Re: GaMapICC v0.5
Re: GaMapICC v0.5
- Subject: Re: GaMapICC v0.5
- From: Klaus Karcher <email@hidden>
- Date: Mon, 25 Jan 2010 20:02:17 +0100
Hi Marco,
many thanks for your feedback.
Marco Ugolini wrote:
I want to let you know that I have released GaMapICC version 0.5:
<http://digitalproof.info/gamapicc/>
Thank you, Klaus.
I tested the application by converting an RGB file to a custom CMYK profile
that I have used for one of my clients.
I used an image-specific gamut mapping strategy with 100 percent gamut
filtering, for print evaluation environment viewing conditions.
I have posted the results at this URL:
<http://files.me.com/marcougolini/8mxtcb>
The compressed archive contains 3 files: one is the RGB source image,
another one is the GaMap-converted one, and the third is the one I converted
using Photoshop (RelCol). Each file name clearly indicates which is which.
To me, the Photoshop conversion looks better: the tonalities more closely
match those of the source image (no brighter, no darker), and the red
saturated area around the blade of the axe reveals more detail. Whereas the
GaMap-converted image appears too light and shows less detail in that same
saturated red area.
Overall, these GaMap results don't look very impressive to me. Please don't
take offense: I am trying to have an honest assessment.
I appreciate your honesty.
I agree with you: the rel. col. Photoshop conversion looks better than
the GaMapICC conversion (at least on screen).
The tonality shift in the GaMapICC version /might/ be intentional and
appropriate depending on your source viewing conditions (I tried md-pp,
mt-pp and pp-pp but was not able to reproduce your results exactly. Did
you apply any changes to the image after conversion? e.g. downscaling or
sharpening?)
Please note also that Photoshop automatic black point compensation for
device->display transformations might affect the screen assessment
(Photoshop doesn't simulate the "true" black point unless you activate
"Proof Colors" and select the "Simulate Black Ink" checkbox in the
"Proof Setup" window).
However there are issues with GaMapICC in respect of Gamut Mapping as
well as source gamut boundary calculation. Your example revealed a
glitch in the gamut filtering code I have to discuss it with Graeme.
Please let me know if other GaMap conversion settings would produce better
results.
Please give me some time to find out what's going wrong before I go into
detail.
And just one additional note in order to guard against
misunderstandings: Image specific gamut mapping is no silver bullet. It
does not magically provide the optimal result straightaway in each and
every case. There are many cases where much simpler methods like black
point compensation and clipping are just the right thing. GaMapICC with
100% filtering should do nothing else but apply exactly as much gamut
compression as necessary to avoid clipping. It can not decide how much
clipping might be good for an image.
In addition GaMapICCs gamut filtering option should provide a
continuously variable control to adjust the gamut mapping between "no
clipping" and "no gamut compression", but obviously this does not work
as expected yet.
Many thanks for your feedback again,
Klaus
_______________________________________________
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