squares and parallel lines

  • Thread starter Thread starter Dennis Herrick
  • Start date Start date
D

Dennis Herrick

I have a Mustek A3 scanner and have been using it for sometime.... One
problem I have is that if scan something like music, I can get the staff
lines perfectly straight, but then the beginnings of the lines are not
square. Or if I scan a record jacket, it's wider at the top than the
bottom -- the corners aren't square. Squares just don't come out square.

As far as I can tell, there isn't an adjustment for this -- I guess it's
just a property of the scanner itself. Is this correct?
 
Dennis Herrick said:
I have a Mustek A3 scanner and have been using it for sometime.... One
problem I have is that if scan something like music, I can get the staff
lines perfectly straight, but then the beginnings of the lines are not
square. Or if I scan a record jacket, it's wider at the top than the
bottom -- the corners aren't square. Squares just don't come out square.

As far as I can tell, there isn't an adjustment for this -- I guess it's
just a property of the scanner itself. Is this correct?
Click to expand...

I would say it is not a property of the scanner. If your scanner is making
trapezoid scans then something is wrong with the scanner.

I have a Mustek A3 scanner and it produces square scans.

I drew a crosshatch pattern on a 8 1/2 x 11 lined yellow pad, I spaced black
lines top to bottom one inch apart. Then I scanned that pad at 200 dpi, I
got an image that was perfectly square. The one inch spaced lines were just
as parallel as I drew them. The rules on the yellow pad, was also parallel.

If the art is not flat on the glass you could get a wedge shape. In the case
of a LP record album cover, it won't lay flat on the glass because the LP
cover is 12 1/4" x 12 1/4". It is wider that the 11.7" width of the scanners
glass.

I have not experienced a great amount of image distortion on an LP cover.
The one that I scanned comes out pretty darn square.
 
CSM1 said:
I have a Mustek A3 scanner and it produces square scans.
Click to expand...
[snip]

Same here.
My recently bought ScanExpress A3 is working fine (for line-art).
For full-colors the preview image does not match the final scan result,
but that's another problem.

However... my HP 7450 manages to 'shear' every scan.
On a A4-sizes 200 dpi scan, the bottom righthand side of the paper is
about 5-10 pixels lower (then the bottom lefthand side).
This also results in all horizontal lines being slightly diagonal.
All this is probably due to wear & tear on the internal mechanism.
(5+ years, 10.000+ scans (excluding previews)).

So... does anybody know some freeware 'de-shear' tool
for use with Windows 2000 (or XP) ?

Most programs only seem to be able to rotate, but using that,
it (obviously) messes up the vertical lines of the scan...

TIA,

E.
 
I have a Mustek A3 scanner and it produces square scans.

I drew a crosshatch pattern on a 8 1/2 x 11 lined yellow pad, I spaced
black lines top to bottom one inch apart. Then I scanned that pad at 200
dpi, I got an image that was perfectly square. The one inch spaced lines
were just as parallel as I drew them. The rules on the yellow pad, was
also parallel.

If the art is not flat on the glass you could get a wedge shape. In the
case of a LP record album cover, it won't lay flat on the glass because
the LP cover is 12 1/4" x 12 1/4". It is wider that the 11.7" width of the
scanners glass.
Click to expand...

I hadn't thought about it not being flat, but this happens with everything,
even single sheet music that does lay flat....
 
I hadn't thought about it not being flat, but this happens with everything,
even single sheet music that does lay flat....
Click to expand...

Is this distortion viewed on the monitor screen, or after printing on paper?
CRT monitors often have a trapezoid adjustment which can warp the screen view
that way. Printing it would eliminate this factor.

In the scanner, carriage guide rods can be mounted off square, so that the
carriage moves slightly sideways as it goes down the page. They are built to
snap together right however.

Wider at one end implies the distance of sensor to paper varied over the
length. The paper distance laying on top is one concern. Making this up, but
if the carriage rods or the top cover could drop or rise at one end, it could
do that too.
 
Is this distortion viewed on the monitor screen, or after printing on
paper?
CRT monitors often have a trapezoid adjustment which can warp the screen
view
that way. Printing it would eliminate this factor.
Click to expand...

I only see it on screen, and since the music is scanned for an electronic
music stand, printing doesn't enter into it. I notice this most dramatically
in Photoshop when I have to rotate the page (usually .3 to .5) to make the
music staff lines parallel, but then the beginnings of the staff lines are
no longer flush to each other on the left side.
In the scanner, carriage guide rods can be mounted off square, so that the
carriage moves slightly sideways as it goes down the page. They are built
to
snap together right however.
Wider at one end implies the distance of sensor to paper varied over the
length. The paper distance laying on top is one concern. Making this up,
but
if the carriage rods or the top cover could drop or rise at one end, it
could
do that too.
Click to expand...

I didn't think of this until now, but when I bought this scanner used, it
was advertised as working but didn't work when I tried it. I took the top
off and moved a few things by hand, then put it back together and it worked.
I always assumed that since it worked that I didn't disrupt anything, but
maybe something happened -- I'll have to check that.
 
I only see it on screen, and since the music is scanned for an electronic
music stand, printing doesn't enter into it. I notice this most dramatically
in Photoshop when I have to rotate the page (usually .3 to .5) to make the
music staff lines parallel, but then the beginnings of the staff lines are
no longer flush to each other on the left side.
Click to expand...

CRT monitors often have a trapezoid adjustment that could cause such
distortion if wrong, but you may see it on other things too. Printing it as a
test would eliminate the monitor to see if there was any difference. If no
difference, it wasnt the monitor. If better, seems like it was. A screen
view is a pretty small image to be printed, but it may be good enough for this
test.

Are you saying the alignment is OK until you rotate it? That would suggest
the rotation of course, it wouldnt be square if you rotate it. I may not
understand your meaning, and I think you mean only 0.3 degrees. You cant
align the paper with the top or side edge of the scanner glass, to keep it
straight? Perhaps what you mean is that rotation is similar to resampling,
which can cause segments of thin line art lines to disappear sometimes. You
might try a grayscale scan once, perhaps at lower resolution to fit the screen
directly, without resamplng to fit. Just thinking... I have not seen that.
 
Back
Top