>> Horses for courses indeed. But I'm just a knackered old donkey
Stephen, while you are the thoroughbred in a dust cloud on the
horizon. I'm sure that your way is better but CMP is 'good enough'
especially when you see what the mag repro does to a file. And it
saves me the hard climb up that interminable learning curve. <<

David, thank you for the vote of confidence <g>. As an aid in the learning
curve - here are some quick pointers for you to play with if you would like
to explore my viewpoint:

My basic steps are:

* Appropriate assigned colour space so that the caste is not enhanced, even
if this is a white lie about the true colour source (sRGB instead of A98 for
example)

i) Ensure the image has neutral white/black/neutral points if appropriate

ii) Measure colours/tones throughout the tonal range to see what/where a
caste is

iii) Once the problem is known, then corrective measures are taken

* If there is one severe colour caste problem throughout the image, sample
the colour caste colour and make a new layer. Fill new layer with caste
colour. Invert the colour in the layer. Set layer to colour blend mode.
Adjust opacity to suit, starting at 50% although lesser is often common.

* Most can be corrected via a curves adjustment layer set to colour blend
mode in Photoshop.

* Less experienced users may prefer colour balance with preserve luminosity.

* Really tough problems are corrected via AB curves in LAB. A dupe is made,
converted to LAB, edited, returned to RGB then blended in colour blend mode
over the original RGB file so that the original luminosity component does
not go through an unwanted mode conversion.

You can also get in colour blend edits via apply image or channel mixer as
well. All of the above often get the same approx. result - although there
are some subtle differences between them all.

As well as the calibrated/profiled monitor in controlled viewing conditions,
I like to use RGB, LAB or CMYK colour sampler and mouse-over readings to
help establish colour values when evaluating and editing known or presumed
colours.

Regards,

Stephen Marsh.

===============================================================
GO TO http://www.prodig.org for ~ GUIDELINES ~ un/SUBSCRIBING ~ ITEMS for SALE

Reply via email to