Requesting Help!! - DEM Bugging and Issues

Tanker46

Fat Director
Hello Everyone!

I'm reaching out of my usual fields and posting in unfamiliar territory again because I have found myself in a bit of a predicament regarding my DEM of the Southern Railway and I could very much use some advice from more experienced DEM Creators.

Currently, my good friend Frank (Borderreiver on the forums) and I, are attempting to create a series of DEM Maps that will essentially cover the entire Southern Region of British Railways in the 1950s. The maps are intended to be mergeable with one another so that a seamless virtual replica of the Southern Network in the twilight of steam and the dawn of electric and diesel. Given the current state of Trainz's Route Capacity, merging them into one massive route is just shy of impossible, but with a few more iterations of the game I can definitely see this massive ambition becoming a reality. My current goal is to create three separate routes that cover the entire networks of the original Railways that governed them - the LBSCR, LSWR, and SECR - and complete them so that in the future they can be combined together into one massive route once Trainz has the capacity to support such a massive creation. This way users will have miles of track to run the Southern Region Locomotives and rolling stock that so many great modelers have created for us to use.

The current DEM we are working on is the Central Portion of the Route or as a lot more people recognize it, the London, Brighton and South Coast Railway (LBSCR). At present more than 75% of the LBSCR's former network has been mapped. All that remains is the lines to Eastbourne and Hastings, the remains of the Bluebell and Oxted Line, the remainder of the Steyning Line to Horsham and Dorking, and the portion of the West Coastway Line that ends at Fratton/Portsmouth - the terminal at the western edge of the LBSCR's Borders.

However, we have run into a significant problem as we have attempted to continue eastward. Frank explained it to me like this:

''Heading east remains something to be overcome, as to get the map panels to join up means altering the co-ordinates to the UTM zone of the London to Brighton panels. What this bakes in is a geophysical anomaly, which will grow as the line progresses east. This will then make it impossible to successfully join routes to the east in DEM built to the correct UTM zone co-ordinates. You will not really see this issue in a UK line aligned north and south but you will encounter it in a line aligned east and west. How far the line can go west before we encounter the same issue I do not know."

My question for the all of you veteran DEM creators is as follows - is there any way to resolve this error we are encountering the further we map eastward? I know many US and AUS Routes that have been DEM'd that far exceed the distances and lengths that we are attempting, and they seem to be able to map their territories with little to no effort. (At least from where I am standing..😅) I desperately would like to know if this is a common issue or if there is a workaround. As it stands, if we cannot overcome this problem, it will leave Kent and the entire SECR Network inaccessible to the project.

I'm asking anyone from any community for help on this. USA, German, Australian, Russian, Japanese etc.. It doesn't matter what your preferred community is. I am desperately requesting assistance from anybody who could help me and Frank figure this out.

Best Regards,

Alex Geisen (Tanker46)
 
This sounds like an error in the original ordinance map position. I ran into an extreme case when I was working on a modestly-sized route that was about 60 miles long. Some of the maps were incorrectly encoded and referenced UTM coordinates that are way out of place from the rest of them. This effectively put the route into the trash at least at the time.

I can't think of a way to get around this unfortunately if the underlying DEM data doesn't line up with the maps that are being positioned on top of them. You could try manually placing the map tiles in the area in Surveyor, but that's a really awful amount of hard work.
 
This sounds like an error in the original ordinance map position. I ran into an extreme case when I was working on a modestly-sized route that was about 60 miles long. Some of the maps were incorrectly encoded and referenced UTM coordinates that are way out of place from the rest of them. This effectively put the route into the trash at least at the time.

I can't think of a way to get around this unfortunately if the underlying DEM data doesn't line up with the maps that are being positioned on top of them. You could try manually placing the map tiles in the area in Surveyor, but that's a really awful amount of hard work.
I see. The current DEM consists of three sections: The South London portion, the Central Portion and the Southern portion. I would most likely have to figure out which section is the culprit and disconnect it from the route before redoing that particular DEM perhaps?

Although it could be tedious, I'm interested in hearing more about this manually placing map tiles in Surveyor. Could you explain to me how one would go about it?
 
The problem is you can't even use georeferenced map images that are converted to textures on top of the exported terrain because the map-tile information is so far off from where they are supposed to be. You would need to generate the UTM tiles objects in TransDEM. They are actually assets that TransDEM generates and then fiddle with them in Surveyor and manually place them on top of the DEM in their correct position and ignore the auto-placement by TransDEM. I have never created these tiles but I have seen the option when exporting routes from the program.

