[argyllcms] Re: Relative And Perceptual rendering intent
- From: Dimitrije Zivkovic <migraf.keramika@xxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Sun, 11 Aug 2013 22:18:57 +0200
Thanks for advice
But I want opposite of your suggestion.
I want to Relative be more like Perceptual, not inversely.
Actually I can't change only Relative intent.
When I looked profile that Xrite software produced, Relative intent look more
accurate and intensive colors than Perceptual, but accurate in color conversion.
Because I use some special color instead of Magenta (It's look like some kind
of red) with Perceptual intent I've got more precise conversion, but with
Relative I've got different color than Magenta is, and this is not good.
The question is, can Relative be more like Perceptual?
I put the link of colors and their conversion with Xrite profile and Argyll
profile.
http://www.sendspace.com/filegroup/MMBNuSbhgbXAF%2Fb0yhTijTRVm%2FqSRny7CceSgjg%2F7cg
Dimitrije Zivkovic
On Aug 9, 2013, at 4:42 PM, Nikolay Pokhilchenko <nikolay_po@xxxxxxx> wrote:
> Dimitrije Zivkovic wrote:
>
> Is any possibility to achieve similar results in profile between Relative
> and Perceptual rendering intent?
>
> Also, without using -s option, profile can't calculate right conversion.
>
> My command looks like this:
>
> colprof -v -qh -kp(…) -Zp -Tp -tp -sAdobeRGB1998.icc -l250
>
> Thanks in advance
>
> Dimitrije Zivkovic
>
> Hi Dmitrije!
> I think You shouldn't use AdobeRGB1998 as source color space for gamut
> mapping. This is because AdobeRGB1998 represents a lot of impossible for CMYK
> printing colors. The compression from excessive high gamut volume to the
> gamut of Your printer leads to heavy gamut compression, to excessive
> desaturation of some colors. I'd recommed You select as source colorspace
> more reasonable, more realistic gamut. For example, ISO Coated v2 (ECI) or
> sRGB - depends on the source of images to print.
> If You can compute the profile or device link for every image or image group,
> You can use real image gamut as source gamut for profiling or device link
> computing. This way You'll got a minimum color compression for perceptual
> rendering intent.
> May be You need just black point compensation? If so, You can select another
> intent in place of saturation or perceptual profile table. For example:
>
> colprof -v -qh -kp(…) -sAdobeRGB1998.icc -tla -Zp -l250
>
> With Luminance axis matched Appearance You'll got most similar to Relative
> intent result but with better blacks.
> I think there should be no sense to use "-T" parameter as long as You've
> select "-s" because there should be only "perceptual" table. So no need to
> define saturation table mapping type.
Other related posts:
