Nick Makris (nick@mcn.org)
Fri, 30 Jul 1999 07:45:57 -0700
The limits of the printing the grayscale have been discussed more than once
on this forum. And even some of the responses to my notes on Dan's book
indicate that they are hard put to discern any appreciable differences when
printing on a 1200-1440 DPI printer. I recall that on one occasion,
(forgive me, I don't remember who said it) someone made the point about not
needing 256 colors of gray to successfully print a grayscale image.
In my recent testing/investigation I have also been using a book called
"Real World Scanning and Halftones" by Blatner, Fleishman & Roth (2nd
edition) Peach Pit Press. I highly recommend this text - it is very
comprehensive and among its other attributes, it does a wonderful job of
analogizing halftone vs stochastic processes. The following formula (a
derivative of one found on page 218 of the text), begins to provide insight
to the idea that the max printer resolution as output on any piece (say
1440dpi) is limited by the combined use of number of grays (say 130 out of
possible 256) and the number of lines chosen for the screened output (say
130): Actual print output in dpi=(Screened output in LPI) X square
root(Number of Grayscale colors to be printed)
By deduction, I have assumed that if the number derived is less than that of
the printer max capability, the effective resolution will be the number
derived above and conversely the efective resolution will be the max printer
resolution. This may explain the complaints about lack of discernable
difference in the use of these new hi-res printers.
So, the moral of the story is use fewer colors and fewer lines in the above
formula to force the desired printer output. In the above formula
130Xsquare root(130)=1482 and therefore the max printer resolution (1440)
will prevail. In my reading, I have found that the numbers used for colors
(130) and LPI (130) above are acceptable for many outputs. Test, Test,
Test!!!!!
As with all investigation, some questions remain unanswered. In this case,
I don't understand what the association between LPI (as in screened output
or halftones) and that of stochastic output which supposedly has no defined
lines.
If you have any answers, comments or corrections, please post.
Many thanks,
Nick
This archive was generated by hypermail 2.0b3 on Thu Oct 28 1999 - 21:40:39