I would like to add that one goal of an ICC workflow is to visually match
the endpoints; i.e., the images presented by display and output devices.
As someone has already duly noticed, simply having files tagged with ICC
profiles--while important--misses the mark of "color management," which is
to be able to do something with the well-defined values in the image.
I believe the lack of measurement hardware support (i.e., not
video cards) is the biggest deterrent to Linux-based color
management and image proofing.
Unfortunately, for my domain, hardware support is the biggest issue in a
Linux ICC workflow. It's not an issue of software, because there are so
many talented developers who can read the ICCv4 spec and implement it and
can read the abundant papers on colorspace conversion, and implement that.
I can even claim to be one of these people. The issue is getting hardware
colorimeters and spectrophotometers to measure monitors and printouts.
Some companies have done a fine service of shipping very usable software
(read, can run on Linux platforms) with their hardware. X-Rite, for
example, ships their older spectrophotometer (DTP41) with a Java-based app
to capture the values from the spec. No device drivers are needed; the
spectro plugs into a serial port.
Unfortunately, being able to measure output values only solves half the
problem. For a full color-correction workflow, I need to be able to
soft-proof, which requires that my monitor is profiled and calibrated.
To do so requires a colorimeter--or spectrophotometer--which can be used
by Linux.
Even performance is not such an issue. For real-time image rendering,
matrix-shaper transforms can be used to give a rough idea of what the
image looks like because typically we are only going from one RGB space
into another. On any reasonable CPU today, we can find matrix
multiplication routines in hardware, which could easily transform a
million pixels in very very little time. Furthermore, image-editing
software can provide a "soft-proofing toggle" which then uses better
techniques (e.g., tetrahedral interpolation using CLUTs) to provide the
highest-quality WYSIWYG capability before pressing the "Print" button.
An ICC-aware (or otherwise color-smart) video card is only solving the
performance issue...It's not resolving the monitor characterization
issue--which is the true variable.
Right now, I cannot choose to sacrifice performance for quality
because I cannot profile or calibrate my monitor. It seems to be
a matter of slow vs. impossible. I would much rather have the
first scenario (which doesn't currently exist) than no choices
with the second scenario. I simply think that visual gamma
matching is insufficient for output-quality workflows.
Now...If someone could just come up with a hardware calibration for
monitors in Linux...I could think about switching back. =)
--troy
On Sun, 23 May 2004, Karl Heinz Kremer wrote:
> Gerhard is correct, the CMS in X11 is not based on ICM profiles. You
> can find some information in the man page to xcmsdb. Sun did implement
> a CMS system based on ICM profiles (using a Kodak system) for Solaris.
> This proofs that it can be done, it's just a matter of time and
> know-how.
>
> I looked into the X11 CMS quite some time ago, and it turned out to be
> not very useful for serious applications (I was toying with a CMS aware
> version of the Gimp, but the X11 CMS was not promising enough, and
> using LCMS in Gimp's display layer was too slow).
>
> Karl Heinz
>
>
> On May 23, 2004, at 5:43 AM, Gerhard Fuernkranz wrote:
>
> > Hal V. Engel schrieb:
> >
> >> X11 seems to have support for ICM profiles. I stumbled across it
> >> yesterday while researching this on the web and even looked at some
> >> of the man pages for it on my SuSE 9.1 box. There are about 43 ICM
> >> related X11 commands. I don't know if it works. And I am sure that
> >> no one is using it because no one even knows it exists.
> >>
> > Yes, X11 does have some color management support (see e.g.
> > http://www.stanford.edu/~lharcke/programming/Xcms), but my
> > understanding it, that X11 color management is very limited, i.e. it
> > seems to support only matrix/shaper for Displays, but is by far not a
> > full-featured CMM, and it *IS NOT* based on ICC profiles, but
> > X11-proprietary.
> >
> > And it is also not application transpararent, but can only be used
> > from applications which explicitly call the provided API functions.
> >
> > However, basically the latter applies to Windows as well. Programs
> > like Adobe Gamma also DO NOT provide full color management for
> > applications which are not CM-aware. My understanding is, that Adobe
> > Gamma just loads three 1D-Luts for the R, G and B channels from the
> > profile's vcgt into the video driver to adjust gamma and to possibly
> > alter the white point. Surely, this implicitly does affect all - even
> > non CM-aware - applications, but they still don't get the whole
> > monitor profile applied, but only the vcgt. See also below.
> >
> > Best Regards,
> > Gerhard
> >
> >
> > ----- quoted from a previous posting to this list -----
> >
> > This has been and still is a topic under hard discussion in many
> > forums :-)
> >
> > Of course this approach has advantages, but to my understanding also
> > has serious problems. Let's resume first how these profiles works.
> >
> > A typical profile using vcgt is a "two step" beast. Fist step is
> > applied on
> > computer startup. The profile loads the videocard with a compensation
> > to make the display behave as a perfect 2.2 (or anything else) gamma.
> > Ok so far. But this does not change phosphor primaries, nor whitepoint.
> > So, all non-color managed applications are still operating in the
> > native monitor space, only the gamma they "see" now is 2.2
> >
> > Then comes the second part. To do compensation for white point and
> > primaries, a color savvy app still needs to use the profile. Just the
> > profile is assuming gamma of monitor as a perfect 2.2. At that point
> > the effort required by the color managed app is exactly same as the
> > profile were not
> > touched the hardware ramps at all.
> >
> > Benefits? well, all non-color savvy apps does "see" a gamma of 2.2 But
> > the primaries and white point are still wrong. Drawbacks? Setting the
> > gamma
> > of videocard is hardly OS-dependent, so, lcms would not be portable if
> > doing so. And more importantly, The profile is doing adjusts on your
> > hardware, which many people probably don't want. I assume a profile is
> > a characterization of how a given display does behave, not a way to
> > configure the display.
> >
> > Anyway, just my 2 cents.
> >
> > Regards,
> > Marti.
> >
> >
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: Oracle 10g
> Get certified on the hottest thing ever to hit the market... Oracle 10g.
> Take an Oracle 10g class now, and we'll give you the exam FREE.
> http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click
> _______________________________________________
> Lcms-user mailing list
> [EMAIL PROTECTED]
> https://lists.sourceforge.net/lists/listinfo/lcms-user
>
-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g.
Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click
_______________________________________________
Lcms-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/lcms-user
