digitizer help

[mike camann]

GRASS net--

I just discovered that I had been dropped from the GRASS list for some
reason, so I might have missed any replies regarding my digitizer
problem that were posted to the net rather than mailed to me directly.
I have renewed my subscription, so please resend your responses.

Here is some additional information about the problem we're having:

This is a new setup; we're setting up a new digitizing machine.  We
have GRASS4.1 update 2 and a Calcomp 9500 digitizer.  We're using the
default digitizer file that digcap.sampl specifies, e.g. binary format
23 from the Calcomp.

We created a special test $LOCATION called test.digit in order to test
the digitizer.  Its parameters are:

GRASS 4.1 > g.region -p
projection: 1 (UTM)
zone:       17
north:      3502000
south:      3487000
east:       251000
west:       238000
nsres:      10
ewres:      10
rows:       1500
cols:       1300

and the ellipsoid is clark66.

OK-- when we attempt to digitize the 1:24000 USGS quad map that this
$LOCATION encompasses, we enter the following initial control points
located near the 4 corners of the quad map (I know, there're only 4
points, but it's a test...):

1)      241000 (X)      3501000 (Y)
2)      250000          3501000
3)      250000          3489000
4)      240000          3489000

Here is the map layout on the digitizer and the relative location of
the control points:

+----------------------------+
| 2                        3 |
|                            |   X (easting)
|             E              |   |
|             |              |   |
|          N--+--S           |   +----Y (northing)
|             |              |
|             W              |
|                            |
| 1                        4 |
+----------------------------+

Now, when we accept the residuals on those points and re-check the
registration on them, we get:

1)      228764.94       3052165.10
2)      237753.39       3051918.11
3)      238061.99       3051909.63
4)      228094.48       3052183.52

Allowing for some sloppiness in checking the registration points, you
can see that they have somehow been translated.  In particular, notice
the compression of the Y dimension!  If we accept them, then define our
digitizing window using the puck at the diagonal corners of the map, we
get a long (width of the graphics window), skinny digitizing window,
compressed correspondingly in the Y dimension.

The strange part is that once we begin digitizing, we can re-check
those 4 registration points and they're spot-on:

1)      240999.18       3501000.29
2)      250000.89       3500997.16
3)      249999.61       3489002.37
4)      239999.70       3489002.19

The same thing is true for other points with known coordinates.  Of
course, whatever we digitize does not appear in the bogus digitizing
window on the graphics display.

Can you help?  If not, can you point me to someone who can?
Thanks....

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Michael Camann                          camann at dial.pick.uga.edu
Department of Entomology                camann at phoenix.cs.uga.edu
University of Georgia                   (706) 542-1388
Athens, GA 30602                        (706) 542-2276

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%