RSS/Atom feed Twitter
Site is read-only, email is disabled
  1. Chat
  2. Useful links
  3. Polls

Featured content

New color mode

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.

2 of 2 messages available
Toggle history

Please log in to manage your subscriptions.
New color mode Troy Large 13 Dec 05:25
  New color mode Graeme Gill 13 Dec 23:56
Troy Large
2011-12-13 05:25:09 UTC (almost 13 years ago)

New color mode

Martin Nordholts gmail.com> writes:

On 07/01/2009 08:21 PM, Emil Assarsson wrote:

On Mon, Jun 29, 2009 at 9:33 PM, Martin Nordholts gmail.com>

wrote:

Maybe the RGB -> HSL conversion is buggy?

I don't see it as a bug... it's more of a problem. Gimp uses HSL(?) and PhotoShop /seems/ to use CIELAB for these operations.

Any comments on this?

If using a CIE Lab based color space for defining and implementing the color layer modes gives better looking and more intuitive results, than that is the color space we shall use.

A patch that uses CIE Lab (or CIE LCH?) for this when using GEGL would be much appreciated.

/ Martin

For sure one should use CIELAB. There is a good write up about that color model here: http://www.handprint.com/HP/WCL/color7.html (you will find CIELAB a little over half way down the page. CIELAB, while quite good in itself is also the preferred system for transformation between color spaces - for example going from RGB to CMYK one is best to go RGB > CIELAB > CMYK, thus, not only would you have a top rate color system, you would be set up for top rate transformation functions as well

Graeme Gill
2011-12-13 23:56:40 UTC (almost 13 years ago)

New color mode

Troy Large wrote:

over half way down the page. CIELAB, while quite good in itself is also the preferred system for transformation between color spaces - for example going from RGB to CMYK one is best to go RGB > CIELAB > CMYK, thus, not only would you have a top rate color system, you would be set up for top rate transformation functions as well

Well, actually no. L*A*B* has often been used in this process as a gamut mapping space (needed when converting RGB to print spaces), but it does have the disadvantage that radial colorspace changes are not constant in hue, particularly in the blue region. This is no handicap for many uses (including an image edit space), but it is for gamut mapping. The more modern approach in gamut mapping is to use a space that does have constant hue in the radial direction, such as CIECAM or IPT spaces.

Graeme Gill.