Re: MeasureTool Compare question
Re: MeasureTool Compare question
- Subject: Re: MeasureTool Compare question
- From: Marco Ugolini <email@hidden>
- Date: Wed, 16 Mar 2005 00:36:56 -0800
In a message on Tue, 15 Mar 2005 12:45:28, Matt Nelson wrote:
> I think you will find this answer to your question in the reply from Mr.
> Orpen a few messages up.
That is a clever script: I tried it and it works, but I think that I have an
alternative procedure which, compared to this one, is definitely less
time-consuming.
You be the judge if I'm on to something here:
1) In ColorLab (the free unsupported application from Gretag Macbeth,
available at <http://www.gretagmacbeth.com/mac/downloadcenter.htm?id=8667>),
open the ECI2002CMYK reference text file from the "File" -> "Open special"
submenu. (Mind you: open the REFERENCE text file, not the TIFF image file.)
To be able to do that, though, you must first place an alias of that
reference file in the "OpenSpecial" folder inside the ColorLab 2.8
application folder in the Finder, since ColorLab does not have the ECI2002
reference text file available in its default state. (Also, remember that the
procedure I am describing also applies to ANY other reference text file, not
just the ECI2002.)
2) Still in ColorLab, with the reference file open, go to the Filter ->
Conversion -> Testchart Define submenu. In the following "Define Strips"
dialog box, go to the "Number Of Patches Per Chart" section, and write in
"20" in the box right below "Horizontal" and "89" in the one below
"Vertical." Leave EVERYTHING ELSE untouched. Click the "OK" button.
3) A new window appears, titled "Target TestChart." You will see that at the
upper left corner it says "CMYK / 1780 Patches (the exact number of samples
in the ECI2002 chart). That means that you have the correct and complete
chart.
4) Now you will go ahead and save this custom test chart BOTH as a REFERENCE
text file and an IMAGE file:
4a) Go to the "File" menu, select "Export...", select "Text document" from
the "Format" pull-down menu, and save the reference text file for this chart
(e.g., "ECI2002.txt").
4b) Go to the "File" menu and select "Export..." again, but this time select
"TIFF images" from the "Format" pull-down menu: this will be your test chart
image in TIFF format, with the values in each sample corresponding exactly
to those in the reference text file you just saved right before it. Name it
and save it (e.g., "ECI2002.tif").
5) Open the TIFF test chart in Photoshop, and you will see that the file is
20 pixels across by 89 down: each pixel is a color sample, with no gaps, for
a total of 1,780 pixels. (This is perfectly appropriate for the color
transform to Lab values. If you so choose, you can save and keep this pair
of reference-file-cum-image-file as a starting point from now on for all
your round-trip CMYK printer profile tests.)
6) Still in Photoshop, apply to this chart the CMYK printer profile you wish
to test, then convert to Lab using the Absolute Colorimetric intent, then
save the converted image in TIFF format (you can name it "ECI2002_Lab.tif",
for example). Do not embed a profile.
7) Go back to ColorLab, and open the converted Lab image you just created in
Photoshop (File -> Open...). Once it is open in ColorLab, go to the Filter
-> Conversion -> Testchart Define submenu. In the "Define Strips" dialog
box, again like you did previously, go to the "Number Of Patches Per Chart"
section, and write in "20" in the box below "Horizontal" and "89" in the one
below "Vertical." Again, leave EVERYTHING ELSE untouched. Click the "OK"
button.
8) All that is left to do now is export the list of Lab values in this
chart. Go to the "File" menu, select "Export...", select "Text document"
from the "Format" pull-down menu, and save the reference text file for this
chart built from the converted Lab TIFF image (you can name it something
like "ECI2002_Lab.txt"). This is your "measurement file 2", with no need for
either tedious manual samplings or scripts, and, seemingly, just as precise
(actually, more precise, since you get decimals, for whatever that is worth,
whereas Photoshop rounds up to the nearest full integer).
Anybody please let me know if you see any mistakes or problems in this
procedure.
Thank you.
Marco Ugolini
Mill Valley, CA
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden