[filmscanners] Re: understanding contaminated blacks--SS4000--newbietype question

["Anthony Atkielski" <anthony@atkielski.com>]

Paul writes:

> If this is merely a variation in the black
> point for the three colors, it should be
> possible to compensate for, more or less, by
> setting the black point independently for
> the R, G and B channels in either Levels or Curves.

There's a pitfall, though:  Whenever you set this points independently for
each primary, you shift the color balance of the entire image.  Sometimes
it's practically impossible to get the color balance right in the image
thereafter--shifting for midtones knocks off the shadows, fixing the shadows
knocks the highlights out of balance, and so on.  I usually set black and
white points for all colors at once for this reason, although with Levels I
set for each color independently.  I've not determined which of these gives
better results; it seems to depend on the image content.  Levels works well
for images with fairly even distribution of light throughout the range of
the film; curves seems to work better for contrasty images, like night
shots.  Sometimes I use both, or I look at the histogram and tweak the
curves to drop resolution from the zones of luminosity that are not used.
For example, in any image with strong highlights and shadows but little
content in the midtones of importance, you can scrunch the midtones and
expand the highlights and shadows to take their place with good results.



----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body