New and updated tools for SRTM mesh design

[Holger 0]

Hi all:Anyone working with SRTM2BGL and SRTM data?just noticed on John Childs' website - http://www.terrainmap.com/ - hat he has (once again) updated Blackart (now v3.65). This time there are two important features for FS mesh designers: (1) he has fixed the .BSQ writer, which previosuly resulted in a horizontally flipped mesh; and (2) he has added a batch processing option, so that users can run the time-consuming interpolation of several files at once (e.g., over night). Thanks to John!Anyone unfamiliar with what Blackart can do, check out this thread: http://forums.avsim.net/dcboard.php?az=sho..._id=1459&page=2Also, I noticed a reference to a "SRTM Void Killer" utility by David Gill in Andrej Drobun's recent SRTM mesh of the Republic of Slovenia. It's available at http://www.dgadv.com/dgvk/Based on the description, it merges GTOPO30/DTED0 with SRTM data. I'm a bit sceptical about anything that markets itself as "The Ultimate Tool" but the screenshots look pretty good. It's hard to tell from the limited example whether the fill algorithm manages to avoid the sharp edges (cliffs) that make simple SRTM-GTOPO30 merges often unsatisfactory. Moreover, at a price tag of US$43 this tool is obviously not targeted at freeware designers. Perhaps others know more about David's utility?Blackart (which is freeware) also has the ability of merging DTED0 with SRTM data but I haven't tried that feature myself. Does anyone have any experience with it?Cheers, Holger

[Guest crashaz]

WooHoo! Thanks for the heads up Holger!:-wave

[Guest stentor]

Holger:Thanks a lot by the tips.I don't tried yet, too, because I had now problemsto download DTED0 from NIMA machine.Yes, the screenshots look really pretty good.Didn't understand if the guy merge both, GTOPO andDTED0 at the same operation.It's redundant.Martin Wright's FSTerrain merges GTOPO30 with SRTM,but at sea level you have problems by loweringaltitudes. Merging with tiles over sea level, itseems nothing happens.Since I have results from my experiments, I'll contactyou. Other way, using Blackart, you adjust offset -3m(I use 5 or 6)to flattening ocean areas. It works.By the way, GTOPO30 or SRTM30? What's difference?My best wishesCelso Junior

[Holger 0]

Hi Celso:I too have problems with the DTED0 data - I can't figure out how to download the data in 1x1 degree tile, as is required by Blackart. I only see the option of defining my own area of interest and that's rather tedious. I guess I should ask John Childs whether there's a direct link to the tiled DTED0 data."By the way, GTOPO30 or SRTM30? What's difference?"As far as I understand, GTOPO30 is the "old" version made by compiling traditional elevation data (i.e., based on stereo airphotos) into a worldwide coverage. SRTM30 uses the same data format and tile boundaries but is derived from the Shuttle Radar 1-arcsec data, with holes filled in by GTOPO30 data. Therefore, SRTM30 should be of slightly better quality for most areas.---> ftp://edcsgs9.cr.usgs.gov/pub/data/srtm/S...0_DocumentationCheers, Holger

[Guest stentor]

Hi, Holger:Yes, I agree SRTM30 should be a bit better than GTopo30, butin my experiments I couldn't see no difference.Doesn't matter. Anyway, merging Gtopo or SRTM over a HGT,using FSTerrain I had following results:1) Using files at sea level, the process corrupt terrainby lowering altitudes in general.2) Files over sea level, doensn't happens nothing. There is no effect.It seems the command just works if0 altitude is detected.Better if you use the .bmp file generated byFSTerrain, and correct sea level areas using a paintingprogram like Paint Shop Pro. It is a hard way, but it works.Just I discover how to get correct files DTED0 to use with Blackart and other programs, I'll contact you.Thanks, cheersCelsoPS.: I sent a mail to Childs, once, asking about DTED0 and reply never came.

[Guest andrejdrobun]

Hi there !well, it all depends on a quality of the data. for my part of the world (terrain is quite mouintainous, with lots of slopes over 50-60 degrees) the SRTM data is 3arc sec with as many as 110.000 voids per one degree. That really makes less than 10% of the total of points, but the problem is that the voids are allways grouped in the areas of the highest peaks with the biggest angle of slope and making of mesh with SRTM 3arcsec resolution is simply impossible. I spent a lot of time trying many things, using only SRTM 3arcsec data with different numbers of interpolation passes within the srtm2bgl, but results were always more or less the same - whole groups of mouintain peaks were missing ( of course they were smoothened out by the XXXXX_interpolate.exe ), the mesh looked not so bad, only the altitudes and shapes of the areas where the voids were corrected were completelly off, our highest mountain was unrecognisable and 1500 ft lower than it really is and so on...I would imagine that interpolating SRTM with anything makes no sense in flatter areas or higher data resolutions (= 1arcsec), but in my particular case only resonable result was achieved by the method you all already know.BTW, it is possible to open *.HGT files and edit them directlly, by direct editing of hgt data I solved problems of sea level and some persistent anomalies in the mesh. I also flattened terrain around LJLJ airport the same way. Otherwise the airport was on a shelf on one side and surounded by little hills on the other, which is of course completelly unrealistic, because it's runway has a significant slope in RL and as we know the runways in the sim are flat.