Digital BW, The Print

Stephen writes:

> Now, are you able to tell me the best way
> (either theoretically or empirically) of
> calculating the number of pixels that a
> printer would use to create an image, given
> a certain image size and printer resolution
> setting?

Theoretically it is difficult, because modern printers don't use simple
halftone screen rasters to create their images; some use stochastic
screening and dithering and other tricks to improve resolution (although it
still remains well below the machine-dot pitch).  Empirically, you can
fabricate a test pattern in Photoshop and then print it, and then examine
the pattern under a loupe to see just how well the printer can resolve
details.

Note also that the situation is different for those occasional
dye-sublimation printers out there.  Dye-subs use transparent inks that can
be overlaid one on top of another, and the darkness of each ink can be
controlled by varying how much is deposited.  As a result, they don't
require screening or dithering, and one machine dot = one pixel.  This is
why dye-sub can produce better images at 300 dpi than an inkjet at 2880 dpi.

I don't know if there is such a thing as dye-sub for black and white,
although it would be easy enough to create.