Mark Brader:
> > The human eye's white point, although adaptive,
> > seems calibrated for a bluish white, to compen-
> > sate for the sky's blueness. Were it not so,
> > everything would look bluish outdoors.
>
> Nonsense. First, the blue of the sky is simply a
> scattering of part of the white light from the
> Sun; the Sun itself looks yellowish due to the re-
> moval of this light. When the skylight and direct
> sunlight are recombined -- either in filtering
> through clouds, or directly on striking an object
> on a clear day -- you get white again.
Thanks for correcting my naive views. The only
thing I am unsure of is the recombination you men-
tioned. I don't think it can yield white again be-
cause of directivity and absorption differences. Or
do you think that the spectrum of the Sun's "recom-
bined" light on Earth is similar to that it really
emits, as measured from space?
> Second, the eye is sufficiently adaptive that the
> claim of being "calibrated for a bluish white" is
> meaningles. People move all the time from day-
> light to fluorescent light to incandescent light
> and hardly notice any differences in color unless
> they're looking for them with special care.
Well, I easily do, because the adaptation does not
compensate for all possible spectrum differences,
although it does help remove color casts, and even
that not entirely. So I think that adaptation only
decreases perceived defferences, never removing
them.
> > Every owner of a decent digital camera can test
> > this.
>
> Standard photography, whether film or digital,
> does not render colors perfectly. It will cer-
> tainly show the differences between, say, daylight
> and incandescent light, but this does not prove
> anything about the eye.
When it comes to the white point and eye adaptation,
I am sure it does. One can calibrate the white
point against a sheet of paper outdoors in sunny
weather, and see how the same paper will look in
cloudy wather and indoors. When the camera does not
adapt automatically, like the eye does, it shows re-
al differences. I think that "traditional photogra-
phy" does render colors quite accurately within a
certain range, because only a color-space transfor-
mation (including gamma-correction to match the
eye's brightness scale) and the setting of a white
point are required for natural-looking results.
Film photography (C41) also requires the removal of
film base color and sometimes a finer film-dependent
color correction (cf. film profiles).