RE: Seams emminent when mosaicing DEMs

[Mike Dunne]

Hi Ryan, Two things are probably happening here - 1. The clipping on the two datasets is probably different - this would account for the different colors but not elevation part. 2. The 2 pixel wide seam is due to the sunshading - which is a filter type process - not running right to the edge of the dataset. There will be a 1 pixel border around the dataset where the kernel is not applied. This is standard practice for filters and kernels.   The solution is to save datasets out to a virtual dataset, and then apply the sunshading to the vds. When you save to vds, make sure the "delete all output transforms" option  is checked.   As an aside -Whether you can do it all at once depends on how many input points you have and the amount of ram that your machine has. It you do need to do it in chunks the seam effect can be reduced, depending on the gridding method you use.   Hope this helps. cheers Mike -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Ryan Sherby Sent: Thursday, 30 May 2002 6:20 AM To: [EMAIL PROTECTED] Subject: Seams emminent when mosaicing DEMs I am trying to mosaic DEMS made from contours that have been converted to DXFs.  I have a rather large area to be gridded so it has to be done in chunks.  I am providing a 1500ft buffer of each chunk to be gridded then cut back before the final mosaic to rid the interpolation during the gridding process.  As soon as I colordrape the image with sunshading the seam appears.  it is usaually a 2 pixle wide seam.  The seam is radically different colors but not elevations?? This has been stumping me for weeks.    Please Help, ErMapper support never could present a solution, except grid everything at once.  Impossible.