I agree, I would figure out who the culprit is prior to redoing the whole dem. It may not be the whole area you are trying to place the maps on and only one or two maps being the culprit. That would be ideal, not really, but a lot better than a whole route being totally off.

Another thought is, you may have to split up the project into parts due to the size of the area. I noticed that Trainz gets weird with really, really large routes. It may load them up okay but when objects are added, and sessions are made, the program still doesn't perform well. Sessions too for large routes become sluggish and the AI drivers do weird things even in the latest beta where that issue was fixed. The issue I see is there's much that needs to be loaded up that the memory and CPU are pushed to their limits.
 
In this case I thinkit might not be size related, but an anomaly in the DEM data. I'm saying this because some years ago I tried to put together a DEM for the Bluebell Railway and ended up with exactly the same problem just south (IIRC) of Horsted Keynes.
 
In this case I thinkit might not be size related, but an anomaly in the DEM data. I'm saying this because some years ago I tried to put together a DEM for the Bluebell Railway and ended up with exactly the same problem just south (IIRC) of Horsted Keynes.
If it's an anomaly in the Data is there a way around or to at least mitigate it? Or is just a consistent no matter what I do? If I cannot expand any further on the LBSCR Map, I'd at least like to make sure that I can still map Kent itself as I still plan on creating a route based from the SECR Regions as well as the other lines throughout Kent.

It's funny you mention the Bluebell Railway as I have the length of the route from East Grinstead to Horstead Keyes in the DEM Currently. Only Newick & Chailey and Sheffield Park are currently missing. I'm going to talk to Frank about filling them in along with the remainder of the Oxted Line. Perhaps a Bluebell Release in the future?
 
Hello Everyone!

I'm reaching out of my usual fields and posting in unfamiliar territory again because I have found myself in a bit of a predicament regarding my DEM of the Southern Railway and I could very much use some advice from more experienced DEM Creators.

Currently, my good friend Frank (Borderreiver on the forums) and I, are attempting to create a series of DEM Maps that will essentially cover the entire Southern Region of British Railways in the 1950s. The maps are intended to be mergeable with one another so that a seamless virtual replica of the Southern Network in the twilight of steam and the dawn of electric and diesel. Given the current state of Trainz's Route Capacity, merging them into one massive route is just shy of impossible, but with a few more iterations of the game I can definitely see this massive ambition becoming a reality. My current goal is to create three separate routes that cover the entire networks of the original Railways that governed them - the LBSCR, LSWR, and SECR - and complete them so that in the future they can be combined together into one massive route once Trainz has the capacity to support such a massive creation. This way users will have miles of track to run the Southern Region Locomotives and rolling stock that so many great modelers have created for us to use.

The current DEM we are working on is the Central Portion of the Route or as a lot more people recognize it, the London, Brighton and South Coast Railway (LBSCR). At present more than 75% of the LBSCR's former network has been mapped. All that remains is the lines to Eastbourne and Hastings, the remains of the Bluebell and Oxted Line, the remainder of the Steyning Line to Horsham and Dorking, and the portion of the West Coastway Line that ends at Fratton/Portsmouth - the terminal at the western edge of the LBSCR's Borders.

However, we have run into a significant problem as we have attempted to continue eastward. Frank explained it to me like this:

''Heading east remains something to be overcome, as to get the map panels to join up means altering the co-ordinates to the UTM zone of the London to Brighton panels. What this bakes in is a geophysical anomaly, which will grow as the line progresses east. This will then make it impossible to successfully join routes to the east in DEM built to the correct UTM zone co-ordinates. You will not really see this issue in a UK line aligned north and south but you will encounter it in a line aligned east and west. How far the line can go west before we encounter the same issue I do not know."

My question for the all of you veteran DEM creators is as follows - is there any way to resolve this error we are encountering the further we map eastward? I know many US and AUS Routes that have been DEM'd that far exceed the distances and lengths that we are attempting, and they seem to be able to map their territories with little to no effort. (At least from where I am standing..😅) I desperately would like to know if this is a common issue or if there is a workaround. As it stands, if we cannot overcome this problem, it will leave Kent and the entire SECR Network inaccessible to the project.

I'm asking anyone from any community for help on this. USA, German, Australian, Russian, Japanese etc.. It doesn't matter what your preferred community is. I am desperately requesting assistance from anybody who could help me and Frank figure this out.

Best Regards,

Alex Geisen (Tanker46)
Are you using TransDem? If yes, post your issue in the TransDem forum. https://forum.transdem.de/viewforum.php?sid=b6854ebdd9856c69a7697a25d321e624
 
Back
Top