Default Track Replacements

[eric]

If you are using the US tracks textures, then the US tracks profile for Dynatracks will work.

I can take a look at one for ATracks.

[scottb613]

Hi Eric,

I was just trying to read through this thread and understand how it all works - still a newb - LOL - I don’t even know what A-Tracks are. Please don’t take on any extra work on my account. I’m familiar with XTracks, Scalerail, and US3 Tracks. I know US3 Tracks only have #10 switches which makes it difficult to make a route. While I know you made new tracks using the US3 Texture - I didn’t know the shape was a correct match for US3 as well.

I’ve tried to convert a route previously with US3 and used Dynatrax on Dynamic Track.

So would Dynatrax remove the texture anomaly with Dynamic Track - as it makes a real shape?

Thanks.

Regards,
Scott

[eric]

There was a revision issued Feb 06 with a new TrProfile.sft and I also included it as an attachment to a post earlier here. If there's an anomaly, it might be due to using the older profile.

What Dynatrax does is:

1. create a shape that matches the profile provided, and is extruded along the geometry of the dynamic track entry
2. performs a shape replacement entry in the W file referencing the shape it created in step 1

In theory.... you could do the same manually using Doug's script in Blender. The US Tracks profile is part of his GitHub distribution, and building out the track geometry JSON from the local TSection.dat entry is pretty straight forward. It might even be a logical add-on script to Doug's work for someone who really understands Python scripting...

[NW 2156]

Eric,

Drop me an email when you get a chance.

I've got some things I'd like to discuss.

Thanks.

Robert

[scottb613]

Hi Folks,

OK - I dove in on the new track.

First Off - Eric - THANK YOU - for all your hard work and drive on this - greatly appreciated - sir. You've certainly moved the ball forward - significantly.

My opinion - US3_Tracks look better than A_Tracks.

The Track Replacement looks absolutely gorgeous.

I spend most of my time on Paul's beautiful CT River and Paul uses quite a bit of Dynamic Track. It seems it's all still actual Dynamic Track and not converted with Dynatrax. With the profile provided in Eric's updated pack - the profile seems close - as posted below. The rails and inner berm seem like a perfect match. The outer berm - where the berm pitches down - is a little off. It seems Eric's track pitches more steeply than the official US3_Tracks profile. Eric's steeper pitch looks better. I've tried using Dynatrax and the official US3 Track Profile - there's like 20 different options - I tried several - none seems to improve the situation.

I looked at the Track Profile - included in Eric's Track Pack - it's above my paygrade. Is there a way to make both profiles match Eric's new track? I'd like to able to use Dynatrax to do the conversions.

Thanks.






Regards,
Scott

[eric]

I'll look later today, Scott.

The TrProfile is intimidating at first, but simply defines a set of polylines to extrude. It can be safely tweaked in a text editor.

Judging from the screenshot, it looks like the outer X values on the ballast slope are off.

Look for two individual Position values where the X value is greater than 2.5 / less than -2.5 with a corresponding Y less than or equal to 0.

Change the X vals to 2.5 and -2.5 respectively and you might be good to go. I'll know for certain which to swap once can compare the JSON to the STF later today.

[scottb613]

Hi Eric,

Thanks - I'll look at it again. When I first looked at the TDB file - I felt the same way - once you get used to it -- it starts making sense.


Regards,
Scott

[eric]

I think you may have an old profile, Scott.

There was a number transposed in the TrProfile.

Here's the newest one: ---> USTracks-TrProfile.zip <----

You can also make the edit yourself without a download:

Open up TrProfile.stf and look for ballast_left and ballast_right towards the bottom....

Ballast_right:
Position ( 2.6 -0.02 ) update to Position ( 2.6 -0.2 )

Ballast_left:
Position ( -2.6 -0.02 ) update to Position ( -2.5 -0.2 )

Here are the relevant sections in the JSON Profile used to generate the track:

{
  "CutoffRadius": 2000,
   "Name": "Ballast",
   "TexName": "US_Track3.ace",
   "ShaderName": "BlendATexDiff",
   "LightModelName": "OptSpecular0",
   "MipMapLevelOfDetailBias": -1,
   "Polylines": [ {
    "Name": "ballast",
    "part": "ballast",
    "DeltaTexCoord": [ 0, 0.2 ],
    "Vertices": [
     { "Position": [ -2.6, -0.2 ], "TexCoord": [ 0.862, 0 ] },
     { "Position": [ -1.7, 0.136 ], "TexCoord": [ 0.7158, 0 ] }
    ]
   },
   {
    "Name": "ballast",
    "part": "ballast",
    "DeltaTexCoord": [ 0, 0.2 ],
    "Vertices": [
     { "Position": [ 1.7, 0.136 ], "TexCoord": [ 0.0342, 0 ] },
     { "Position": [ 2.6, -0.2 ], "TexCoord": [ -0.1389, 0 ] }
    ]
   }
  ]
 }​

And in the TrProfile.stf:

        LODItem (
            Name ( "Ballast_Slopes" )
            TexName ( "US_Track3s.ace" )
            ShaderName ( "BlendATexDiff" )
            LightModelName ( "OptSpecular0" )
            AlphaTestMode ( 1 )
            TexAddrModeName ( "Wrap" )
            ESD_Alternative_Texture ( 1 )
            MipMapLevelOfDetailBias ( -1 )
            Polyline (
                Name ( "ballast_left" )
                DeltaTexCoord ( 0 0.2 )
                Vertex (
                    Position ( -2.6 -0.02 )
                    Normal ( -0.17068 0.985326 0 )
                    TexCoord ( -0.1389 0.0 )
                )
                Vertex (
                    Position ( -1.7 0.1359 )
                    Normal ( -0.17068 0.985326 0 )
                    TexCoord ( 0.0342 0.0 )
                )
            )
            Polyline (
                Name ( "ballast_right" )
                DeltaTexCoord ( 0 0.2 )
                Vertex (
                    Position ( 1.7 0.1359 )
                    Normal ( 0.17068 0.985326 0 )
                    TexCoord ( 0.7158 0.0 )
                )
                Vertex (
                    Position ( 2.6 -0.02 )
                    Normal ( 0.17068 0.985326 0 )
                    TexCoord ( 0.8774 0.0 )
                )
            )
        )​

Not a lot of difference between them, so perhaps the ORTS team might consider changing over from STF to JSON and make life a little easier?... Far less room for error if the track profile used to generate track could also be used to render it in the sim...

[defect_detector]

Eric, just dropping in to say thanks for this massive undertaking. Keeping the community alive for years to come, if only we could get some of the massive userbase back...

[eric]

Appreciate you being here, Marshall. Keep producing your eye candy, and perhaps people will take another look one of these days. I've seen a slow but steady trickle of people returning after several years absence.

[scottb613]

Hi Eric,

Figured it out - thank you.

It looks like a fat finger mistake in the original - the 2.6 was a good value - the other value on that line was set to "0.02" - elevation - way too short - needed to be "0.2" to match up with your tracks.

I noted the Dynatrax header in this file - so can I use this same profile to run Dynatrax? Is there any performance gain with Dynatrax?

        LODItem (
            Name ( "Ballast_Slopes" )
            TexName ( "US_Track3.ace" )
            ShaderName ( "BlendATexDiff" )
            LightModelName ( "OptSpecular0" )
            AlphaTestMode ( 1 )
            TexAddrModeName ( "Wrap" )
            ESD_Alternative_Texture ( 1 )
            MipMapLevelOfDetailBias ( -1 )
            Polyline (
                Name ( "ballast_left" )
                DeltaTexCoord ( 0 0.2 )
                Vertex (
                    Position ( -2.6 -0.2 )
                    Normal ( -0.17068 0.985326 0 )
                    TexCoord ( -0.1389 0.0 )
                )
                Vertex (
                    Position ( -1.7 0.1359 )
                    Normal ( -0.17068 0.985326 0 )
                    TexCoord ( 0.0342 0.0 )
                )
            )
            Polyline (
                Name ( "ballast_right" )
                DeltaTexCoord ( 0 0.2 )
                Vertex (
                    Position ( 1.7 0.1359 )
                    Normal ( 0.17068 0.985326 0 )
                    TexCoord ( 0.7158 0.0 )
                )
                Vertex (
                    Position ( 2.6 -0.2)
                    Normal ( 0.17068 0.985326 0 )
                    TexCoord ( 0.8774 0.0 )
                )
            )
        )​

Regards,
Scott

UPDATED US3 Track Profile
TrProfile.zip

[scottb613]

Hi Eric,

LOL - sorry - disregard - I didn't see your response. At least I learned something.

Regards,
Scott

[eric]

No worries. The more people who understand the structure the better.

I've looked at the .profile format that Tim used for Dynatrax, and it's all the same elements arranged differently.

To be honest, as long as the STF files are able to generate the same profile as static track, there's really not much of a need for Dynatrax anymore. If there's a performance hit, it's going to be minimal.

Dynatrax was a valuable tool for those of us who were generating static track pieces, but the python script for Blender that Doug created can do all that and more.

I'm normally one who is late to the game on new tools (i.e. it wasn't until late 2023 that I started anything in Blender, and I didn't use TSRE until 2020), but this is one area where I'd rather be on the bleeding edge for a change.

[CSRX]

Eric,
Maybe it's just me, but I noticed the M-Tracks is set up to match that of the acleantrack files. Is there a chance to have an option set to use the US3Tracks? I only ask, as the mix of the two on the KDSub Enhanced, looks a bit strange.

[eric]

Is there a chance to have an option set to use the US3Tracks?

Yes, that's in the works. Possibly by end of week.

The testing set is available as 88502 but I didn't get any feedback on those. Feel free to give it a spin and let me know.