TrainzDem & TS2010

steamboateng

New member
I have a route covering approx 35 miles downloaded and converted to DEM files in TrainzDem 2.1.1. I have also imported 3 large USGS raster maps and georeferenced them to the DEM maps. My questions........
1. I plan to draw the route as a vector data to establish the baseboard layout in Trainz. It is my understanding that TrainzDem will automatically trim the baseboards to one or two boards either side of the route. Yes/No?
2. When step one is completed, I can then convert the remaining raster map images to 1km x 1km UTM objects, which I can save and import into the Trainz baseboards. Correct/Not?
3. The dems were generated from 1/3 Arc/Sec data, at a resolution of 10 meters and the dems grid in TrainzDem was set at 10 meters. I will convert the data to TS2010 baseboards. Will it make any difference using 5m or 10m baseboard grid in Trainz, considering the resolution of the original dem data.
4. Has anyone dowloaded and imported into Trainz, data from the 1/9 arc/Sec US Seamless server? What was the result?
Regards
 
I have a route covering approx 35 miles downloaded and converted to DEM files in TrainzDem 2.1.1. I have also imported 3 large USGS raster maps and georeferenced them to the DEM maps. My questions........
1. I plan to draw the route as a vector data to establish the baseboard layout in Trainz. It is my understanding that TrainzDem will automatically trim the baseboards to one or two boards either side of the route. Yes/No?
2. When step one is completed, I can then convert the remaining raster map images to 1km x 1km UTM objects, which I can save and import into the Trainz baseboards. Correct/Not?
3. The dems were generated from 1/3 Arc/Sec data, at a resolution of 10 meters and the dems grid in TrainzDem was set at 10 meters. I will convert the data to TS2010 baseboards. Will it make any difference using 5m or 10m baseboard grid in Trainz, considering the resolution of the original dem data.
4. Has anyone dowloaded and imported into Trainz, data from the 1/9 arc/Sec US Seamless server? What was the result?
Regards

1. yes if you set the route filter

2. not sure what you mean by "remaining" but yest you can make rasters or ortho photo map tiles or just plain map tiles as utm's and import them and they follow the same route filtering... Ie 2 on either side of the route...

3. not much if anything.

4. Yes I have and it's pretty memory intensive but the results are nice. I did Avondale La and the riverbanks and trackbeds were really well defined...
 
Thanks for the relevent reply, mr scsi. You relieved a few concerns I had about TrainzDem.
What I mean about the 'remainig' raster map images is this: The map images I have georeferenced and overlayed onto the DEM images are large rectangular images, each of a complete 7-1/2 deg. USGS raster map. The rail route is only a portion or portions of the maps. Once the route is filtered to 1 or 2 baseboards each side, lage portions of the map will no longer apply. It makes sense to convert the applicable portions of the raster maps to UTM objects after route filtering to reduce the number of UTM objects generated. Can I do it that way?
Im quite pleased to hear that the 1/9 Arc/Sec NED's improve the DEM geography. The area I'm modelling is largely lowland with rather moderate hills dotting the countryside. I'm sure a more detailed DEM model will add much to the route. I'm inclined to download a full set of 1/9 arc/sec NED data and reapply the (already) georeferenced raster maps.
Regards
 
Last edited:
2. When step one is completed, I can then convert the remaining raster map images to 1km x 1km UTM objects, which I can save and import into the Trainz baseboards.
The actual extent or size of the georeferenced image is not relevant when exporting to Trainz. TransDEM looks on top of all the a raster map images loaded the same way they appear in the TransDEM main window. It then iterates either over baseboards (720 m grid) or over UTM squares (1000 m grid).

For the ground texture it processes along the baseboard grid. For each baseboard it looks at the raster map stack, selects the images which are relevant for the current baseboard, temporarily clips these images to the extent of the baseboard and then re-samples the temporary clipping to either the 10m or 5m grid.

Similar for UTM tiles. This time it is the 1000m UTM grid. For each square TransDEM analyses all contributing images for scale. The one with the largest scale will determine the texture resolution (nearest power of 2). Then TransDEM temporarily clips the images as before and re-samples to texture resolution.
 
Thanks for dropping in, geophil.
So as I understand it, and you describe it, TrainsDem will clip the large raster maps to baseboard size, as determined by the 'route filtering' function, and either clip those maps to 750m sq. clippings as a texture or 1000 (500)m sq. clippings as utm objects. automaticaly assigning those clippings to the DEM for import into Trainz.
Or am I over simplifying?
 
Thanks for dropping in, geophil.
So as I understand it, and you describe it, TrainsDem will clip the large raster maps to baseboard size, as determined by the 'route filtering' function, and either clip those maps to 750m sq. clippings as a texture or 1000 (500)m sq. clippings as utm objects. automaticaly assigning those clippings to the DEM for import into Trainz.
Or am I over simplifying?

Basically yes.

Clipping will occur square by square (720 or 1000m). Once a square is finished, the temporary data will be deleted immediately and memory will be freed.

The area needed from the larger raster map image will be stamped out, so to speak, leaving the original raster image more or less unharmed. This temporary clipping will then be re-sampled. (Internally it's probably the same operation which does both the clipping and the re-sampling.)
 
Thanks for clearing that up. I have some fine 3/30/11 dem data with 1943-1949 raster map data (1927 North American Datum) georeferenced to the dem. It worked out perfectly. the raster maps overlay the dem (or vise versa) with no apparent misalignment of contours. I'm impressed.
Regards
 
Back
Top