[argyllcms] Re: How wide is the colour gamut covered by dispread?
- From: "Gerhard Fürnkranz" <nospam456@xxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Thu, 30 Apr 2009 22:29:03 +0200
-------- Original-Nachricht --------
> Datum: Thu, 30 Apr 2009 06:49:10 -0700 (PDT)
> Von: Stephen T <stwebvanuatu@xxxxxxxxxxxx>
> An: argyllcms@xxxxxxxxxxxxx
> Betreff: [argyllcms] How wide is the colour gamut covered by dispread?
> I found out why my laptop does not profile well: it has a white
> LED backlight and inadequate spectrum.
>
> At least I can calibrate it OK and there's no more blue colour cast.
> Colour temperature is good, tones are satisfactory, everyday colours
> (mixtures) are reasonably OK but pure colours are off, red looks
> orangeish. The colour spectrum of the white LED backlight in my
> display can't make pure red and the gamut of this display is surely
> less than sRGB. BEWARE! My old Dell with CCFL backlight has better
> colours!
>
> Anyway, I notice that Argyll uses very many color patches. Can
> dispread profile wide gamut displays, like an LCD with RGB LEDs?
Sure, why not? I'ts rather a matter of the (currently available) monitor
colorimeters, which may be less accurate on these displays (unless you have an
instrument which is especially calibrated for measuring these kinds of displays
- like e.g. instruments bundled with such displays).
> How wide is the gamut covered by dispread? NTSC? AdobeRGB?
"targen -f <large-number-of-patches> ..." creates a set of patches covering the
whole gamut of the display. Dispread just measures the actual colors (XYZ) of
these RGB patches on your display.
> How does dispread handle reduced gamut displays that can't
> generate pure R, G, B as per sRGB gamut?
Dispread still just measures the actual colors of the patches generated by
targen. It is eventually colprof, which fits a mathematical model to the .ti3
data in order to establish a mapping between device RGB numbers and the
corresponding actual colors (XYZ), and records this mapping in the generated
ICC profile.
> On my "crap" display it seems to
> oversaturate, trying to make colours that are impossible.
Any output device (display, printer, etc.) can only reproduce a limited subset
of the colors that can be seen (and distinguished) by the human vision. How
out-of-gamut colors are reproduced on a device is a matter of the rendering
intent. Colorimetric intent is supposed to reproduce in-gamut colors as exact a
posible, and to clip out-of-gamut colors to the device's gamut boundary, which
may not give pleasing results when the image gamut is significantly larger than
the device's gamut. Perceptual intent on the other hand is rather supposed to
do a pleasing reproduction by compressing the source gamut so that it fits into
the target device's gamut (which desaturates in-gamut colors as well). Matrix
profiles can only describe the colorimetric intent, though. If the gamut of
your display is significantly smaller than sRGB and if you want to get a
pleasing, gamut-compressed (though not necessarily colorimetrically correct)
reproduction of sRGB images, then you could try to create
a CLUT profile with perceptual tables (e.g. "colprof -qh -cmt -dmt -S sRGB.icm
<name>") and to apply this profile with perceptual intent.
Regards,
Gerhard
--
Psssst! Schon vom neuen GMX MultiMessenger gehört? Der kann`s mit allen:
http://www.gmx.net/de/go/multimessenger01
Other related posts: