Re-encouraging track modeling

GT VIRUS

In my opinion, track-modeling has always been low in interest. There has never been much interest from modders in creating new tracks, but these days it seems even less so. Since the fall of RSC, I've struggled to find any new material on creating new tracks. While BTB has helped with the creation of new tracks, the standard of these tracks, overall, is lacking. And it's no wonder, as the techniques for creating the truely beautiful tracks aren't being shared anymore.

Even now, 5 years after the release of rfactor, tracks are still being poorly converted and not finished correctly, with shaders being basic and textures not correctly updated. It's a tough job sure, but I feel as the community has been starved of great tracks, that we have become more accepting of sub-standard tracks, to the point were it seems acceptable to release a basic conversion without need for any improvement.

I used to make tracks, unforcinatly never getting to the stage to producing a release-able track. For me, the lack of motivation has been what has stopped me continueing with track-making. With the release of iRacing and it's laser-scanner tracks, every attempt I have made to restart track-making has felt sub-standard, and that now stops me from continuing on.

So what can be done? I do not know. It is obviously not possible to just give everyone laser scanners and tell them to crack on, but the development of techniques to improve track accuracy and feel I think are vital, and hopefully rf2 will help on this front.

I hope that this forum will bring back all the track modellers together, and maybe they can start sharing again. There are techniques in track-modelling now that i'm sure the community can help embrace, and bring a new standard to tracks. Not ever track will be of top quality, but hopefully we can begin to improve most tracks, and at least make them modern.
 
Would help if people learnt how to use 3ds Max instead of using other tools which end up with butchered mapping and bad smoothing etc. 3ds Max isn't so hard when you know how, but admittedly the learning curve is steep. With a little bit of persistence it's not that hard to pick up the basics though.
 
Would help if people learnt how to use 3ds Max instead of using other tools which end up with butchered mapping and bad smoothing etc. 3ds Max isn't so hard when you know how, but admittedly the learning curve is steep. With a little bit of persistence it's not that hard to pick up the basics though.

But how many can really pay the price of a 3dmax-license ? :p BTB, 3DSimEd are tools you can LEGALLY make your tracks with (and in great quality too)
 
i think isi should make plugins for gmax which has most of the basics of 3ds max, which would be good enough for modeling tracks and cars
 
Agree with GT virus. Make this the biggest rfactor comunity. Now modders are on one side, some webs in other side, some make bad copies of another, some modders in other side... disperssed.
 
Well, as with rF1 the release alone will of course generate interest again, which is a good start. But game art, especially environments, is not easy by any stretch. I've been doing this since 1999 and I still learn something new every day. What people don't understand is that to do this well takes a LOT of time and dedication. Someone who has never done this before can't sit down and immediately start building Watkins Glen, for example, and expect it to be a work of art. The other issue is that a track builder HAS to have a good understand of race craft (especially if they're creating fictional tracks)--how will cars drive it, enter turns, exit turns, the flow and rhythm of the track, curbs, camber, etc. It all matters....

And there's the problem. As soon as people realize just how hard it is they get discouraged. What needs to happen really is that the experienced modders have to start taking these new folks under their wing, as it were. Find the ones that are really serious and perhaps start working on a track with them. They'll get valuable experience, and you'll actually have help getting stuff done. This is what makes some of the larger mod groups so strong really....

But really, this is why there never seem to be a lot of track builders--it's just a LOT of work and out of 100 people who think they can do it you may only find a handful that have the skill to do it, or the patience and commitment to LEARN the skill to do it....

For our part, Luc and I certainly have plans, after the sim is done and finished, to work on tutorials to explain the new features and how to implement them. Of course, it's just time that's required, which is something we never seem to have enough of :)
 
That's a good point. Without the help of the iDT team, particually Alex, I proberly would never of developed any skills. I proberly still don't have the skills, as some things (particually combination curbs :S)
 
Well, I think to create fictional race tracks it's not a huge work but, by other hand, to recreate a real-life circuit it's not a job for a single person but for a staff.

Firstly, it's almost impossible get correct layout once the official blueprints of circuits rarely are available for a lot of reasons. Many people believe that layouts based on Google Earth images are accurate, but they're not at all since the satellite images have tons of distortion because both their wrong elevation wrapping and non-orthographic (90°) photo strike. Some cities have detailed surveying available for free but mostly is paid (and expensive).

Secondly, the ideal solutions for track modeling are really expensive, mainly the Autodesk suites (IMHO, the best software for topographical modeling by far – Autocad and 3dsMax). Even Photoshop is not a cheap tool. Personally, I started learning by using my company’s licenses. Bob’s Track Builder is not the ideal solution for several reasons. 3dSimed is a fine tool but it’s more for conversion and texturing purposes. Edit geometry is almost unable.

Finally, I’ve never found a really collaborative discussion forum except at BTB Forum. Mostly, the forums have tons of users and readers but few answerers. Very often to see more than 1000 readings and zero replies. The only practicable way to learn something is by opening the existing good tracks and trying to figure out how the modeler has sorted the things out.
 
Well, I think to create fictional race tracks it's not a huge work but, by other hand, to recreate a real-life circuit it's not a job for a single person but for a staff.

This is SO not true, in my opinion :D If anything I'd say it's harder, and here's why. With a real track you have reference, you have defined turns, camber, elevations, architecture--it is what it is. With a fictional track you have to make it all up yourself, it has to race well, it has to make sense architecturally, you have create believable elevation changes, curb placements, camber, etc. All this is the difference between a good, believable fictional track, as opposed to a bad one. It is hard to get all those elements right... :)

But yes, one central location where information can be found and questions can actually be answered would be wonderful. For one person to answer all those questions would be a full time job, and sadly I don't get paid to sit and read/reply on the forums all day :p
 
I never found making believable fantasy tracks layouts that hard, but then again I've spent far too much time in the past doodling track layouts when I should have been doing something else... :D

The surrounding scenery is usually what I have trouble with tbf.
 
True, but I didn't mention that create a GOOD fantasy track is easy... I recognize the ISI’s hard work on RF1 official content, there’s great fantasy tracks on there… It’s not easy but to create a functional simple track is achievable without massive work. It’s more about good imagination and inspiring and some technical knowledge.

In other hand, although real-world tracks has well defined cambers, bumps, angles and topography once that they are already built, these content are not available for a couple of reasons. We should drive on it for feel how different surfaces pull or push your tires in each track sector, be able to take a huge bunch of pictures, have access to the papers to check the actual cambers and so on…

Just for an example, I live in neighborhoods of Interlagos Circuit (Brazil) and I’m almost unable to drive on that because of too expensive prices and restrict opened periods. I don’t believe that observation from grandstands or onboard Youtube’s videos are enough for a good track building at least in a track with lots of height variations. That’s the why I said it’s a work for a team.


This is SO not true, in my opinion :D If anything I'd say it's harder, and here's why. With a real track you have reference, you have defined turns, camber, elevations, architecture--it is what it is. With a fictional track you have to make it all up yourself, it has to race well, it has to make sense architecturally, you have create believable elevation changes, curb placements, camber, etc. All this is the difference between a good, believable fictional track, as opposed to a bad one. It is hard to get all those elements right... :)

But yes, one central location where information can be found and questions can actually be answered would be wonderful. For one person to answer all those questions would be a full time job, and sadly I don't get paid to sit and read/reply on the forums all day :p
 
True, but I didn't mention that create a GOOD fantasy track is easy... I recognize the ISI’s hard work on RF1 official content, there’s great fantasy tracks on there… It’s not easy but to create a functional simple track is achievable without massive work. It’s more about good imagination and inspiring and some technical knowledge.

In other hand, although real-world tracks has well defined cambers, bumps, angles and topography once that they are already built, these content are not available for a couple of reasons. We should drive on it for feel how different surfaces pull or push your tires in each track sector, be able to take a huge bunch of pictures, have access to the papers to check the actual cambers and so on…

Just for an example, I live in neighborhoods of Interlagos Circuit (Brazil) and I’m almost unable to drive on that because of too expensive prices and restrict opened periods. I don’t believe that observation from grandstands or onboard Youtube’s videos are enough for a good track building at least in a track with lots of height variations. That’s the why I said it’s a work for a team.

Good points, for sure. The long and the short is that each type, real or fictional, has their own specific challenges, true enough. :) But certainly, yes, for a beginner you could create a very simple fictional track to learn the basic skills required for track making. Trying to tackle an extensive fictional track, or a real track with all its many fine details would CERTAINLY not be a good way to start learning :)
 
Bob's track builder is a great foundation for building tracks on, some of the best tracks available today were made with BTB, but the developer has stopped developing it and it is lacking some vital parts.
The shaders for example are not a complete system, the animation is limited to 2 texture frames, you cant weld veritces, you can't remap the mesh, you can't give certain objects an individual name, it exports way to many repeated materials and a few other issues.

On the brighter side, you can import gps data, you can add camber and make smooth corners, you can use lerp and blend textures (terrain only, track blending was planned but never got done). You can add objects and move them around, you can use string objects which makes for adding things like light poles very easy.
As for the gps data, btb becomes a great time saver for setting the layout of the track, from there you can import it into max and continue on.

It would be great if ISI would take up BTB and continue developing it then release it as a tool with rFactor2. It seems like a logical thing to do, you have a very modable base with rFactor, but the tools to mod it are hard to use and not available to everybody. I have met people who have major potential to build amazing tracks and with BTB their potential was put into reality. The community would have missed out on those tracks if it was not for BTB (if that makes sense).

Whilst on this subject I have a question for Scott. In your tutorial you use a bend modifier to make the corners, I have wondered why you chose to do that instead of having it follow a spline? I know there has to be a good reason, but I can't seem to figure it out.

Cheers
 
Whilst on this subject I have a question for Scott. In your tutorial you use a bend modifier to make the corners, I have wondered why you chose to do that instead of having it follow a spline? I know there has to be a good reason, but I can't seem to figure it out.
Cheers

Nope, not my method :p That tutorial was created by the original track artist for rF1, long before I arrived on the scene, and that was his preferred method. I use a spline, extrude method myself.
 
Nope, not my method :p That tutorial was created by the original track artist for rF1, long before I arrived on the scene, and that was his preferred method. I use a spline, extrude method myself.
I had to refer back to the tutorial then, I see now you added a few notes in it which led me to think you made it :eek: . Ok cool thanks for that, Now I don't have to learn a new thing :)
 
I have to agree, using a spline is my prefered methood these days. I remember my first track was using a simular technique to that original tutorial, except using a Pathdeform on the section created. However i found later on in the build that it creates when streching the run-off area's using FFD boxes like originally intended in the model. Created some very thick walls that i didnt notice until it was too late :( these days its splines for tracks and walls, and joining them as nessiscary with fill
 
Bob's track builder is a great foundation for building tracks on, some of the best tracks available today were made with BTB, but the developer has stopped developing it and it is lacking some vital parts.
The shaders for example are not a complete system, the animation is limited to 2 texture frames, you cant weld veritces, you can't remap the mesh, you can't give certain objects an individual name, it exports way to many repeated materials and a few other issues.

IMHO the lack of any feature to cut / merge / weld one track to another is the biggest problem of BTB. Doesn't seem be acceptable to join tracks by moving points of two independent GMTs without at least any SNAP helper. The lack of any cut tool to cap the tracks in any direction is a bit disappointed aspect too. In BTB, all track panels must be parallel including its ending… Thus, you have to hide the rest of the track underneath the second one, without any coincident points… Not a good solution. Try to imagine create the track joints as below using BTB… it seems to be like bend over backwards to me…



Uploaded with ImageShack.us
 
Bob's track builder is a great foundation for building tracks on, some of the best tracks available today were made with BTB, but the developer has stopped developing it and it is lacking some vital parts.
The shaders for example are not a complete system, the animation is limited to 2 texture frames, you cant weld veritces, you can't remap the mesh, you can't give certain objects an individual name, it exports way to many repeated materials and a few other issues.

On the brighter side, you can import gps data, you can add camber and make smooth corners, you can use lerp and blend textures (terrain only, track blending was planned but never got done). You can add objects and move them around, you can use string objects which makes for adding things like light poles very easy.
As for the gps data, btb becomes a great time saver for setting the layout of the track, from there you can import it into max and continue on.

It would be great if ISI would take up BTB and continue developing it then release it as a tool with rFactor2. It seems like a logical thing to do, you have a very modable base with rFactor, but the tools to mod it are hard to use and not available to everybody. I have met people who have major potential to build amazing tracks and with BTB their potential was put into reality. The community would have missed out on those tracks if it was not for BTB (if that makes sense).

Whilst on this subject I have a question for Scott. In your tutorial you use a bend modifier to make the corners, I have wondered why you chose to do that instead of having it follow a spline? I know there has to be a good reason, but I can't seem to figure it out.

Cheers

You can do all of that in max quite easily.

I have some documents that I'll see if I can adapt for public consumption at some time.
 
i personally loft rather than extruding which Scott does, it achieves the same effect though and quicker imo. Yes I loft a single line to create the roads, and then i add extra polys on the road at a later date by a mixture of chamfering and connecting edges. I use the border of the roads to create new lofts or shapes for my grass pieces too.
 
Interesting, I never thought about using extrude, I can see some advantages to it, like being able to mould the terrain around the track as you go. I build cars too, so it would probably suit me more to use extrude rather than loft.
Thanks for the reply, I'll have a play with both techniques and see what I like the most :)

Cheers
 
I ran into a slight problem, I found height data for a track that was recorded on an AIM, but each point is far apart, I spent all last night trying to figure out how to smooth the up/down axis of the control arms so I don't have a steep drop. Is there a function that will let me do this? or is it something that has to be done manually?

Also, when I tried to use extrude, it was ok for the start along a straight, but as soon as I hit a corner the edges didn't pivot with the spline, is there a way to get the edges to do this?
 
Also, when I tried to use extrude, it was ok for the start along a straight, but as soon as I hit a corner the edges didn't pivot with the spline, is there a way to get the edges to do this?

After you extrude the spline up by a set amount--say 10 meters for example, grab the upper perimeter edge (using edit poly) and then apply a "push" modifier. Adjust the push values so the track is as a wide as you need it (you'll use ffd's later on to tweak width, add camber, etc). With another edit poly grab the same perimeter you pushed over and move it down by the same amount you extruded it up. As long as you had "generate UV's" checked when you first extruded you should now have a standard track surface...

After you have the basic camber, elevations, and width changes made, you can grab interiors edges and you "connect" to create subdivisions for crowns, bumps, whatever... It is a bit slower than lofting a cross section, but I prefer to attack the track in stages, and this method allows me finer control of each step. Just personal preference really....
 
I mainly do short ovals and have tried multiple ways to create the track surface. Splines for the track surface just never seemed to work for me. Bend modifier was even more painful. In the end, Im sorta using the ISI tut way, creating road, walls, fence, texturing, etc. in about 50 m sections. Then I will add 2 or 3 sections together and use the ffd 4x4x4 to bank and bend. With the current way I do my dirt ovals, I will loft the walls and fences after I create the track surface. Granted I tend not to stray from anything over 400m long, it works for me.
 
After you extrude the spline up by a set amount--say 10 meters for example,

Thanks for the info Scott, I'm just confused with 1 thing, are you extruding the spline you laid out as a path for the track, or the spline from a cross section?
I tried the process on the spline I laid out as a path and it all went good. But I'm confused because when I think extrude I think extrude a cross section bit by bit (similar to building a car) and snap it to the path spline.
 
Thanks for the info Scott, I'm just confused with 1 thing, are you extruding the spline you laid out as a path for the track, or the spline from a cross section?
I tried the process on the spline I laid out as a path and it all went good. But I'm confused because when I think extrude I think extrude a cross section bit by bit (similar to building a car) and snap it to the path spline.

Well, the spline is the outer perimeter of the track layout. use the extrude modifier on it and it will create a "wall" of geometry, with mapping coordinates already generated. The upper perimeter of that is what you will push over to create the tracks width... That's what you're looking for, yes?
 
Well, the spline is the outer perimeter of the track layout. use the extrude modifier on it and it will create a "wall" of geometry, with mapping coordinates already generated. The upper perimeter of that is what you will push over to create the tracks width... That's what you're looking for, yes?
Ok yep I get it now, I was picturing a spline in the centre of the track, not along the outer perimeter. Now it all makes sense to me and I can already see how this technique will overcome a lot of issues I have always found when using a centre spline. Thank you very much for taking the time to help me. :)


Cheers
 
Ok yep I get it now, I was picturing a spline in the centre of the track, not along the outer perimeter. Now it all makes sense to me and I can already see how this technique will overcome a lot of issues I have always found when using a centre spline. Thank you very much for taking the time to help me. :)


Cheers

No problem at all. And, by the way, you COULD use a center spline if you'd rather. You just need to make a clone of it. You'd extrude one up and push over to the inside of the track. The second you'd extrude up and push to the outside. But, that will make adding camber a bit more work as now you have the inside, middle and outside edges to worry about....
 
And, while I'm at it, I always wanted to try using a spline cage to create the track surface. You'd create splines for the inside and outside of the track, and then add some splines for cross sections. The geometry gets filled in automagically. I use to model EVERYTHING this way, but have never tried it with a track yet. Might be a good experiment after rF2 is finished when I finally have some free time again :p
 
And, while I'm at it, I always wanted to try using a spline cage to create the track surface. You'd create splines for the inside and outside of the track, and then add some splines for cross sections. The geometry gets filled in automagically. I use to model EVERYTHING this way, but have never tried it with a track yet. Might be a good experiment after rF2 is finished when I finally have some free time again :p

I use this technique for most grass areas's, makes things fairly easy. Usually i'll make a spline from the edge of the track to where i want to make a wall. Doing will allow you to use less poly's for the walls, and create an optimised mesh between the wall and track
 
I had a bit of a mess around with it last night, one thing I got stuck on was, how to make all the track panels an equal distance apart?
I also tried a spline cage too and couldn't find a way to generate the mapping co-ords.
I can't get loft to work either lol, (I must be doing something very wrong). I create the spline in top view, I create a cross section in front view, I click loft and select path, but the object that's created doesn't follow the path, it goes up in the air and is back the front, I tried many many ways to get it right, rotating the cross section, resetting the xform, etc etc, but no luck.

On the subject of mapping, I usually use Blender, there is a neat unwrap function in Blender where you can unwrap the mesh following the active quad. You select the quad at the start of the track first to make it active, then you select all and when you click 'unwrap follow active quad', it unwraps the whole mesh in a straight line. I haven't yet been able to find a function like this in 3dmax.
 
After you have the spline all shaped, you'd add verts (using refine) with fairly equal spacing, probably a bit tighter in the corners than on the straights. Then, turn off "optimize" in the spline's properties, and set the numbers of iterations you'd want between each spline vert. When you extrude you'll get a pretty nice even spacing of the segments.

To unwrap the spline cage you could create a simple version of the track with the spline method and project the mapping down onto the detailed version of the track. I do this a lot--works great.
 
Ok, cool, I finally got it to a point where I am happy with it. Thankyou. Hopefully from now on I have less questions :D


Cheers
 
I had a bit of a mess around with it last night, one thing I got stuck on was, how to make all the track panels an equal distance apart?
I also tried a spline cage too and couldn't find a way to generate the mapping co-ords.
I can't get loft to work either lol, (I must be doing something very wrong). I create the spline in top view, I create a cross section in front view, I click loft and select path, but the object that's created doesn't follow the path, it goes up in the air and is back the front, I tried many many ways to get it right, rotating the cross section, resetting the xform, etc etc, but no luck.

On the subject of mapping, I usually use Blender, there is a neat unwrap function in Blender where you can unwrap the mesh following the active quad. You select the quad at the start of the track first to make it active, then you select all and when you click 'unwrap follow active quad', it unwraps the whole mesh in a straight line. I haven't yet been able to find a function like this in 3dmax.


Select the path before clicking loft, then get the shape. Also its the pivot point and spline direction that matters in lofts.

Also check out the normalize spline modifier, rather handy ;)

In the later versions of 3ds max there is spline mapping, although it'll probably require some manual editing too.
 
Thanks Alex, that worked :D , I haven't tried the normalize spline modifier yet though. Curiosity got the better of me and I'm going through the process using the bend modifier atm, trying out all different ways :).
You know what would be very useful? If you could set out the track like in the rf tutorial with a 10m section, then once you have set your elevations, make it attach itself to a spline.
The biggest problem I am finding is setting smooth transitions between elevations. Setting them in a straight line is the best way I have found so far.

I have a question about fdd's, I don't think I am setting them up properly, if I load enough points to be able to manipulate things, it really slows my system down and as a square/s over a twisting and turning track I find it hard to manipulate things smoothly, I also tried doing it little bit by little bit and found it a lengthy process, I found that grabbing each vertice and moving them one by one faster so I must be doing something really wrong :D

[ED] I think I have got a handle on the ffd modifier, breaking the track up into smaller parts made it a lot easier to work with.
 
Last edited:
(I edited a chunk out)
I used to make tracks, unforcinatly never getting to the stage to producing a release-able track. For me, the lack of motivation has been what has stopped me continueing with track-making. With the release of iRacing and it's laser-scanner tracks, every attempt I have made to restart track-making has felt sub-standard, and that now stops me from continuing on.

So what can be done? I do not know. It is obviously not possible to just give everyone laser scanners and tell them to crack on, but the development of techniques to improve track accuracy and feel I think are vital, and hopefully rf2 will help on this front.

The laser-scanned tracks are one reason why I'm modeling ghost tracks and minor tracks. We can expect the big boys with laser scanners to eventually do all the major tracks, but the minor tracks and ghost tracks are where individuals will make the most difference.
 
I have come to a situation with this track where I have got some elevations wrong and I need to fix it. Is there a way to attach a spline to an existing track so the track can be altered? Like pairing the spline with the mesh so the spline controls the mesh.
 
GT Virus, Alex, Scott, I just wanted to pop back in and say thanks heaps! You got me started on this and now I can't stop :D I really appreciate all your help and advice.
It probably doesn't mean much to you guys but this is something I have always wanted to get a grip on and have found it really hard to get started on it. All I needed was a kickstart and since you guys have pointed me in the right direction I finally am able to get into it. Once again thanks heaps!

Cheers!

 
It's not matter of price, you need to learn how to model for a reatime engine, the uvmapping and the shading of a mesh.
This can be done in Blender or in Zmodeler or gMax, cheap or totally free.
A lot of 3d i did on first release of Varano, was done in zmodeler, terrain and buildings, with mapping too.

BTB.... mmh, i'm with Alex Sawczuk, from my point of view is really useful for previews but not for really good stuff.
A waste of polygons, bad shading control, i don't like it but i can understand that is a lot easy to learn, but imho, a 'cheap' shortcut for making racetracks. :)
 
I would love to make tracks, dirt track to be exact. I live in central Iowa and have raced on at least 10 tracks in the area, which I would like to construct a few for the game. I make a living using CAD, but the tutorials I have found make my head spin. I have 3ds max, 3dsim, turbo cad, rhino (dont like), and solid works. I have even downloaded Bobs track builder. I've made several cars for the dirt track portion of the game and understand how to create them,but not the tracks. Is there any step by step tutorial out there that can help me. I have tried the ones listed on this site with no success. With winter starting I will have a lot of free nights to work with. Please help.
 

Back
Top