CHROMiX

variation on "old color" issue

Our printer (Epson with an Onyx RIP) was profiled and behaving nicely. After several months we noticed that the color was “off” (ie the files made when we first set up the system are now printing redder). Apparently this happened a few weeks ago and since then (unaware of the problem) we have been making all our files a little more cyan to correct for this. Now that we know about the problem and know the solution (re-linearizing gets us back to the original color) we have the dilemma of what to do with all those files made during the “off” color period.

Worst case scenario we could open all the files (they are all RGB tiffs) and manually apply curves in photoshop and hope for the best.

But I am sure there must be some way to create a profile of the “off” color, apply it to all the “off” files, and then convert back to “good” color. I just cant figure out he right way to do this. Any suggestions?

thanks,
Kate

The “right” way is to adjust the color of your tifs to output correctly with the correct calibrations… I can see where this could be a challenge!

I think it would be possible to embed a source profile which would describe the image in such a way that it would be “corrected” to a normal state… assuming your workflow retains profile data and does not strip that information out. Alternatively you could use that profile to permanently alter your tifs, essentialy automating what you could do by hand.

How you would obtain that profile and how you would measure its accuracy is a mystery to me, however! I suppose you could purchase a profile editing package and make manual adjustments until a satisfactory output is acheived. That kind of trial and error is often futile, however.

A resolution to this issue would interest many, I’m sure. This is a challenge every printer faces at some point. I hope one of the greater minds will chime in here.

At 5:29 AM -0800 11/28/05, kateowp wrote:

Our printer (Epson with an Onyx RIP) was profiled and behaving nicely. After several months we noticed that the color was “off” (ie the files made when we first set up the system are now printing redder). Apparently this happened a few weeks ago and since then (unaware of the problem) we have been making all our files a little more cyan to correct for this. Now that we know about the problem and know the solution (re-linearizing gets us back to the original color) we have the dilemma of what to do with all those files made during the “off” color period.

Worst case scenario we could open all the files (they are all RGB tiffs) and manually apply curves in photoshop and hope for the best.

But I am sure there must be some way to create a profile of the “off” color, apply it to all the “off” files, and then convert back to “good” color. I just cant figure out he right way to do this. Any suggestions?

Hi Kate,

I’ve dealt with a similar issue before. Let’s see here…

  • first, the file needs to be in printer-space. So you need to convert the file as if you were going to print it.
  • if the files looked good while the printer was printing poorly then they are in ‘bad-printerCMYK’. An important note here is that the files were converted originally to ‘good-printerCMYK’ but you were printing on a bad printer so they didn’t look right and you edited them to look right. That’s what put them into BPcmyk.
  • If you printed a profiling target at the time of bad printer behavior (or can return it to that set of curves) then you could capture what BPcmyk was.
  • then you’d simply assign the BPcmyk profile to the image and convert it to RGB and then back out to good-printerCMYK or just directly to GPcmyk.

If you didn’t print a target at the time of bad printer behavior and can’t return it to that state then I think you are in correct-by-hand land.

Looking this over I see another scenario. That you edited the files in RGB working space prior to printing. That should be OK as long as you have the BPcmyk profile. You would convert the file to good-printerCMYK first, then assign BPcmyk, then convert back. much like above.

does this make sense?

Regards,

Steve


o Steve Upton CHROMiX www.chromix.com
o (hueman) 866.CHROMiX


Post generated from email list

thanks steve! perfect sense!
kate