David Thompson kindly had a look at my program, and he has located source of
the main problem.  My coordinates are of the order of (2000000,6000000),
delta in the elevation data is 1.0, and in this frame the TIFF delta is
0.24.  If the coordinates are all translated by (-2000000, -6000000, 0) the
image after Map (to put elevations into the TIFF-like field) and RubberSheet
is much improved.  It no longer has the odd patterning which David says was
caused by the large number of degenerate triangles that I was getting.
These were presumably caused by a loss of accuracy when very small delta
(0.24) was added to big coordinates, apparently making the triangle vertices
colinear.

David suggests that someone else in the group may be able to suggest why
this is happening.  I can see that numbers like these could create a problem
if DX was using single precision.  Is this the case?

Although the image is greatly improved when the origin is shifted in this
way, it is still not as good as it could be.  The final step of Collecting
the output of RubberSheet with an AmbientLight cleans it up.  Can anyone
suggest why removing the default lighting should solve this problem?

Gib

NOTICE - This e-mail is only intended to be read by the named recipient.  It
may contain information which is confidential, proprietary or the subject of
legal privilege. If you are not the intended recipient please notify the
sender immediately and delete this e-mail. You may not use any information
contained in it.  Legal privilege is not waived because you have read this
e-mail.

For further information on the Beca Group of Companies, visit our web page
http://www.beca.co.nz

Reply via email to