CHROMiX

Color Management Myths Question...

Was reading through some of the myths which spurred some questions which I’d love to hear anyone’s response to, byt especially Steve’s. :smiley:

With respect to Myth #29

Would you consider a RIP that has you perform a density linearization prior to profiling to be “very linear” with respect to the number of patches you would need to produce a “good” profile?

With respect to Myth #30

Would you limit ink in the profile, the RIP or potentially both, IF the limit was not IN ADDITION TO the profile?

Thanks for any insight you may provide.

Yes but regardless I will mostly use the ECI2000 target. Even if the rip is pretty basic I’ll still try harder to get a very neutral balance before building the profile. My attitude is that if the profile has to do less work then I’m going to get a better result.

I will always limit my inks in the RIP and then build my profile set to 400% Total ink. If I find that I have a little too much ink I can then simply rebuild the profile at 380 or whatever. If I find that I have to limit ink any further at the profile build stage then I’ve obviously done something wrong in earlier steps so will always then prefer to start again.

Scott,

Thanks for the reply and insight.

At 2:57 PM -0700 7/21/06, Doug Rhiner wrote:

sorry for the delay in my reply…

overall, yes. Although I would be careful about the term ‘linear’. Density is kind-of linear from a perceptual point of view but the calibration may or may not be actually linear (it may follow a predefined tone curve of some sort).

But yes, if I understand your question correctly, linearization will reduce the number of patches required to build a good profile… although lot should probably still be used to deal with overprint and hue-shifting issues.

probably just the RIP. Anything done in the profile will be in addition to the RIP, you don’t have any choice about that. Now if there were some reason to not fully limit in the RIP then you may be able to leave some up to the profile but I can’t think of a real-world case of this where it’s a better idea than just doing it in the RIP…

Also, profiles almost never give you the ability to ink limit on a per-channel basis. This is something you will need to do in the RIP.

Regards,

Steve


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


Post generated from email list

Linearization prior to profiling certainly helps. Density linearization is good, chroma linearization (that neutralizes a grayscale) is even better. Profiles shouldn’t be forced to correct for a poor linearization to begin with.

To be verbose, generic profiling targets perform best when the device’s behavior is consistent and perceptually linear. MonacoProfiler does have a custom target generation option (they call it linearization) that generates a custom profiling target for a devices irregularities. This option can generate superior profiles for devices that can’t be perfectly linearized prior to profiling.

Several years ago, RIPS that allowed for custom linearization allowed for significantly better quality than drivers that were inherently non-linear, perceptually. Recent drivers have really reduced this advantage.

Limiting in the RIP is the way to go. Per channel ink limiting comes first, then composite (TIL) ink limiting. If these are done right there shouldn’t be any reason to manually limit further in the profile, aside from what will be done automatically.

Cheers,

Scott Martin

The problem is… were do you limit? 400%? 360%? 340%? 300%?
I did a limit test and looked for bleeding etc etc, what I find out was that 270% would be a good choice. I did a profile with 270% then I did one without using the limit in the RIP and used 400% in the profile. Comparing the two for delta E, To my suprise this profile was much better then the one with 270%!!!

Would it not be better to evaluate the profiles first before deciding what TIL you should use? If you do many profiles with different settings and compare them with ColorThink and Delta E. Pick the best ones and do more tests then you will have the TIL there. Put the TIL in the RIP from the result and you are set?

Anything wrong with this workflow?

Soren

Generally, the ink limit needs to be set in some way before the profiling target is printed. In general, I almost always go set up a printer in this order:

1 Print a linearization target, read the target, set individual ink limits.
2 Print the linearization again with the ink limits set then linearize.
3 Print an ink limit test and set the ink limit.
4 Print the Profiling target and create the profile.

You can also generate a custom target from your application with a predefined ink limit and skip the in rip ink limiting. Any ink limiting you do after profile creating will effect the quality of your printing since the ink limits are not accounted for in the profiling process.

Question 1: We linearize all our presses based on both density and a gray scale ramp prior to profiling them.

Question 2: We limit ink in the profile. The main reason is because our color management consultant didn’t know the nuts and bolts of this somewhat proprietary RIP, but understood Monaco Profiler. He opted to go with what he knew, so we limit in the RIP for no better reason than that.

We print on pressure sensitive vinyl (decal material) so your mileage may vary. However, the following procedure works wonderfully for us.

We use our trusty old X-Rite 404 densitometer to decide total ink limit and we set it in the profile after we put it through some testing.

We use a test file created for our conditions by our color management consultant.
We simply print it, wait 24 hours for the solvents to outgas, then take density readings of all the patches.
On each patch, the density values increase until they stabilize or decrease. We take the patch with the highest density and use that as our TIL.

Too much ink on the substrate will diffuse the reflection and the densitiy values can actually decrease. That is where we set our limit.

Our ink limit is a 300% and we set it in the actual profile. Our RIP doesn’t have express control over each individual color in that regard. The profile is the best place for us to do it based on the RIP we use.