Re: Screen profiling

Gernot Hoffmann wrote:

Let's have a look at the standard scenario as valid for commercial
programs (Photoshop for image processing).

1. The monitor is calibrated by monitor hardware as near as possible
to G=2.2 and D65 and luminance 100cd/m2 (my settings).
2. Necessary corrections (calibration) are applied by Eye-One.
Then the monitor is profiled, using Eye-One and commercial
software. This delivers a unique ICC profile which is independent
of any paper. In fact, it describes the relation between physical
monitor colors and color numbers.
3. Darkness is called L*=0. Full white is called L*=100.
Then the Lab graypatterns in my doc can be used for checking the
achieved result.
4. Once a CMYK profile for a specific printer, paper and process is
available, one can use Photoshop's Softproof. Here we have the
option 'Use papergray'. The result looks often ugly, but after
some
adaptation, at least the general shift should be shown correctly.

I think we are approaching the problem form different angles.
You use commercial programs, feed them with your Lab values and
observe what they do. Putting L=0 in your file does not make
your computer actually display it.
I write commercial programs and am wondering what is
the right thing to do with the values you feed in.
I thought the whole idea of calibration
is to have one value to describe the color, like for example
CIE L=15 a=0 b=0 at D65 measured on a dark fabric.
My understanding is that calibrated monitor will display something
close to RGB=0,0,0. Because and LCD black will still be much
lighter than that piece of fabric. Viewed at any conditions under
any light, even with drunk observer.
Just like I have never measured a yarn with L higher than 92,
if I omit the fluorescent ones ;-)
So having the L=0 and L=100 in your file is the same as putting
L=25 b=-100 a=100, it just test how does the system behave on
out of gamut colors, not its actual color accuracy.

I could be mistaken and will be glad to improve my
colorimetry engine, if you have a better idea on
what to do.
For example, I would like to try CIE Lab D65 to sRGB
conversion, if you have a formula for that or any
good web addresses to share. I know sRGB is not regarded
highly in this group, but I am prepared to try anything.

Best regards,

Dušan Peterc
http://www.arahne.si
.

Relevant Pages

  • Daily Report #4585
    ... WFPC2 CYCLE 15 INTERNAL MONITOR ... the calibration pipeline. ... these grains are subject to size-dependent ... A Paschen-Alpha Study of Massive Stars and the ISM in the Galactic ...
    (sci.astro.hubble)
  • Re: Olympus low pricing; Never again?
    ... >> calibration software to match their output, ... > What does a printer profile have to do with monitor calibration and ... The monitor should be hardware calibrated to "the ... > the local lab, the image on the monitor looks the same. ...
    (rec.photo.digital.slr-systems)
  • Re: Olympus low pricing; Never again?
    ... > will look like it does on my monitor. ... So your lab has a screwed up profile. ... a whacked monitor calibration that ONLY works at that one lab who is using ...
    (rec.photo.digital.slr-systems)
  • Re: Olympus low pricing; Never again?
    ... If I take that file to my lab to get it printed, ... >> will look like it does on my monitor. ... > So your lab has a screwed up profile. ... > a whacked monitor calibration that ONLY works at that one lab who is using ...
    (rec.photo.digital.slr-systems)
  • Re: Olympus low pricing; Never again?
    ... > calibration software to match their output, or don't you actually have any ... What does a printer profile have to do with monitor calibration and viewing ... the local lab, the image on the monitor looks the same. ... the monitor to match some other output device. ...
    (rec.photo.digital.slr-systems)