Mesh on small islands

[Guest phpdk]

Hi,I used Ferranti's DEM's to create a mesh for the Faroes Islands. I did several inter- and extrapolations, I up- and downsampled to several LOD's to get just the right look for those fine islands. Using Microdem and the 2000 and 2004 SDK's. Everything went fine, except for one thing:The mesh just does not show on the very small islands. One islands is app. 1 km squared, so there should be many elevations points (up to 400 m), and it shows up just fine in the different DEM readers, and (as bgl)in the TMF viewer. But not in the sim, even if all sliders are at maximum and mo matter what value I use for TERRAIN_MAX_VERTEX_LEVEL. I have tried all possible (?) combinations, without any luck.Is this some kind of limitation in the sim, and if so, does anybody know exactly how it works?PeterPerhaps I should mention that I am resampling 3 merged tiles into one bgl, but that I have also tried with just one tile.

[Holger 0]

Hi Peter,there are probably two issues that work against you. One is that the default shoreline files are notoriously coarse and often misplaced spatially. Thus, if the mesh location doesn't coincide with the land mask (the corresponding HP9*.bgl file) then it will be flattened by the surrounding 0-m water masks in the HP9 bgl file.Moreover, Jonathan's data north of 60 degrees are not currently in the WGS84 datum but in the Pulkovo 1942 datum of the Russian 100k topos. He points this out on his website and actually just uploaded a revised version of the Scandinavia tiles (though not yet the Faroes data) re-projected to WGS84. The result of the datum discrepancy is a shift in E-W direction, which makes it more likely that your mesh doesn't fit the shorelines.What you can do is:* create your own accurate shoreline data, e.g., with Jim's Slartibartfast: http://www.jimkeir.co.uk/FlightSim/index.html* re-project your mesh data to WGS84* wait for an updated version of the Faroes data* adjust the Lat/Lon values in your mesh inf file to better fit the existing maskIf your project is intended to be of longer "shelf life" then I'd suggest to ensure that you're using the correct datum rather than shifting the mesh but the choice, of course, is yours.Cheers, Holger

[Guest phpdk]

Hi Holger,Thank you for your answer. Your explanation about the water mask flattening seems to be true; strange that I didn't notice problems elsewhere in the scenery. I deactived HP946100 and the small islands appeared.About the datum: I did notice Jonathan's remarks about the Pulkova 1942, and I did try in Microdem to convert it to WGS84. From the information in the data header I thought I had succeded. Now I am in doubt, and I shall look into it. (Whether this is the cause or not - I do not think it is, as the shift involved here seems to be N-S, and bigger than the expected few hundred meter). I am certainly not going to cheat on this, however long the expected shelf life ;-)Peter

[Holger 0]

Hi Peter,try Google Earth or NASA's Worldwind to determine the correct locations of some prominent landmarks along the coastline. That should give you a good indication of how badly the default shorelines are misplaced.Cheers, Holger

[Guest phpdk]

Hi again,Jonathan has just uploaded new files in WGS84. Thanks to him.However that confirms that my problem was not there (I had converted correctly) but with the default coastline.Thanks Holger.Peter