Canon vs TurboPrint ICC Files

[riclin]

A query regarding possible differences between genuine Canon ICC profiles and those generated locally with TurboPrint, ArgyllCMS and a ColorMunki....

I recently discovered that using genuine Canon ICC profiles as Argyll preconditioning profiles results in significant numbers of read failures with the Argyll chartread tool, which reported ΔE read errors of as much as 160 that no amount of strip re-reading would correct. I have theorized that the Canon and TurboPrint drivers may map their colour generation using different ink allocations and that the Canon profiles are therefore incompatible with the TurboPrint colour logic (and v.v. - profiles generated within the TurboPrint ecosystem would not be suitable in a generic Canon environment.) Rather than "guiding" Argyll, the Canon preconditioning ICC may instead be conflicting with the TurboPrint colour logic. Replacing the Canon profile with local TurboPrint generated profiles for preconditioning appears to completely resolve the issue.

Does this make sense or should I look elsewhere? I'm not overly concerned since as I generate new profiles, I'm also creating a library of similar ICCs that seem to work seamlessly as preconditioning files for future profiles. But I am curious as to whether my theory is correct, or whether there is some deeper issue that requires further investigation?

TIA,

Richard (riclin)

[zedonet]

Hello. You are right, the high ΔE are probably caused by the fact that ICC profiles cannot be interchanged between the genuine Canon driver and TurboPrint. The reason is that color reproduction is dependent not only on paper and ink but also on the driver / RIP software: different printer drivers have different color output, depending on halftoning algorithm, linearisation, use of light (photo) inks vs. regular inks, black generation etc.

I suggest the following procedure

1. create an ICC profile without preconditioning
2. use this ICC profile as preconditioning for future profiles

For best results, separate preconditioning profiles are recommended for the paper types "Plain paper", "Matte coated inkjet paper" and "Glossy / Semiglossy paper".

[zedonet]

PS: what is your Canon printer model?

[riclin]

Thanks for your reply - you have confirmed my suspicions. This particular profile run is for a Canon IP8700 printer using Canon GI-63 (or regional equivalent) inks. I have prepared 920 patch profiles without preconditioning for Gloss, Satin, Matte and FineArt paper types and intend using these as preconditioning profiles for more general 460 patch profiling runs with other papers. A similar workflow will be used when generating profiles for my Pro-10S with Octoink ink - potentially omitting the gloss papertype.

Sound reasonable?

Thanks again for your reply,

Richard Lindner (riclin)

[zedonet]

Yes, this sounds good.