair M. Robinson" writes:

Is there any way of retrieving an image's parasites from within its display filters (i.e. can the filter reference the image), or am I limited to global parasites?

No, there isn't yet but this is one of the changes that I have mentioned that need to be done to make color display filters useful.

Has any thought been given to setting up global display filters so that the user doesn't have to select them for every newly-opened image?

The same answer applies here.

Finally, someone (Sven?) mentioned something about widgets (or more precisely, the colour selectors) being able to make use of the display filters; how far has this progressed?

The display filter stack is in libgimpwidgets so it is useable from the color selectors that are also in libgimpwidgets. What's missing here is integrating the display filter routines to GimpColorArea, GimpPreviewArea, GimpColorScale, ... This also depends on the framework for global display filters mentioned above.

However I don't see a way how color selectors could be dependant on an image. I don't think the image color profile can in any way be applied to the color selectors. Simply because the color selectors don't know about the image.

Sven

Simon Budig
2004-08-05 18:54:36 UTC (over 19 years ago)

Colour management and the proof module

Sven Neumann (sven@gimp.org) wrote:

"Alastair M. Robinson" writes:

Finally, someone (Sven?) mentioned something about widgets (or more precisely, the colour selectors) being able to make use of the display filters; how far has this progressed?

[...]

However I don't see a way how color selectors could be dependant on an image. I don't think the image color profile can in any way be applied to the color selectors. Simply because the color selectors don't know about the image.

One could assume a standard profile for the color selectors (sRGB?) and transform the color to the image profile when the tools accesses the foreground for painting in a specific image.

Probably this should be optional for the people who need exact RGB-Values in their image.

Bye, Simon

Alastair M. Robinson
2004-08-05 19:32:29 UTC (over 19 years ago)

Colour management and the proof module

Hi,

Simon Budig wrote:

However I don't see a way how color selectors could be dependant on an image. I don't think the image color profile can in any way be applied to the color selectors. Simply because the color selectors don't know about the image.

One could assume a standard profile for the color selectors (sRGB?) and transform the color to the image profile when the tools accesses the foreground for painting in a specific image.

At this stage it has been decided that the core of GIMP shouldn't be affected by colour management issues - the general consensus is that such drastic changes should wait until GEGL time. This means that in general, images will have to share a common source colour-space for the time being. While I think this is less than ideal, I can live with it as long as changing the working profile isn't too much of a chore!

I also see no reason not to allow the source profile to be over-ridden by a parasite, so I'm encouraged that it will ultimately be possible for the display filter to retrieve them from the image. This will make some interesting tricks possible:

For example, the separate plugin has one mode in which it generates an RGB image with solid Cyan, Magenta, Yellow and Black layers, and places each plate's data in the layer masks. Each layer's mode is set to Darken Only, which gives a rudimentary full colour display of CMYK data. This is equivalent to the naive R=1-C style CMYK->RGB conversion, so it gives over-saturated results, but I've generated a source profile that significantly improves this display.

For tricks like this, it wouldn't matter one jot if the colour selector doesn't match the image display

All the best, --
Alastair M. Robinson