OK, then let's get down to the specifics and see what we need to avoid
the horrors of spatial aliasing.
Do we need to have optically pure, non-resampled, non-interpolated data
of at least 720 ppi, and then printed at input source of 720 dpi to
avoid this defect? What kind an image demands this? How commonplace is
this type of image feature? At what viewing distance is it visible?
Can we develop some reasonable rules to follow, such as: when printing
highly contrasted monochromic images, spatial aliasing becomes severe
enough to warrant using XXX resolution with Epson printers to avoid
aliasing visible at 12" from the print on an A4 print?
Surely not all image features demand this type of treatment to avoid
perceptible spatial aliasing? Does a complex color images with low
contrast require this? Hell, do the vast majority of everyday images
require it?
Personally, I think overuse or underuse of USM (unsharp masking) is a
much more real problem in digital printing than images requiring 720
ppi/dpi source and input data to avoid spatial aliasing.
I am not suggesting that the theory doesn't sometimes manifest itself in
degrading the result if it has not been followed through upon, but that
the circumstances where such exacting image quality is required to make
a print that looks good to the naked eye are statistically small
compared to a plethora of other problems the majority of people
encounter in printing digitally, from clogged to misaligned print heads,
to using wrong paper types, to processing the image incorrectly in the
image processing software, to problems introduced in scanning, to
problems introduced by digicams image sensors or during image
compression or storage.
My concern is that the impression may be left that anything below 720
dpi/ppi pure data will make a bad print, and that certainly isn't true.
That is why I ask for clarification as to under what extreme conditions
is this extra "cost" of image processing worthwhile?
Art
