Microdem now saves in 16-bit BSQ format

[Guest coptered]

The latest update to Microdem will now save in the 16-bit BSQ format that resample.exe will recognize without any further conversion. Microdem also gives you the rows, columns, and NW corner coordinates when the save is complete. You will need the current Microdem_setup.exe installed, then replace the microdem.exe file with the new one contained in microdem.zip dated 25-Aug-2003 located here: http://www.usna.edu/Users/oceano/pguth/microdem/win32/Thanks goes to Dr. Peter Guth at the US Naval Academy for making this update for us!Eddie

[Guest TMY]

The news could be also interesting. But the site from you signalled it has been being for months that it is inaccessible.

[Guest sgreenwood]

Hello Eddie,Thanks for the heads-up.>The latest update to Microdem will now save in the 16-bit >BSQ format that resample.exe will recognizeYes, although it seems a bit slowwwwww.>Microdem also gives you the rows, columns,>and NW corner coordinates There still seems to be an minor (?) issue in the way it handles USGS seamless NED data, however.I believe they are the still the wrong coordinates for our purposes. He does seem to have fixed his calculation so it reports the correct NW corner of the map now, but we need the location of the NW elevation point. A difference in perspective which the mapping software I have looked at does not address clearly: PixelAsPoint vs PixelAsArea.His coordinates are shifted NW by 1/2 the resolution of the data. You may want to use the coordinates in the .*.tfw file if you have not made a subset that includes that corner, or at least check them to confirm the discrepancy. If you have created such a subset, you will have to do the calculations yourself.For what it's worth, in its default mode, Manifold shifts the NW corner the same way when Exporting Bil files. And keeps shifting it if you continue to Import and Export the data!I'll have a web page up in a few days explaining the problem in more detail if it is not sorted out in the meantime.Stevewww.fs-traveler.com

[Guest sgreenwood]

>He does seem to have fixed his calculation so it reports the >correct NW corner of the map nowI stand corrected. In general the coordinates are still wrong. Below is a link to a page describing my understanding of the situation. The image also shows how I believe the USGS SDDS server fulfills our requests when specifying boundary coordinates. It would explain why additional rows and/or columns are sometimes included.Using this interpretation, I am able to calculate all the different coordinates reported in the metadata and by Microdem and tie them to the requested coordinates.How serious is this shift in the data? 90m SRTM seamless data will be shifted about 63m from the correct location.10m NED seamless data will be shifted about 7m from the correct location.(I haven't tested subsets with the new version)www.fs-traveler.com/sdds-test.htmlSteve