> It's also trivially easy to prove that the image is resampled at 360ppi or
> 730ppi, using a "nearest neighbor" algorithm, by constructing a test image
> of alternating 1-pixel black and white lines, and printing it at various
> resolutions. You can do it yourself in ten minutes.
>
I posted a file (
2288 kB)
that checks the print driver's resolution. The pdf file contains the same
test image repeated many times at various resolution from 120 ppi up to 1440
ppi. The test pattern contains variable contrast, crossed, sinusoidal bars
with period of either 2 or 4 pixels. Around the outer edge of each pattern
the bars are in one direction only and at 100% contrast. These outer bars
on each pattern are extremely sensitive to resampling artifacts. The second
page contains a similar pattern, but with periods of 8 and 16 pixels per
cycle. This second page was added mainly so that the form of the test image
would be more obvious.
If you print it out, be sure to UNcheck (under Acrobat reader options) both
"Enlarge Small Pages to Paper Size" and "Shrink Oversized Pages to Paper
Size. Look for uneven teeth around each image. Also look at how the low
contrast grid pattern merges into gray.
On my system (MacOS X 10.2.1, standard Epson print drivers) and my printer
(Epson 2100/2200 or Epson C80) it is obvious that only ppi's that are evenly
divisible into 720 print without being resampled. Contrary to what has been
reported here, the image at 720 ppi does not print correctly. Only 360,
240, 180, (no 144 ppi image) and 120 ppi print correctly. In addition, I
saw very little difference between printing at 240 and 360 ppi. The HP
postscript laser printer does not print any of the targets without artifacts
- it use a halftone screen that does not match the ppi of any of the images.
For real images, I see little, if any, difference once the resolution
reaches 240 ppi and I have never seen anything attributable to the
resampling.
> --------
> Dean Shough
> dean.shough@lmco.com
>
----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or
body
