SCOtopo - Modern Open Rails Cloud Terrain Builder (USGS 3DEP)

[scottb613]

That sounds like a plan and welcomed addition....I subscribe to the topic...hopefully that will let me know when you release...I have a few routes I want to try it out on.

Hi...

Yeah - I'll post here when I'm done - it's getting close.

One thing - I'm finding the terrain mesh is so much better - that I can see the actual ROW pretty clearly. I have a new version of then LHR that's been a WIP for years - about 80% of the track was placed - now that I can see the ROW - I need to rework much of the track. Previously I had used the map tiles as gospel - dutifully placing the track on the line even if it didn't always make sense based on topography - now the berms and cuts for the track are clearly visible. The shift is not that great - maybe off by like two or three track widths in places. Still - to pull track and reset grades - turns into a huge ordeal.

Regards,
Scott

[eric]

I'm currently running it against one of my existing routes to see how it comes out.

Over 2600 tiles to create....

Scott, is there by chance a way to overwrite the terrain mesh but still preserve any existing terrain texture mapping? I'm assuming Append only creates what is missing.

Sent from my SM-S911U using Tapatalk

[scottb613]

I'm currently running it against one of my existing routes to see how it comes out.

Over 2600 tiles to create....

Scott, is there by chance a way to overwrite the terrain mesh but still preserve any existing terrain texture mapping? I'm assuming Append only creates what is missing.

Sent from my SM-S911U using Tapatalk

Hi Eric,

You'll be there for a while - LOL..

Now that's a STRESS TEST.


I think it took me about an hour for almost 600 tiles. I have an older machine - though. Also - it seems the USGS cloud is sometimes snappier than it is at other times.

Hopefully you got passed the initial memory error on that behemoth.

By default - it does not - or should not - destroy any tile texture or water on tiles - updates topo geometry only.

If you have tiles that are messed up - or suspected corrupted - then you select the "Clean Tile Wipe" - it creates - new - bare naked - tiles.

Regards,
Scott

[eric]

Cool..... It's been running for four hours, and at 1100 out of 2668. I'd done photo terrain on all of the route already, as well as some selective terrain painting, which is why I didn't want to lose that. Will see how it turns out later tonight I guess. I can already see an improvement. Embankments where they didn't used to exist, and a much more defined Chicago River.

[amtrak115]

Scott, when you say marker files are meaning KML/kmz files that follow the track and thereby gives you the tiles or is it something different??

[scottb613]

Scott, when you say marker files are meaning KML/kmz files that follow the track and thereby gives you the tiles or is it something different??

Hi...

It's the standard old marker file used from MSTS's inception. Just a file with a series of points - after I get this out - I can add a function to read KML as well.

Why they chose to put Longitude before Latitude - I'll never know - the programmers clearly weren't navigators....


Marker File

SIMISA@@@@@@@@@@JINX0I0t______

Marker ( -74.2155 41.4828 01|Maybrook 2 )
Marker ( -74.2156 41.4671 02|Burnside 2 )
Marker ( -74.2284 41.4444 03|Girarde 2 )
Marker ( -74.2456 41.4231 04|Hamptonburg 2 )
Marker ( -74.2377 41.4008 05|Farmingdale 2 )
Marker ( -74.2395 41.3845 06|Craigville 2 )
Marker ( -74.2554 41.3618 07|Greycourt 2 )
Marker ( -74.2727 41.3516 08|Chester 2)
Marker ( -74.2854 41.3209 09|SugarLoaf 2 )
Marker ( -74.2900 41.2957 10|Lake 2 )
Marker ( -74.3082 41.2829 11|Durland 2 )
Marker ( -74.3234 41.2733 12|Stonebridge 2 )
Marker ( -74.3705 41.2522 13|Warwick 2 )
Marker ( -74.4095 41.2421 14|Milford 2 )
Marker ( -74.4328 41.2364 15|NY/NJ 2 )
Marker ( -74.4553 41.2232 16|DeKays 2 )
Marker ( -74.4825 41.2068 17|Vernon 2 )
Marker ( -74.5108 41.1953 18|SandHills 2 )
Marker ( -74.5383 41.1785 19|McAfee 2 )
Marker ( -74.5745 41.1494 20|Hamburg 2 )
Marker ( -74.5882 41.1232 21|Franklin 2 )
Marker ( -74.6402 41.0986 22|LakeGrinnell 2 )
Marker ( -74.6724 41.0641 23|Sparta 2 )
Marker ( -74.7458 40.9971 24|Andover 2 )
Marker ( -74.7999 40.9494 25|Tranquility 2 )
Marker ( -74.8205 40.9317 26|Allamuchy 2 )
Marker ( -74.8405 40.9215 27|LongBridge 2 )
Marker ( -74.9103 40.8712 28|GreatMeadows 2 )
Marker ( -74.9335 40.8516 29|Townsbury 2 )
Marker ( -74.9778 40.8308 30|Pequest 2 )
Marker ( -75.0061 40.8318 31|Butzville 2 )
Marker ( -75.0762 40.8203 32|Belvidere 2 )
Marker ( -75.1688 40.7763 33|MartinsCrk 2 )
Marker ( -75.1970 40.7049 34|Hudson 2 )

Marker ( -74.6768 40.9074 35|PortMorris 2 )

Regards,
Scott

[pwillard]

Oh no kidding... that's messed up.

