Re: MeasureTool Compare question
Re: MeasureTool Compare question
- Subject: Re: MeasureTool Compare question
- From: ignacio ruiz de conejo <email@hidden>
- Date: Fri, 18 Mar 2005 08:20:20 +0100
Marco, Bruce:
Anybody please let me know if you see any mistakes or problems in this
procedure.
I am in agreement with the procedure, but would like to here
about a related subject, which is the need to use a different
testing data set than the one used to generte the profile.
You both have mentioned using the ECI2002 target, but that target
is often used by some profilers. Since I want to use a target,
independently of the generating target, I have developed the
following process:
1- Generate a Lab target that samples the space, with given increments
(defaults: dL=5, da=10,db=10; 0<L<100, -100<a,b<110)
2- Determine the gamut of the profile, using Jan Morovic's segment
maxima method, implemented in his Public Gamut Mapping Algorithm
3- Filter the out of gamut points of the original Lab target.
For a 'normal' CMYK profile, that leaves around a thousand points.
4- Follow the rest of your procedure
The goal is to avoid using points that were used to generate the
profile. And the process above addresses that, I think.
(different criteria could be applied here to test a profile: giving
more importance to colors in the gamut border, more frequent colors,
etc... but so far I am just using a regular distribution in Lab)
How to test a profile for out of a gamut colors, is a different topic
I also would like to here about.
Best regards,
Ignacio
--
______________________________________________________________
Ignacio Ruiz de Conejo Tel. +34 93.5821342
Color & Imaging Group, R&D Fax +34 93.5824816
Hewlett-Packard Española, ICD e-mail: email@hidden
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden