I wonder, perhaps then Shadow cars has different AI setting to replicate the alien simracers better, thus it would make sense if they use more road then, or should I say - more barriers
That’s really odd, it’s almost like the FASTEST line is shifted to the right. I’ve checked the track in dev mode and there are no specific named paths so it’s not that that’s causing the issue.
Does someone know a way to replace the AIW of S397's mod with that of SRHF1's version? Is it locked out? That would solve this problem right away as theirs works with any cars (or skins).
I don't know too much about AIW but have noticed on other tracks that different cars drive different lines. I was wondering if maybe the cars have a different reference point that reacts to the race line. Say car A has the reference at right front tyre, car B has reference at center and car C has reference at rear left? Like I say I don't know enough about modding but am curious to find out if that's a thing.
@JimmyT I don't think so, but rcd files can determine how far ahead cars look which can have quite an influence on their line approaching and through corners, especially slower ones.
That makes sense. But it does not explain why the SRHF1 mod of Monaco has no issues with the same cars as S397's version does. Something must be incorrect in the making of the file. Coutie posted that they used the old ISI F1 2012 to make the AIW. I also know nothing about how to make one of these the same as JimmyT but does it make sense to make it with an almost 8 year old ISI mod instead of a current S397 one? When someone makes an AIW, do they try multiple cars on it to test it out or just one. I do not know, do you?
Won't work because the two versions of the track will have different coordinates. However, you can edit the AIW unless it is locked...
AI learning, would that be any good in this type of situation? This is an old post so dunno if it’s still current => https://forum.studio-397.com/index.php?threads/ai-learning-step-by-step-guide.47657/
As I said earlier, if the shadow cars have a different class the rcd may not apply to them. I'm just guessing here.
I have tried training the AI once and did not achieve good results. But the AIW on the track I tried it on was extremely poor.
Based on the video it looks like the car is too wide for the racing line which makes the AI constantly and laterally hit the barriers. The only way to fix it is to record a specific racing line in AIW file for this F1 car in order to keep the center line points clear from the walls and guardrails. . If narrower cars are running fine this is a good possibility. Check the AIW file using MAS tool and look for specific car lines. If it was a problem with corridors all cars would have the drunk behaviour as shown in the video. The problem is, I don't know if those DLCs tracks are still editable in ModDev anymore, if it be the case then would be easier to set a custom fix. If they're not longer editable you'll have to pray for a miracle.
Yes it would, but... type 2 training would be required and that only works on an individual livery. With the MP4/8, it's easy enough to train one and copy it to the other very few liveries. However, since it is a favorite track for GoldenBear, he's likely to want all cars he uses to be trained and that takes time.
That could possibly be why the Senna also crashes on the track being a wide car although it only does in the tunnel at the apex. The F1 1979 and F1 2013 3rd party mods also crash as well as some other tin tops. The GTE cars from S397 seem fine on it.
Car width should adjust automatically, it seemed to adjust when I tested with cars including the boxmaster.
Greeting! Last night I was testing Circuit d' Ajure with a new 8GB video card running a field of 19 Tatuus (mix of classes) vehicles. Because I was new to the track, I was doing most of the crashing. Don't recall any "bottlenecks" in the tunnel. When I ran the "tin tops" I did have major AI problems. I haven't done any modding for rF2 yet, but when working with AMS one could make some adjustments in the srs file (the old rfm rf files). Sample: OvertakingRadiusRiskBypass = 600 //radius in meter, from this radius and above the AI will be bypassing, bellow it will not OvertakingWaypointsAhead = 15 //Number of waypoints ahead of the current one to calculate the radius Does rF2 have something like this, or is even the issue? Just throwing against the wall here. Cheers!