Hi Neil & Mark,
Thanks both of you for your comments & suggestions.
>It's probably more reasonable in your case to stick with Epson, upgrade your rip to the LFP version, and continue using the workflow you know. The only change would be now with 16 bit drivers you may not need the rip any longer, and replace it with a PC and Qimage.
Very sensible suggestion to stick with a familiar workflow, thanks. But sorry, what do you mean by the "LFP version" of ColorBurst RIP?
Also, not to get off-topic, but FWIW I would respectfully challenge the assertion that a 16-bit printer driver obviates the need for a RIP. I've found that going through the tedious process of "linearizing" the printer in ColorBurst before creating a custom profile makes an enormous difference in shadow separation.
Maybe in theory the extra resolution of the 16-bit driver can improve print quality. But consider that an 8-bit driver is supposed to be able to generate 256 distinct ink levels per cartridge (if I understand correctly). Realistically, of course, we users often get suboptimal print results with 8-bit drivers, when printing images with only 10 or 20 steps in a particular ink primary (e.g. using a classic "ramp" test image). A priori, I don't see how being able to generate 65K distinct ink levels is going to solve this problem.
My experience is that a *lot* of improvement can be made in the 8-bit regime, using a good RIP & a spectrophotometer. I wouldn't forgo using a 16-bit driver (currently I don't believe they're available from Epson for Windows), but I'll need to see the results of extensive, rigorous testing & comparisons of the two paths before I believe that 16-bit drivers are a quantum leap forward.
Back on topic now...
Thanks for reminding me of the 4000 / 4800 scandale. I appreciate your suggestion to look around for a rental lab, but to date I haven't found any in Chicago
Also, I should say that I'm not really in a one-off situation with this commission (as my original posting may have suggested) b/c I also have a major show coming up next April, which I want to print in larger-format for as well.
Based on Neil's suggestion, it feels like the "safest" path for me, in the sense of minimizing the unknowns, risks, and changes, may be to go with the Epson 9880. I would definitely love to have that wider set of inks in the Z3100, but OTOH I don't need the built-in spectrophotometer. I'll be sorry to be limited to PK ink with the Epson -- that still really irks me -- but I guess I'll have to accept the evil.