Gaps in mesh?

[Guest tunger]

I'm another new mesh developer with a question. Here's the overview:I got some data from the USGS seemless data distribution to cover the Escalante region of Utah. I chose 'BIL' format, which turned out to be exactly the format needed by resample. The data came in two segments. I created '.inf' files (see below) and generated two '.bgl' files. The problem is that each 'bgl' file seems to have a margine around it, a region of 3000 ft or so where no data was created even though there is data in the source file. This left a 1 mile gap between the two regions. I can see this gap in tmfviewer and in FS. I later created a single '.inf' file which generated one '.bgl' from the two data sources and got a continuous region. This solves my immediate problem, but leaves several questions I'm hoping more experienced folks can answer: o Why the gap? Does resample always create empty borders? If I'm going to cover an area with multiple 'bgl' files do I have to use overlaping source data? o What is the balance between many small files and few large files? The individual files are 7MB while the combined is 15MB. What is too large?More Detail:Data source, 'inf' files, and 'bgl' files are on my web site at: http://www.tumtum.com/download/fsdata/One 'inf' file looks like this:## 57908423## 37.78754629633217# -111.03078703708712 -110.682175925986# 37.1940277778303#[Destination] LOD = 12 DestDir = "scenery" DestBaseFileName = "57908423" NorthLat = 37.78754629633217 SouthLat = 37.1940277778303 EastLong = -110.682175925986 WestLong = -111.03078703708712[source] Type = ElevS16LSB SourceDir = "57908423" SourceFile = "57908423.BIL" Lat = 37.78754629633217 Lon = -111.03078703708712 NumOfCellsPerLine = 3765 NumOfLines = 6410 CellXdimensionDeg = 9.259259259e-05 CellYdimensionDeg = 9.259259259e-05 It does not seem to matter if I specify the bounding region or 'UseSourceDimensions' I have followed the discussion about LOD 10 vs. 12. I'm interested in the highest detail possible for some specific regions in which I have hiked. I think that I do see a difference. If further testing bares this out I'll post screen shots.I've written some perl scripts to automatically generate the '.inf' files from downloaded metadata. Once I have the above issues resolved I should have a pretty good tool for automatically creating new mesh from downloaded data.

[Guest sgreenwood]

Hi Tom,"The problem is that each 'bgl' file seems to have a margine around it, a region of >3000 ft or so"Probably the size of a quadrant at the LOD you used.While you can avoid the issue by combining files with inf and resmple, there is a limit to this approach, especially if you want to automate the process. Here is a link to my discussion of your problem.www.fs-traveler.com/mesh-gaps.shtml"If I'm going to cover an area with multiple 'bgl' files do I have to use overlaping source data?"Yes! Your best approach is probably to request areas small enough so they do not get split into parts, giving you control of the process. When requesting adjacent areas, be sure they overlap by 1 LOD quadrant. This will insure complete coverage. If you anticipate using multiple LOD values, select the data based on the lowest LOD value you expect to use. Higher LOD values will just add a bit of overlap in the mesh. Going the other way will leave you back where you are now."What is the balance between many small files and few large files?"Small files are easier to distribute/download, and are easier to replace later, if necessary. I don't beleve the size impacts performance much.My bgl file sizes are determined largely by the number of DEMs I can merge conveniently in Microdem and then process with my software, without using the Windows swap file (usually 1-2MB). Steve