With today's resources I could see how a pure 16-bit path makes sense, but also perhaps why other replies - people with the actual printer unlike me! - report they didn't see big quality jumps. That brings back bad memories of 4GByte addressing limits in wide format with 32-bit drivers. So a 16-bit buffer could be problematic then, particularly with wide format. That was an age of 8-bit integer graphics and lower h/w resources of course. I seem to recall Graham Gill, the developer of Argyll, came to similar conclusions.

Interestingly though 8-bit input was sufficient. Things like dot gain mean you need the extra precision. By processing I mean internal processing within a specific stage like multiple intensity processing. Greater than 8-bit processing is definitely required for optimal printing. Several years ago I did some investigations into 8 vs 16-bit printing at the various stages of the print pipeline. At least I hope it was a technical decision and not pure marketing. It is good to see someone doing that and even more impressively that Canon agreed it going into production. The XPS print model might have various advantages be it, high bit depth as they report for the driver, or something like better colour management tagging. I assume someone in Canon was creatively assessing the print models in Windows and seeing where improvements could be made.

I have no specific experience with that particular driver and printer to share but in your position I would definitely try it. Other than the added bit depth, what features and capabilities do I gain - or lose - by moving to the XPS driver? What is the argument for using, or not using, the XPS driver? I've Googled around on this, but haven't found any info that I'd consider authoratative.

Canon office xps class driver download