Car Goes to Space When Doing AI driver Change From AI to Me.

Hello all. I was doing a recreation of the 24 Hours of Daytona on the DX11 mod converted by SJ in the RSS Cadillac DPi. Everything was going great, that is until my AI driver pulled into the pits after his stint, and stopped in the box. I pressed the 'I' key to do a driver swap. When I did, I heard a sudden crash noise, and saw my car being hurled into the air. I could not press escape, so I had to exit my game, therefore meaning my replay didn't save. This is the 2nd time I've attempted this race and failed to make it further than halfway (the other reason was just the game crashing after an issue with the AI on the mod track itself). