Modified Track Path Editor + Tracks/Discussion

Ouch! Is it broken?

On a different topic, here is a new version of the elevation editor. The old wave feather mode has been upgraded with a new sine wave algorithm. It now produces much smoother transitions, as you can see in the screenshot below:

View attachment 601472
Nice 1 eran :cheers:

As for the disk, I'll try & upload the latest build of Okutama. I'd like to see if it crashes on exit for anyone else as it seems that all CM tracks are crashing on exit after the 2nd load up. It's almost like it can't get back to the main menu :(

A new disc is only £10 from argos so I'm not that fussed if it's the disc.
 
Sorry for spamming updates, but this is a pretty nice one so I hope you'll forgive me ;)

Changes in 1.0.6:

  • The slope calculation has changed. It used to calculate the slope of the line between the height point and the previous height point, but now it calculates (and displays) the slope of the actual point. It makes it a bit more intuitive to read the slope graphs.

  • The big news! Cubic Hermite spline interpolation! Hover the brush over a range of height data points and press the S key to perform a spline interpolation between the endpoints of the range. This new interpolation mode creates a smooth transition between the endpoints, taking into account not only the elevation but also the slope. It also has the benefit of allowing you to smoothen a smaller section of the track, rather than smoothen the entire track at once (which may not be what you want if you're happy with the entire track except for that single bump that you want to correct).

    The video below provides a short demonstration.

 

Attachments

  • elevation editor 1.0.6.zip
    9.2 KB · Views: 25
Well it looks like I'm out of the game for a bit, this is a picture of my track's page :)
20161029_130813.jpg

as you can see, no track's :(

Also official track's don't load now either, grrr.

You guy's will have to wait a little longer for the finished Okutama until I pickup a replacement disc :(
 
There is any modified TPE with the possibility straight start/finish line less than 600 meters?

In principle the elements of the start/finish line can be made shorter, but i haven't tried it yet. I guess the outcome will look strange and there can be problems with pitting on such short pit lanes. But you can place the pit lane and buildings on a (not too extremely) curved part of the track, so it's possible to make a shorter start/finish straight and leave the pit lane and buildings at 700m length.
 
In principle the elements of the start/finish line can be made shorter, but i haven't tried it yet. I guess the outcome will look strange and there can be problems with pitting on such short pit lanes. But you can place the pit lane and buildings on a (not too extremely) curved part of the track, so it's possible to make a shorter start/finish straight and leave the pit lane and buildings at 700m length.

I tried placing the pit lane on a curve when I worked on that figure 8 jump track a while ago, and although it was fine visually, the pit wall collision mesh didn't appear to bend and instead it ran straight across the track.
 
@Mr Grumpy

I believe I have already told you how to achieve that? Actually I believe I've told it in the forums. Even made a tool for it :)

EDIT: Yeap, have shared it like four months ago :D
https://www.gtplanet.net/forum/thre...racks-discussion.337816/page-18#post-11415435

After you replace the modified xml file in the APK and install it, it applies the changed elements to existing tracks as well so you can shorten them without making them again.

Also: https://www.gtplanet.net/forum/threads/shortest-track-made.342725/#post-11258766
 
Last edited:
@Mr Grumpy

I believe I have already told you how to achieve that? Actually I believe I've told it in the forums. Even made a tool for it :)

EDIT: Yeap, have shared it like four months ago :D
https://www.gtplanet.net/forum/thre...racks-discussion.337816/page-18#post-11415435

After you replace the modified xml file in the APK and install it, it applies the changed elements to existing tracks as well so you can shorten them without making them again.

Also: https://www.gtplanet.net/forum/threads/shortest-track-made.342725/#post-11258766

Maybe you did, memory like fish me lol.
Anyways I prefer the tools I got now, so much control over everything.

BTW got a new disc so I'm back in the game. Expect Okutama later tonight, after the Mexican GP :)
 
Testing a new random bump algorithm. It applies a random elevation change to each height point and the magnitude is determined by generating a random floating point number between 1 and 0, and then multiplying that number by itself a number of times.

The end result is that most of the generated bumps become very small, and then there are occasionally some bigger bumps.

I tested the algorithm on the track below, any feedback would be appreciated :)

Note that the two jumps at about 2/3rds around the track were created manually, and not by the bump algorithm.

https://www.gran-turismo.com/gb/gt6/user/#!/friend/eran0004/course/2548774/
 
Look at this, its @NingDynasty track. Somehow he's managed to shorten the S/F straight within the editor.
View attachment 601993
look how long the whole track is.

This is as far as I got at 1.70 miles
View attachment 601995
Also look at the remaining miles between the 2 pictures.

Yeah I actually made the track a while ago and must have forgotten to post it. Razerman helped me learn how to edit the files manually and recompile the APK. It's not the actual starting line for the track but see the curved pit lane issue above. I was able to use the data points for each anchor of my Sebring track so it is identical to the original just shortened to the club circuit layout.

@Mr Grumpy

I believe I have already told you how to achieve that? Actually I believe I've told it in the forums. Even made a tool for it :)

EDIT: Yeap, have shared it like four months ago :D
https://www.gtplanet.net/forum/thre...racks-discussion.337816/page-18#post-11415435

After you replace the modified xml file in the APK and install it, it applies the changed elements to existing tracks as well so you can shorten them without making them again.

Also: https://www.gtplanet.net/forum/threads/shortest-track-made.342725/#post-11258766

This guy has made so much possible and I'm happy to see the rest of you continuing to push the boundaries. Haven't upped it here yet but I also did Anderson RacePark which is a kart circuit here in Florida using the techniques shared.

Testing a new random bump algorithm. It applies a random elevation change to each height point and the magnitude is determined by generating a random floating point number between 1 and 0, and then multiplying that number by itself a number of times.

The end result is that most of the generated bumps become very small, and then there are occasionally some bigger bumps.

I tested the algorithm on the track below, any feedback would be appreciated :)

Note that the two jumps at about 2/3rds around the track were created manually, and not by the bump algorithm.

https://www.gran-turismo.com/gb/gt6/user/#!/friend/eran0004/course/2548774/

Look forward to testing this, perhaps you could assist with applying to my Sebring variants. That's really all that is missing to take that track closer to perfection. I really need to get up with this new stuff.

Have you tested trying to emulate cracks? Basically parts of Sebring are ~20 foot concrete patches and the bumps are the cracks between them. What kind of impact would a ~1 inch section every 20 feet being 2-3 inches below the normal track surface give in game?
 
Last edited:
Have you tested trying to emulate cracks? Basically parts of Sebring are ~20 foot concrete patches and the bumps are the cracks between them. What kind of impact would a ~1 inch section every 20 feet being 2-3 inches below the normal track surface give in game?

Well, the track is divided in sections. Typically, a straight consists of one section and a corner consists of two sections (you can of course add more anchors to these straights and corners and get more sections).

For each section of the track, there is a variable that says how many data points the section has. It doesn't specify where those are located, instead they're spread evenly across the section. For instance, if you have a straight that is 100 meters, and there are 20 data points, then there is 100/20=5 meters between each point. So that is the resolution for elevation changes and banking changes.

You can add more data points, but it's kind of tricky to do that without distorting the elevation changes you already have. And I'm not sure if it's possible to add so many data points that you get that 1 inch resolution you need for the cracks.
 
Well, the track is divided in sections. Typically, a straight consists of one section and a corner consists of two sections (you can of course add more anchors to these straights and corners and get more sections).

For each section of the track, there is a variable that says how many data points the section has. It doesn't specify where those are located, instead they're spread evenly across the section. For instance, if you have a straight that is 100 meters, and there are 20 data points, then there is 100/20=5 meters between each point. So that is the resolution for elevation changes and banking changes.

You can add more data points, but it's kind of tricky to do that without distorting the elevation changes you already have. And I'm not sure if it's possible to add so many data points that you get that 1 inch resolution you need for the cracks.

10-4 perhaps just changing every other section by and inch +/- can simulate that.
 
10-4 perhaps just changing every other section by and inch +/- can simulate that.

I guess that it might be possible to edit the starting point for each section, so that they overlap. That way you'd be able to get two data points close together. You'd need three data points for a complete crack, so this would be more like some kind of "steps" but if it's doable it's probably as good as you can get it.

Here is the concept. The dots are data points and the white dots mark the start of a new section. The blue lines mark the start end end points of each section.

The top set shows the original distribution of data points. The white points have been lowered an inch or two.
The bottom set is the same, only now the start point of each segment has been adjusted so that it begins before the previous section has ended. This way the first data point of the new section gets close to the last point of the previous section, and the result is a sharper drop between these two points.


cracks.png


This is all just a theory though, it might not actually work this way. But it might be worth testing.
 
I'd be careful reducing the track segment lengths to much (20m minimum) as it can cause strange graphical problems like disappearing segments when you approach them on track. Also track overlapping causes issues aswell.
 
I guess that it might be possible to edit the starting point for each section, so that they overlap. That way you'd be able to get two data points close together. You'd need three data points for a complete crack, so this would be more like some kind of "steps" but if it's doable it's probably as good as you can get it.

Here is the concept. The dots are data points and the white dots mark the start of a new section. The blue lines mark the start end end points of each section.

The top set shows the original distribution of data points. The white points have been lowered an inch or two.
The bottom set is the same, only now the start point of each segment has been adjusted so that it begins before the previous section has ended. This way the first data point of the new section gets close to the last point of the previous section, and the result is a sharper drop between these two points.


View attachment 602630

This is all just a theory though, it might not actually work this way. But it might be worth testing.

Thanks for this. It certainly gives me a good idea of how it works since I'm still yet to see it. Looks like we have a couple possible solutions. Will start downloading the files needed and hopefully get started tonight.

The overlapping is a really great idea. I know Mr. Grumpy was talking about overlapping scenery elements has anyone overlapped track elements? Ha answered below, got it.

I'd be careful reducing the track segment lengths to much (20m minimum) as it can cause strange graphical problems like disappearing segments when you approach them on track. Also track overlapping causes issues aswell.

Thanks for the advice. Not sure we'll need to reduce the segment lengths much as long as there are enough data points to edit. Bummer on the overlapping. Looks like we'll have to use the block effect.

I guess the biggest concern would be the pit straight since that's a large chunk of the bumpy section. I do recall though that it's something like 6-7 sections at eifel so each would have ~20 data points. Working it out in my head I think we'll have to go down the adding data points road for the pit straight at least.
 
Last edited:
I'd be careful reducing the track segment lengths to much (20m minimum) as it can cause strange graphical problems like disappearing segments when you approach them on track. Also track overlapping causes issues aswell.

The bank segments would actually become slightly longer. I don't think the road segments would be affected, but I guess it needs to be tested.

It would distort the elevation changes and the banking transitions though. The elevation can always be fixed later on, but we don't have control over the banking of each individual data point so that would be a bigger issue.
 
The bank segments would actually become slightly longer. I don't think the road segments would be affected, but I guess it needs to be tested.

It would distort the elevation changes and the banking transitions though. The elevation can always be fixed later on, but we don't have control over the banking of each individual data point so that would be a bigger issue.

Luckily there is no banking involved.
 
I guess that it might be possible to edit the starting point for each section, so that they overlap. That way you'd be able to get two data points close together.
It would distort the elevation changes and the banking transitions though. The elevation can always be fixed later on, but we don't have control over the banking of each individual data point so that would be a bigger issue.

Great idea! You could also lengthen each segment a bit (add to the vlen member of the banks/segments array) so that it overlaps the next segment. That way the elevation should be altered only slightly because the points of the heights shift only a little bit for each segment and the changes don't add up. The banking transitions can be reset (in theory) by subtracting the added length from each transition_next_vlen member in the banks/segments. You can find out how much leeway you have by calculating vlen/divNum for the segment – that gives the length between height points. I'm curious if it works in practice... :)
 
The bank segments would actually become slightly longer. I don't think the road segments would be affected, but I guess it needs to be tested.

It would distort the elevation changes and the banking transitions though. The elevation can always be fixed later on, but we don't have control over the banking of each individual data point so that would be a bigger issue.

Great idea! You could also lengthen each segment a bit (add to the vlen member of the banks/segments array) so that it overlaps the next segment. That way the elevation should be altered only slightly because the points of the heights shift only a little bit for each segment and the changes don't add up. The banking transitions can be reset (in theory) by subtracting the added length from each transition_next_vlen member in the banks/segments. You can find out how much leeway you have by calculating vlen/divNum for the segment – that gives the length between height points. I'm curious if it works in practice... :)

My concern with this, and this is based off my use of the xml editing, if we lengthen the sections we are talking about will this lengthen the associated pit scenery and cause overlapping issues with that?
 
I don't think so, because this is only editing the Bank data. The scenery is handled by the Road data.

This is a very important difference between what I've been working with in the past via the XML editing. Those modifications would impact both the scenery and track section. Glad to know this type of editing is possible.

I like the fact that @NingDynasty has got involved because all of a sudden this thread has become alive with activity again :)

Happy modding people :cheers:

I've always been lurking but you tagged me so I figured I could try to do some more editing.
 
Is it possible to further decrease the turn radius to allow for smaller tracks?

There is a modded version of the app that allows any turn radius. You should be able to find that APK earlier on in this thread.

---

I'm making some major changes to the elevation editor. During the process I'm testing some tracks to see that the output files are okay and here is the track I'm testing right now. The start/finish straight gets the honour of illustrating the mathematical beauty of the sine wave :)

IMG_7788.JPG
 
There is a modded version of the app that allows any turn radius. You should be able to find that APK earlier on in this thread.

---

I'm making some major changes to the elevation editor. During the process I'm testing some tracks to see that the output files are okay and here is the track I'm testing right now. The start/finish straight gets the honour of illustrating the mathematical beauty of the sine wave :)

View attachment 603014
What I'm saying, basically to make the collision mesh shorter, as when a turn is too tight, graphical glitches and invisible barriers appear on the road
 
What I'm saying, basically to make the collision mesh shorter, as when a turn is too tight, graphical glitches and invisible barriers appear on the road
Unfortunately the mesh wraps around itself when making the turn too tight and there is no way around it imho. Approx. 20m turn radius is minimum, depending on the road width, i.e. the tightest road side geometry is 13m wide, plus half the road width (and some safety margin) gives the minimum turn radius.
 
Unfortunately the mesh wraps around itself when making the turn too tight and there is no way around it imho. Approx. 20m turn radius is minimum, depending on the road width, i.e. the tightest road side geometry is 13m wide, plus half the road width (and some safety margin) gives the minimum turn radius.
That sucks, so making tracks like any of the ebisu circuits is impossible
 
Unfortunately the mesh wraps around itself when making the turn too tight and there is no way around it imho. Approx. 20m turn radius is minimum, depending on the road width, i.e. the tightest road side geometry is 13m wide, plus half the road width (and some safety margin) gives the minimum turn radius.

Yep this was key in developing Anderson RacePark. I had to force the narrowest road element to replace all other road elements. It did create some pretty tight stuff though.
 
Back