On Thu, 17 Apr 2008 10:51:56 -0700, C J Campbell
wrote:
: I am well aware that there are plenty of software filters out there and
: I use them. I think Nikon's filter plug-ins for Capture NX are even
: better than these guys. However, software cannot always replicate the
: effect of real filters, especially when using spot gels or mixed
: lighting.
:
: Neither is it possible to correct white balance in software and get the
: same results as getting white balance right in the first place. You can
: sometimes get pretty close, but it is definitely not the same.

I just don't believe that. You'd have to prove it by reliably identifying
software vs hardware correction in a statistically significant sample of
pictures you didn't take.

: There is a huge difference between approximating an effect in software
: and nailing it in the original image. I know there are photographers
: who think they are artists and that they can 'feel' the color
: temperature, or that they can walk into a room and tell you what the
: white balance should be. I am not one of them. Furthermore, I think
: that guys who claim they can 'feel' the color temperature are deluded.
: I have never seen one of them who actually could get the white balance
: right in tricky lighting situations, or even in an ordinary office with
: fluorescent ceiling panels.

But here I think you're absolutely right. No one can correctly assess ambient
color temperature unaided. The WB correction of the human eye, which cannot be
effectively defeated, is just too good to allow it. When a photographer walks
into a room, his perception of the color temperature depends strongly on the
color temperature of the room he just walked out of. And the longer he stays
in the room, the less reliable his perception becomes.

Fluorescents are particularly tricky, probably because their "white" light is
composed of relatively few discrete frequencies. The very concept of "color
temperature" breaks down in that environment. To get the WB exactly right in
software, you'd have to know the frequencies and their relative strengths. But
to do it with a filter may well be impossible.

Bob