>
> Then colorspace interpretation/translation is a non-linear process?
>
Absolutely. That's the entire purpose behind a colorspace. In an 8-bit per
color (24-bit) file, each color has a maximum of 256 values. As Moose so
accurately pointed out, as a general rule in the A-D process, the top stop
worth of light values take up half of the potential bits. This doesn't bode
well for editing of anything in the other stops. Therefore, what happens is
we rob bits from the top stop and spread them out for the rest of the stops.
How the bits are spread out is the magic. The aRGB colorspace assigns a
different number of bits to each stop as sRGB does.
Again, this is different than the original RAW file, which really has a
non-usable "colorspace" as it is essentially just voltage measurements. It's
meaningless to edit a RAW file as there is no standard there at all and each
and every camera stores the data with different meanings. Inotherwords, a
green channel measurement of 647 might be one color/brightness in one
camera, but the exact same color/brightness in another might be 989.
aRGB and sRGB are "working" colorspaces. sRGB does have the added benefit of
being the colorspace used by most electronic displays, though. Whenever a
printed output is involved, there is a conversion of the working colorspace
into the colorspace used by the output device itself.
If your output colorspace is sRGB and you have little serious color/contrast
adjustments to make, it is always best to work in sRGB and avoid
conversions. There are a few exceptions to this rule, but...
AG
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|