T3ED Updates (NFS4 Track Editor) by JimDiabolo

[pete9516]

I've made a video on how to use the "fixed heading" replay camera and how to create animated objects. But unfortunately, i've made many mistakes at the animated objects part. Pls forgive me. At the end of the video, i do it correct ^^
https://www.youtube.com/watch?v=Ib8sztUcs5g

that's what video editing is for lmao

Yes but the video goes beyond an hour and recording another 40 or 50 minutes just because i did a single mistake wouldn't have been worth the effort. I rather post comments with infos where i made a mistake and in overall i think it's better when it's a livestream-ish video instead of a hardcore-edited video where the red thread of getting along with the content of the video goes just way too fast. You can get along better if someones makes something live or more direct, not cut too much. It just takes longer to watch. And i even edited the video because i had to make a cut to find that damn mistake i did with the -1 instead of -16384

[pete9516]

But in the end you are right, when having a look again what i'm actually doing in the video i should have just remade it because it's like a live analysis of what the heck i was doing wrong and then correct it. Not really necessary for the video but i just didn't want to make it all again for some reason even though it would have been more clear to understand and stuff but i just didn't see the mistakes i did and it was all completely new to me with the animated objects so it had to happen

[JimDiabolo]

Here is T3ED_Dev4.4.0.433_Beta_05.
The stupid input of negative numbers for animations of extra objects is fixed.
Moving the speeedfile data by two points is now included. But I'm not sure if it really helps or I just got better at making the opponent lines.

[pete9516]

I don't know about that negative inputs. Never experienced it. Couldn't try out your new T3ED version yet. But i've noticed something when editing Autobahn 3.
This function right here does not work properly. The threshold for detecting non passable polygons is too high. The thin polygons won't be detected. Also when there's a tunnel section where the tunnel consists of track polygons the detection will also not work. Means you have to manually put the passable lanes in tunnel sections because the function flags the passable lanes as non passable. Other than that, this function is a god sent. Unfortunately, it does not work on heavily edited tracks as it seems
EDIT: Here's a good example (reference) of when the function decides that a non passable polygon will be passable

[JimDiabolo]

I don't know about that negative inputs. Never experienced it. Couldn't try out your new T3ED version yet. But i've noticed something when editing Autobahn 3.
This function right here does not work properly. The threshold for detecting non passable polygons is too high. The thin polygons won't be detected. Also when there's a tunnel section where the tunnel consists of track polygons the detection will also not work. Means you have to manually put the passable lanes in tunnel sections because the function flags the passable lanes as non passable. Other than that, this function is a god sent. Unfortunately, it does not work on heavily edited tracks as it seems
EDIT: Here's a good example (reference) of when the function decides that a non passable polygon will be passable

It actually worked as planned. I searched for the polygons to the right and left of the VRoad and then went one further. However, if the polygons are so narrow, one could be skipped.
The new version only searches for the polygon slightly to the left of the VRoad point. The next one is then simply taken as the right one.
It would be nice if you had given me the track and block with the problem. Then I could have tested directly there.
Try it out to see if it works better.

[pete9516]

Ah well i've forgot to put the track in here really. I'm currently working on Autobahn 3. I will test your new version as soon as possible
EDIT: Thanks so much for your quick update

[JimDiabolo]

A updated Autobahn would be nice.

[pete9516]

A updated Autobahn would be nice.

Still gotta work out the lane widths, backwards direction's brake points and i think i have to redo the VR stuff at the last section before the finish completely new. The AI just won't work like that. I have to use either the left or the right lane for both forward and backwards direction. Anything else trying to use both lanes as they're supposed to being used by traffic rules won't be possible as of now. The problem is that there are fences all the way through and i would have to remove them for a better visual experience so that cars wouldn't be able to drive through the fences and stuff.. It's a complete mess. The track itself is designed beautifully, but it won't work out for game's artificial intelligence

[pete9516]

I don't know about that negative inputs. Never experienced it. Couldn't try out your new T3ED version yet. But i've noticed something when editing Autobahn 3.
This function right here does not work properly. The threshold for detecting non passable polygons is too high. The thin polygons won't be detected. Also when there's a tunnel section where the tunnel consists of track polygons the detection will also not work. Means you have to manually put the passable lanes in tunnel sections because the function flags the passable lanes as non passable. Other than that, this function is a god sent. Unfortunately, it does not work on heavily edited tracks as it seems
EDIT: Here's a good example (reference) of when the function decides that a non passable polygon will be passable

It actually worked as planned. I searched for the polygons to the right and left of the VRoad and then went one further. However, if the polygons are so narrow, one could be skipped.
The new version only searches for the polygon slightly to the left of the VRoad point. The next one is then simply taken as the right one.
It would be nice if you had given me the track and block with the problem. Then I could have tested directly there.
Try it out to see if it works better.

Oh, unfortunately that hotfix didn't work. I still get the same results. But i can understand that it must be hard to fix something like this. The impassable polygons are very thin

[EvoX]

I need to study Pete's videos to catch up with you, I see...

[pete9516]

I need to study Pete's videos to catch up with you, I see...

Unfortunately, they are very long. But the longer they are, the more understandable the content is

[pete9516]

Autobahn 3 is giving me a hard time. I don't think i will be able to fix the AI as of yet. Blocks 123-125 are very problematic. Even the alternative route which takes much longer doesn't work

[JimDiabolo]

Here is another beta with new VRoadLanesRecalc function. I tested which Autobahn3 (Block 0 & 1) and it seems to be good now.

[pete9516]

Here is another beta with new VRoadLanesRecalc function. I tested which Autobahn3 (Block 0 & 1) and it seems to be good now.

YESS it works now!! Great update. I've tried to double check on different sections but i think it works now in a very good way. Needs further testing i guess but if i find anything i will tell it to you. Still need to try out your new way of placing VR floats. Will do that, too

[pete9516]

Oh i think i've found something already. For some reason, the function will take the next right edge here as the virtual road point middle. Maybe it is only good as an alternative way of calculating virtual passable lanes. Actually, both ways are needed and important. For normal tracks, the initial calculation is perfect, but for tracks like Autobahn 3, maybe the newer way of calculation should be used because of the very thin divided road sections. I will keep both versions for calculation i guess