New BO workflow tests
2002-11-30 by Steven Karafyllakis
Hi All; Here are some further observations and tests (confusion?) on the BO workflow: 1) A quick note on dot size: if your head isn't aligned perfectly, leaving the `high speed' on will make your prints look much grittier- test, align it if it needs it, or turn it off. It makes no difference on the density/contrast of the print. Also, for those of you who can do 2880 (and have the proverbial patience of a saint), that setting produces a much smoother, almost grainless print. I would suggest trying it after getting the process working at 1440. 2) There are several ways to control output and the final print; these tests eliminate a couple so we can figure out what the others do. So the printer driver controls remain zeroed-out, printer gamma is left at 2.2, and PS density-contrast controls are not used to affect the output. The remaining controls I'm trying to understand are the `embedded profiles' (controllable from the `assign profiles' box in the `image>mode' menu) printer profiles, (controllable from the `print space' drop-down menu in the `color management' section of the PS print dialogue box) and the interaction of the two. 3) I went through a 15-print matrix: 3 with embedded profiles only and no printer profile, 3 with printer profiles but no embedded profiles, and 9 combinations of the two. In each case I used the dot- gain 10%, 15% and 20% profiles. First test sequence was with embedded profiles but no printer profile, and includes Clayton's settings, which as he said is near- wyswig, but in fact prints a bit light; I was hoping to get closer without having to tweak the image itself. So this is a summary of the results: Using embedded profiles only, going down from 20% to 15% and 10% LIGHTENS the image progressively about a half-stop each time. Conversely going up to 25% and 30% darkens it, though the effect seems a bit less pronounced, probably because in per-cent terms it is less of a change. Using the printer profiles and no embedded profile does exactly the opposite: lower numbers give a darker print. The combination of the two was fairly predictable: if you start with any given printer profile and go up the scale with the embedded numbers, the print gets darker. Start with any given embedded profile and go down the scale with the printer profiles, the print gets darker. And vice versa. In the end, when I tested a new image that had not been adjusted for printing BO, I remained frustrated with this group of numbers- the darkest combination, 10% printer and 20% embedded profiles, was still a little weak. I didn't get full wysiwyg until I interpolated a bit and tried a 10% printer and 25% embedded profile. That produced a print of good contrast & density that looked very much like my monitor image before assigning the 25% embedded profile-the monitor image changes. I do that step last, and then undo it to compare the final print, on the theory that since dot-gain 20% is the default pre-press standard, I should stay centered around it. Subsequent test prints indicate that most images will print well at the 10%printer/25% embedded setting-but images that have a lot of very low-key important detail need the slightly brighter 10%-20% setting. So, the settings that seem to work for me are as follows: Printer: Black only, 1440 DPI, Gamma 2.2, high speed off unless your head is aligned really well. Image in Greyscale, with an embedded profile of dot-gain 20% or 25%. Color management on, with a dot-gain 10% as a printer profile. The points to remember are as follows: Increasing the dot-gain% for the embedded profile darkens the image, decreasing it lightens it. Increasing dot-gain% in the printer profile does the opposite- lightens the image, while decreasing it darkens it. Also: changing the embedded profile is I believe, the same as doing a tone-density edit, so don't save an image you've done that to unless it's a copy, and don't change the same image more than once, start with a fresh copy. Your mileage may vary, but I hope this gives anyone interested a basis for finding a workable combination of settings. My next project is going to be trying to make a six ink printer behave like a four ink printer by replacing two of the inks with clear base stock. Does anybody have any experience/opinions/suggestions that might be helpful? Steve K