[argyllcms] Re: Can I calibrate crap?

-------- Original-Nachricht --------
> Datum: Mon, 27 Apr 2009 13:57:32 -0500
> Von: Leonard Evens <len@xxxxxxxxxxxxxxxxxxxxx>
> An: argyllcms@xxxxxxxxxxxxx
> Betreff: [argyllcms] Re: Can I calibrate crap?

> > One thing I don't  understand about the original question. I'm
> > surprised to hear of a laptop with an LED backlit display. It
> > would eem surprising for a manufacturer to produce a `crappy'
> > such display.

In some recent displays white LEDs indeed replace the CCFL. But replacing the 
light souce still does not make the TFT panel better than it is, nor does it 
make these display high-end ones. IMO the major reason for using (white) LEDs 
in these display is reduced power consuption.

High-end displays with LED backlight have different objectives, e.g. wide gamut 
and separately adjustable red, green and blue LEDs (not white ones!), in order 
that the whitepoint chromaticity can be adjusted via the backlight. Low power 
consumption is not necessarily the goal here. The XL20 for instance consumes 
more power than displays of comparable size and brightness with CCFL backlight 
do consume (it even needs a fan for cooling). Obviously the used red, green and 
blue LEDs are not as efficient as CCFL or white LEDs...

> Let me add that were it really LED backlit, then you probably
> can't use a colorimeter to calibrate/profile it. You would need
> a Spectometer of specialized colorimeter designed for such
> illumination.

One should of course not blindly expect that spectrometers are the universial 
remedy, but one should always keep in mind that they suffer from various errors 
as well. One source of inaccuracies is the finite bandwith. For measuring the 
color of narrow-band light sources accurately, bandwidth/interval should be IMO 
<= 5nm (or even better <= 1nm), however the i1 Pro for instance reports at 10nm 
bandwith/interval (in regular mode). Or another example, the Spectrocam has a 
BW of even 20nm (despite it reports at 5nm interval). The following diagram 
illustrates how the (measured) spectral locus gets distorted by different 
bandwidths:
http://home.arcor.de/gfuer/tmp27958/effect_of_bandpass.png
[i.e. an error of up to 0.05 chromaticity units can be indeed expected at 20nm 
BW]

Regards,
Gerhard

-- 
Neu: GMX FreeDSL Komplettanschluss mit DSL 6.000 Flatrate + Telefonanschluss 
für nur 17,95 Euro/mtl.!* 
http://dslspecial.gmx.de/freedsl-surfflat/?ac=OM.AD.PD003K11308T4569a

Other related posts: