Digital BW, The Print

> Hi!

Hi Igor.
 
> I'm about to buy a densitometer, probably an X-rite, for measuring my 
> B&W prints, and before I buy such an expensive -even when second hand-
>   device, I want to know for sure: is a COLOR densitometer suitable 
> for B&W?
> Or is it better to buy a densitometer that does only B&W?

Speaking as an X-Rite color densitometer owner, I'd say no. Save your money and get a black and white densitometer, or spend more money and get a spectrophotometer. I consider the classic CMY color densitometer to be an evolutionary dead end.

First, my particular X-rite (I'll dig the meter and model number up tomorrow) cannot combine its cyan, magenta, and yellow readings to give you a black and white reading, so I usually just read the magenta channel and call that "density". That gives you readings that are a little off, but not bad. To get an accurate reading, there's a little excel spreadsheet I can punch all three readings (C, M, Y) into, but that's a pain. Three readings for every single reading out. I've never gotten the computer interface working right, so it's a manual process.

Second, the color densitometer is rather limited. The color filters are narrowband, they do not have a spectral response resembling that of the human eye. This means that the densitometer suffers from observer metamerism. When used with process colors (colors made from a mix of cyan, magenta, and yellow dyes) it's not bad. It's only really accurate when used on conventionally printed color photographs. When used with colors produced any other way, colors that appear neutral gray to the human eye will not meter as neutral on the densitometer. The "problem" colors include anything produced by silver gelatin prints (especially if toned), any alternative process (cyanotype, platinum, etc), any "small gamut" ink set, guadtone or hextone process, or any of the new printers like the Epson R800 (cyan, magenta, yellow, red, blue), Canon 9900 (cyan, magenta, yellow, red, green). And it's always a bit off on Epson pigmented inks, even the older cyan, magenta, yellow sets in the 2200, 4000, 7600, 9600. The densitometer might tell you the gray is too magenta, while your eye tells you it's too cyan.

Something like the photospectrometer that Austin mentioned can measure 31 bands of color, instead of three, and put them together in any number of useful ways: true B&W density, process CMY density, or human vision correct LAB color coordinates.