[eric from trainsim]

PA1930, on 19 March 2013 - 03:32 PM, said:

Can you give me an example on where to add such a parameter on the .CON file?
Does it crash with MSTS if such alteration is done?
I thought the trains would tilt "extra" if the parameter "max tilt" was increased on the .eng and .wag files.

It's not a new parameter in the .CON file -- it's a naming convention used **for** the .CON file.

As I understand it.....

If you have AMTKAcela.con, copy or rename it to AMTKAcela_tilted.con and it will enable super-tilting in ORTS.

You'll also need to update the TrainCfg parameter in the .CON file to in order to keep everything happy...

Example:

TrainCfg ( AMTKAcela  

becomes

TrainCfg ( AMTKAcela_tilted

Because the file format and parameters aren't changed, MSTS won't notice anything different.

[PA1930]

Thanks for the reply! Though I wonder then that the value of "max tilting" is then adjusted on the .eng and .wag files as I mentioned?

[_o_OOOO_oo-Kanawha]

I have been reading through the dynatracks.cs file from the Trunk/Source/RunActivity/Viewer3D/ branch of the source code.

There is mention of a 'DynaTrackProfile' which I believe generates the dynamic superelevated curves from the standard ACleanTrack1 (ballast) and ACleanTrack2 (railhead and sides) textures.
Im a noob in programming, but the code leads me to conclude there is a possible .xml or .stf (or canned which means hard coded?) track profile file from a TrackProfiles folder inside your route folder.

When I create this folder and put a dummy TrProfile.xml file in it, my WP3 route shows nice blank spaces instead of the ugly default MSTS track in curves.
Also there are some file access errors in the logs, mentioning missing 'trp' instances from each tile I think, which means the code is working.

Is it possible to create a DynaTrackProfile which uses ScaleRail?

[Blillpers]

DBTracks, narrow gauge and UKFinescale profiles would be nice too!

[JTang]

I believe there are such profiles exist, just need to find them somewhere.

[_o_OOOO_oo-Kanawha]

JTang, on 20 March 2013 - 12:10 PM, said:

I believe there are such profiles exist, just need to find them somewhere.

That would be nice for experimentation.

There is commented-out line of code that reads:

// FOR DEBUGGING: Writes XML file from current TRP //TRP.TrackProfile.SaveAsXML(@"C:/Users/Walt/Desktop/TrProfile.xml");

Which I believe can be used to create a sample track profile.
The dynamic shape file of both the track and rail cross section needs to be found and adapted to ScaleRail size and model hierarchy.
The ScaleRail is constructed of 5 pieces, while the MSTS default track is constructed of 3 pieces in the Model Hierarchy of Shape Viewer.

Or course one then needs to find the proper ScaleRail textures and the handles to the different parts inside the .ace files that make up ballast, ties, tie plates, rail heads and sides.

[ytracks]

I believe this to be a good idea. Old Routes are not designed to have superelevation. So I think you should not spent too much effort to convert old routes to the new style.
regarding new routes there are some possibilities:

-you can do superelevation depending on the length of the transition curve. At prototype the longer the transition curve the higher the superelevation. So without transition curve the track will stay horizontal. (BTW this will eliminate the problem with switches).
-you can invent an additional superelevation parameter inside the worldfile for trackobjects, comparable to the collide flags. In this way it is on the route developer to manage SE and you can assign SE even to straight tracks whereever needed. And you can finetune the amount of elevation. At prototype the amount of superelevation is not necessaryly the highest at sharpest curves.

keep on the good work

Ytracks

[_o_OOOO_oo-Kanawha]

This is the "Default Dynatrack profile" I managed to create from reading the source code.

Pull the .stf file out of the archive and put it file in a new folder named TrackProfiles under your route folder.
Like in the picture below:


The .stf file will be parsed without errors by the RunActivity.exe. It passes by with a TRP(.stf) entry in the log.

Feel free to experiment, putting different shapes, sizes and positions in the values. Have the source code at hand to get an idea of what goes where.
This .stf uses the default A1CleanTrack textures and is only valid for 1435mm standard gauge.
There is only one LOD, but the "canned response" in the source code also has a 1200 meters LOD and a dynamically rendered value LOD.

I think that for ScaleRail to work one needs 2 more LODItems but since I don't know how to disassemble .ace files, I don't know what they are called and how the textures are oriented. Also the LOD's are different.

DB Tracks is more complicated still, with yet other LOD's and still more LODItems.

I also think there is only one profile at a time per route possible, so one is currently unable to differentiate between mainline and spurs for superelevation.
Perhaps in a future release more definitions could be possible and an extra variable added to the TrackObj definitions in the .w files will choose the appropriate definition, or the textures needed could be read dynamically from the track piece the train currently travels over.

