This is a follow-up to my prior post/complaint (both here and on the Photopxl forum) about Epson not supporting its newly released Epson Legacy Baryta II paper with profiles for many Epson printers still commonly in use. Taking the suggestions of Andrew Rodney, Mark Segal, and Kevin Raber, I trialed several profiles made for the P800 and other baryta papers. Specifically, I used the profiles for the original Epson Legacy Baryta (oELB), Canson Baryta Prestige II (CBPres2), and Canson Baryta Photographique II (CBQue2). I printed the widely known Atkinson test page and a synthetic grayscale step tablet. This paper still has limited availability; my box was sourced from
www.ITSupplies.com, located near Chicago, Illinois, US.
Viewing the Atkinson test pages under 5000K LED lamps showed no large differences among the three profiles. All three profiles produced very acceptable prints. The prints left the impression that any one of them could be "tuned" in Photoshop to match either of the others without undue difficulty. I did note specifically that there was no blue-magenta crossover in the arch and mud ripple images; with all three profiles the sky within the rock arch moved smoothly from lighter to darker blue without acquiring a magenta cast, and the water in the mud ripples reflected only magenta, and never became blue.
Three very small visual differences among the profiles were noted on the Atkinson test page:
1. contrast in the lava and in the aspen forest was slightly different among the three profiles: oELB > CBPres2 = CBQue2 for both scenes;
2. midtones generally printed very slightly lower with the CBQue2 profile than the other two;
3. while there were a few spots where colors appeared very slightly more saturated with the CBQue2 profile, this could not be separated from the fact that these values seemed to be printed very slightly darker as well.
The grayscale step tablet was scanned (Epson V750, Silverfast ai8, Lab colorspace) and measured in Photoshop. The measured L values from the print were plotted against the step tablet L values (0 to 100 by 5) and the plots of the profiles were compared. The plots were consistent with the small differences seen in the Atkinson test pages. The measured L maximum and minimum did not vary significantly among the three profiles. The CBPres2 plot was the most linear with a slight rise above linear beginning below the L35 step and curving gently down to dmax. The CBQue2 plot was very similar but the L values measured were about 2 units lower than CBPres2 between about the L30 and L60 steps; this explains the slightly darker midtones noted in the Atkinson test page. The oELB plot demonstrated a mild midtone contrast increase, centered at the L50 step; it was about 2.5 units above linear from the L60 to L85 steps, and about 1.5 units below linear from the L40 to L30 steps; this explains the very slightly higher contrast noted in the lava and aspen forest scenes on the Atkinson test page.
Summary:
1. In lieu of a printer-specific manufacturer profile any of these three profiles could generally be an adequate substitute, as far as can be determined from the information above (at least for a P800).
2. Of these three, the slightly preferred choice might be the Canson Baryta Prestige 2 (CBPres2) profile due to its general linearity through the mid- and lighter-tones.
3. Nothing about this limited evaluation adresses the gamut limits of these profiles on this paper. Caveat emptor.
Obviously, the measurements were done with "prosumer" level scanning equipment and not a direct read colorimeter. That said, the results are internally consistent and did correlate well with what was seen in the Atkinson page prints, leading me to think they are at least generally correct as far as they go.
My thanks to Andrew Rodney, Mark Segal, and Kevin Raber for their very helpful suggestions to try these alternative profiles.
Mark Gunion
