From: Keith Gerling (kgerling@ameritech.net)
Date: 11/14/01-08:07:45 AM Z
Judy says:
>I was doing this in Photoshop 4 ... I tried both UCR & GCR, also another
>where you made yr own curve, then made VERY careful (for gum) tests
>printing the seps. No correlation in the print to what was "programmed,"
>or none I could see.
If this is the case, then something is seriously amiss. Changes between UCR
and GCR should be profound.
I've not used 4.0, but in 5.0 and 5.5 (and I have to think in 4.0) the
changes in CMYK color setting are
only invoked when the mode is changed. For instance, the way I do this is
1) after arriving at whatever I want
to use as a finished work, convert the image to LAB or RGB. 2) Make several
duplicates (Image/Duplicate) 3. Make a change to color settings (Edit/Color
settings) 4. Convert an image to CMYK 5) Repeat 3 and 4 until I arrive at a
satisfactory separation. The gist of this is: you might not have even
changed your color settings AT ALL, unless you did a fresh convert to CMYK
after every alteration. Converting from CMYK to CMYK will NOT work.
>So far I THINK best control is in curves -- higher contrast neg for black
>neg... but this is NOT a separation per se, it's for serial differential
>printing in whatever color(s).
Apples and oranges. Curves are essential, but are not the best tools to use
in altering color separation.
Another trick (while I'm on the subject): Swap black channels from
different color spaces. Use the procedure I describe above where several
different CMYK images are produced, and "steal" the black channel from a
Light GCR and put it into a UCR image. Then, if you want more or less black
Use the Calculation function (Image/Calculations) to effect the black
channel based upon input from another black (or even better) from a color
channel.
Keith
This archive was generated by hypermail 2b30 : 12/10/01-11:12:21 AM Z CST