Attached File(s)

•  TrProfile.zip (783bytes)
  Number of downloads: 746

[Csantucci]

Thank you for this valuable info. Using it and your .stf template file I was able to create a (rought by the moment) profile for the NG Bernina route.

[_o_OOOO_oo-Kanawha]

If your route uses the MSTS default track you are safe.

Here is my more definite ScaleRail Track Profile for use with the Western Pacific Feather River route.
I have changed the 'Chordlength' to 1 meter so the straight 'chords' that make up a curve are less conspiguous.
I have changed all dimensions of the rails 'vertices' to match the ScaleRail contours.
I have changed the MSTS ballast texture, added ties and tieplates so it looks very close to wooden ties ScaleRail this route uses.

Since there is only one type of superelevated track generated dynamically, the special curved bridge, tunnel and concrete types will be replaced by this definition from the .stf.

Also there is still some Moiré patterning, small gaps between track here and there, and the shadows seem to play a particular game on the rail sides.
All of this superelevation is of course still experimental, tested on X 1502.

Attached File(s)

•  TrProfile (WP3rd_SRv2).zip (1.14K)
  Number of downloads: 843

[Csantucci]

_o_OOOO_oo-Kanawha, on 23 March 2013 - 05:24 AM, said:

If your route uses the MSTS default track you are safe.

Sorry, I don't understand well your sentence. The Bernina railway is narrow gauge, so it does not use MSTS default track, with the exception of Tirano, where there are both the narrow gauge and the standard gauge station.

[_o_OOOO_oo-Kanawha]

Csantucci, on 23 March 2013 - 05:31 AM, said:

Sorry, I don't understand well your sentence. The Bernina railway is narrow gauge, so it does not use MSTS default track, with the exception of Tirano, where there are both the narrow gauge and the standard gauge station.

Sorry, C, I thought that your route was built using the standard gauge tracks, and your 'narrowed' them.
I do have the Bernina installed on some other computer, will you share your .sft file as well for testing?

Nice route, BTW, but a little sparse in scenery since I have not yet copied all those default shape files and such.
I have the Bernina in a folder that uses OpenRails exclusively, so it has not even the standard MSTS files and folders.

SOB is another nice Swiss route, are there any more?

Will try to "superelevate" the Milwaukee Road RMD next as it also seems to use custom track textures.

[_o_OOOO_oo-Kanawha]

By looking carefully over the ScaleRail track and its textures I found that the track is a little asymmetrical, the ties are not 100% centered to the ballast.
This shows a little in those tie plates when you hover really low over the track and move from left to right.

In my intepretation of superelevated ScaleRail there seems to be more Moiré patterning than in the 'flat' ScaleRail.

How can I reduce this effect, caused I think by overlaying textures?

I have tried to avoid 'Z-buffer fighting' by having each layer 2 cm above the other. Perhaps some not-rounded value is better since I eyeballed this first attempt?

Also, I think there are no actual .s files used to render the dynamic superelevated track. Everything is constructed "on the fly" using those LODItems and their "polylines" and "vertices". This means one can recreate the proper ScaleRail ballast contour and rail shape. One just needs to find the right measurements and texture file coordinates.

[Csantucci]

To Edwin: I must admit that apart the Bernina, the SOB and Vone's old Albula route, I haven't tried other Swiss routes, so I don't know what the situation is. There is a payware Gotthard route, but I haven't bought it. It's true that parts of the Bernina route have few buildings, but I preferred not to insert buildings not corresponding to reality. And there are large parts of the route where you won't see buildings in real life!

Referring to using superelevation and a specific .stf file in the Bernina route, release 1505 led to a further improvement: by defining 1000 mm as gauge the non superelevated track is no more drawn in the same track sections where also the superevated track is drawn. However after having done some tests I think that by the moment I will use superelevation only in some runs because:
- due to the fact that in the .stf file only one type of track can be defined, I have the standard track not only where other types of track should be drawn, but even in tunnels
- moreover in tunnels and only in tunnels non-superelevated track and superelevated track are both drawn. But this could be the fallout of some trick I used to avoid white stripes on the ground at rail joints.

On point 1 I hope that JiJun will make the miracle...

PS: I saw only today that I was requested to share the track profile that I developed for the Bernina Freeware. Here it is:
 TrProfile.zip (975bytes)
Number of downloads: 521

[_o_OOOO_oo-Kanawha]

I have updated my posting #72 with a more definite and accurate ScaleRail profile for use on the Western Pacific Feather River Canyon route.

Apart from the slight Moiré patterning, there is hardly any visible difference.
But what about performance in FPS? Does this eye-candy detract from performance?