[scottb613]

Hi Folks,

https://trainsim.com/forums/filelib-search-fileid?fid=91086

OK - uploaded - TREAD LIGHTLY - make sure you BACK UP any routes you are playing with.

The directory structure must be kept intact - but - you should be able to place it anywhere.

I included a small little program in the parent directory - double click on it and it should build a shortcut on your desktop. Move the folder - remove the old shortcut - just click on it again to build a new shortcut.

Directions included in root directory.

Capture.jpg

(5) Sources to select files.

Use Route Tiles:
Uses existing route terrain tiles.

Use Text File:
Uses ROUTE\SCOtopoTiles.txt. This processes exactly the listed tiles and does
not use Tile Radius.

Format:
-04bea740
-04bea744
-04bea748

Use Marker File:
Uses ROUTE\[RouteName].mkr. Tile Radius expands around each marker tile.
Will add new tiles.

Use KML File:
Uses ROUTE\[RouteName].kml. Tile Radius expands around KML-derived tiles.
Will add new tiles.

Use Track Database:
Uses ROUTE\[RouteName].tdb. Tile Radius expands around tiles referenced by
the track database.
Will add new tiles.

----

It writes a log file to your desktop - any failed tiles will be listed at the end - for an easy cut and paste into text file to try those tiles again.

Regards,
Scott

[eric]

Not bad -- did the entire 2600 tiles for the CNW Harvard in just over 9 hours. Will try the UP Gila next and then Packerland.

<ducking>
Next mod.... pull in the Terrtex from Google Maps or Mapbox with an API key provided... and an adjustment by x / z in meters to adjust for the places where the imaging is it's slightly off...
</ducking>

[scottb613]

Not bad -- did the entire 2600 tiles for the CNW Harvard in just over 9 hours. Will try the UP Gila next and then Packerland.


Next mod.... pull in the Terrtex from Google Maps or Mapbox with an API key provided... and an adjustment by x / z in meters to adjust for the places where the imaging is it's slightly off...

Hi Eric,

Whew - glad it didn’t crash - well - you’ve certainly proved I am handling memory in a reasonable manner.

Thanks for testing it and the feedback - that’s some serious run.

LOL - I like your tags,


Yeah - not sure how I can verify which source is accurate - for minute corrections - I’ll mull it over.

Regards,
Scott

[eric]

Yeah - not sure how I can verify which source is accurate

I'd allow for the adjustments on the DEM... Spot checking downtown Chicago, I noticed that it's usually a 10-15m difference east-west but north-south looked fine.

It's really obvious where LIDAR detects an underpass or there's a rock cut....

capture_5100685.jpg capture_5100683.jpg capture_5100682.jpg

In all these cases, I'm positive it's the imagery that was off. Getting two sources to line up exactly is never going to be perfect, and it doesn't take a whole lot to mess up precision ( I went to floating point with 8 decimals from 7 in TSRE to fix some of the TDB errors like flipped vectors that I was getting ).

Do not get me wrong -- this is a huge improvement. I'll have to see if adjusting the cursor for the photo imagery in TSRE is possible.

[scottb613]

Hi Eric,

That’s helpful. I can’t think of anyway I am shifting data - but I can look at it again - I’ll double check my tests and see if the same holds true. If it’s always the I case I can shift data. Adding a control to shift the data NS or EW should be pretty easy - leaving it to the end user to determine truth.

Regards,
Scott

[scottb613]

Not bad -- did the entire 2600 tiles for the CNW Harvard in just over 9 hours. Will try the UP Gila next and then Packerland.


Next mod.... pull in the Terrtex from Google Maps or Mapbox with an API key provided... and an adjustment by x / z in meters to adjust for the places where the imaging is it's slightly off...

Hi Eric,

I did check to see if I was doing something weird - with the Goode projection used by Open Rails - nothing I can find.

I already had the offset function built when I was testing on the CLI. Just updated the GUI with "Advanced Geo Bias" sliders - or numeric fields.

I'll finish testing the sliders and upload a new copy tomorrow.


image.png

Regards,
Scott

[scottb613]

Hi Folks,

UPLOADED - new version v0.61 - bias sliders, aesthetics, and some bug fixes.

This should fix your issue Geary.

The bias sliders move the entire selection of geodata (mesh) - including distant mountains - offset by the indicated amount.

Any issues noted - please post the - full log file - here - - so I can review it.


Capture.jpg

Regards,
Scott

[TrainGuy6550]

I have a few routes I want to try it out on.

How? Pardon me if I'm wrong, but those routes must already have objects placed? How would this work on existing routes?