ULTRA-DETAILED TERRAIN

Still no Mac version, so a hard pass here. I'm sure I could run it under Crossover, but I only want to do so much of that. In the mean time, I have Trainz, which runs good on Macs and is optimized to run Apple Silicon Macs.

Still curious how is Trainz going to support the ultra terrain and DEM data.
 
It is amazing to me how every single announcement of new features for Trainz is greeted with demands for features stolen from other train sims. If you don't like Trainz then go play your other train sim. I'm pretty sure no one here cares.
 
It is amazing to me how every single announcement of new features for Trainz is greeted with demands for features stolen from other train sims. If you don't like Trainz then go play your other train sim. I'm pretty sure no one here cares.

I agree; I think N3V should be congratulated on introducing new features that continue to put Trainz above the competition.

An improved weather system would be nice, but I have no doubt that in time we will get it. For now, I for one am excited about the new updates including the new terrain grid, DirectX 12, the new audio system and continued updates. Well done N3V!

Cheers,

PLP
 
Lighting changes are coming with the ray tracing system (which we're currently working on).

As for priorities of Feature A vs Feature B, I know one thing for sure - we'll never please everyone.

Regarding bugs, there's a form here: Trainz Bug Report. I strongly encourage you to submit a DETAILED bug report - one report per bug. We fix most bugs where we have a reliable reproducible issue. (Ask John how many of his reports have been fixed - not sure of the % but the total is "lots").

We're currently at build 118989, and each submission to our source code repository roughly equates to a bug fix or new feature. TANE was in the 7xxxx series, so that's about 50,000 updates checked in in the past 7-8 years.
 
Regarding bugs, there's a form here: Trainz Bug Report. I strongly encourage you to submit a DETAILED bug report - one report per bug. We fix most bugs where we have a reliable reproducible issue. (Ask John how many of his reports have been fixed - not sure of the % but the total is "lots").

I can confirm that. I have lost count of the bug reports I have submitted and while not all of them have been fixed (as of today), the majority have. I also post a forum message about each bug report I make and the response from QA when it comes. There was one occasion when I was the only person reporting a very obscure bug, based on exactly zero responses to my forum post, and it took a long time (several SP releases) before it was eventually fixed, but fixed it was.

To those posters who complain about every new Trainz release being "a buggy mess", how many of you have filled in a bug report as well as complaining? It is far better to do both.

My opinions.
 
Tony,
As the ultra res terrain seems a complex and big change, timeframe has to be unknown. My main question is how is DEM data going to be handled? Current tools operate with high resolution DEM data, however they have to export to Trainz using lower resolution. Will there be a intermediate Trainz ground format that 3rd parties can manipulate and is documented?
 
Tony,
As the ultra res terrain seems a complex and big change, timeframe has to be unknown. My main question is how is DEM data going to be handled? Current tools operate with high resolution DEM data, however they have to export to Trainz using lower resolution. Will there be a intermediate Trainz ground format that 3rd parties can manipulate and is documented?

I was wondering about that too, but as you noted, there is no timeframe on this, and anything can happen in between. There's also the possibility that lower resolution terrain can be imported and then tessellated as needed, which means that the current DEM files supplied by the National Geologic Survey map servers can be processed as they are through TransDEM. Ideally, if Dr. Ziegler could do this internally in the program, then we could have a direct import.
 
I think the current DEM process will continue to work, however we can’t take advantage of the increased resolution until software is updated or Trainz can directly import DEM data. There is a lot of complexity to DEM data though.
 
Will the new terrain allow the sculpting of overhanging cliffs, caves or true tunnels (ie. tunnels without digholes)?

That would be quite an accomplishment. Somehow you'd have to store multiple Z values (altitude) at the same X,Y coordinate. (And tell the program which side is air and which is solid.) I don't know of a simple way to do that. It would be a neat feature though!

Edit:
I guess that I should add that the higher definition terrain looks impressive. It will be interesting to see how the surveyor tools interact with higher def terrain. I wonder about file sizes though. I already consider resulting file sizes when I import terrain with a 5m grid vs the 10m grid.
 
Last edited:
That would be quite an accomplishment. Somehow you'd have to store multiple Z values (altitude) at the same X,Y coordinate. (And tell the program which side is air and which is solid.) I don't know of a simple way to do that. It would be a neat feature though!

Edit:
I guess that I should add that the higher definition terrain looks impressive. It will be interesting to see how the surveyor tools interact with higher def terrain. I wonder about file sizes though. I already consider resulting file sizes when I import terrain with a 5m grid vs the 10m grid.


Well even the simplest 3D creation program can do cliffs, caves, tunnels etc. It’s all just XYZ vertices and the use of normal vectors to distinguish outside from inside. So maybe it’s like that.

As for finer resolution without blowing the file size budget, I would imagine it requires a complete change in the way the terrain is created and how its data is stored. I’ve heard the term “procedural terrain” mentioned but I don’t really know what it means. Perhaps its a form of data compression based on a set of equations for the topology rather than a huge table of XYZ coordinates. Non-expert speculation makes everything seem possible and easy. I should work in the Marketing department. :eek:

.
 
TransDEM already deals internally with higher resolution than current Trainz maps but N3V would have to share the details of the format of the new grid system for TransDEM to output in that format.

N3V back in the day must have shared the details of their route terrain file system with the author of TransDEM, so it's likely they would do so again.
 
Well even the simplest 3D creation program can do cliffs, caves, tunnels etc. It’s all just XYZ vertices and the use of normal vectors to distinguish outside from inside. So maybe it’s like that.

As for finer resolution without blowing the file size budget, I would imagine it requires a complete change in the way the terrain is created and how its data is stored. I’ve heard the term “procedural terrain” mentioned but I don’t really know what it means. Perhaps its a form of data compression based on a set of equations for the topology rather than a huge table of XYZ coordinates. Non-expert speculation makes everything seem possible and easy. I should work in the Marketing department. :eek:

.

Procedural Terrain is used in Unity could be similar to this? https://www.youtube.com/watch?v=64NblGkAabk

Could be Nvidia's method? https://developer.nvidia.com/gpugem...erating-complex-procedural-terrains-using-gpu
 
Definitely not the Unity type of terrain. It’s essentially the same as existing Trainz terrain which cannot do caves, tunnels or even vertical cliffs or overhangs.

The ingenious nVidia method, based on voxels, can do all the things I mentioned and is fast because it uses the GPU. Laughing because I remember when voxels were just dumpy British family cars.


.
 
Definitely not the Unity type of terrain. It’s essentially the same as existing Trainz terrain which cannot do caves, tunnels or even vertical cliffs or overhangs.

The ingenious nVidia method, based on voxels, can do all the things I mentioned and is fast because it uses the GPU. Laughing because I remember when voxels were just dumpy British family cars.


.

I'm inclined to think it will be the nvidia method, given we already have TurfFx from them.
 
Back
Top