Doug
In this reference images from three different sources all look slightly dark. While the way it appears is subjective and a function of comparison with the monitor or viewing light level, the ability to resolve tonal values is not subjective (or not very subjective). The print is slightly darker - or at least I am unable to resolve the darkest 2-3 cells on a 20 or 21 cell chart. I don't have a reference print from another printer to compare. Can you guys resolve a grayscale chart with a range of 20-21 cells?
I'm trouble shooting that to figure out what adjustment makes sense. I don't think I have an incorrect setting but someone here might have an idea.
Now maybe I am being too picky, so Mark's comment about the inability of the printer to resolve the bottom end makes sense. But overall, I'd like to see the images all a little brighter across all tonal values.
There are two issues you bring up. The first, the ability to discern the darkest patches in a 20 patch neutral set, is largely a function of the paper and ink combined with how well illuminated a print is. You need to have about 300 to 500 lux on the print to get similar results to what you see on a monitor and prints almost always have a lower DMax than a monitor which means the darker patches will appear less discernible on a print than on a monitor.
However, the second point, that the prints appear darker across the tonal curve than the image on the monitor, is a pretty common perception. Even when viewing prints under good illumination. If you print with a white border then you should set your background color in Photoshop to white. This will make the monitor image appear a bit darker and should be a closer match to a print.
But here's the thing. Human vision is highly subjective. Even viewing a reference print is subjective. They can vary quite a bit before they look wrong unless you have a known good reference print to compare them to. And there is another problem with many of the reference images out there on the web. Images with colors that are out of gamut will have their colors, when printed, shifted to colors that are in gamut. They do get printed after all. And different profiles, printers, and paper can shift these colors differently.
My choice of a reference image is the old Kodak "Photodisc" which can be found, usually with a few alterations, on the web as PDI-Target-DCP.jpg. The reason I like that image is that the vast majority of it is printable and in gamut. So if you print it using Relative Colorimetric Intent, you will get a print that will look the same on any printer or glossy/pearl/luster paper if you have a good, color managed system. And what it looks like on a monitor is completely irrelevant.
Once you have a known good reference print, save it, and use it to check new printer paper combos.
The problem people have is getting that first known good print in the first place. And that's why I recommend printing a reference Colorchecker image. Everyone (almost) has an actual colorchecker and can use the colorchecker as a kind of reference image. It's limited, as Mark points out, in that the lowest L value is about 20. But if you are seeing darkened prints over the tonal curve then a colorchecker image print will tell you right away if it is an illusion or from poor lighting. It's a way to verify the printer side of color management. The colors are in gamut or extremely close for virtually any printer/paper and this means that the printed images will look the same since out of gamut colors do not have to be shifted into those that are printable.
Just remember to print it using Abs. Col. and in the same size as your physical colorchecker. Then trim the paper's white edges so that it will look as close as possible to an actual one. The only differences should be those related to surface characteristics. CC's have a matte finish. But you can avoid the reflection from non-matte prints by slightly tilting them from the light source. The other issue is OBAs. CCs don't have any response at all to uV but paper can. They can cause a print to look slightly more bluish in daylight if using a profile made with M2 (uV cut) or, oppositely, slightly red/orange if the profile was made using M0 or M1 and the print is viewed with low or no uV content light. To minimize this print a reference image on paper with no added OBAs.
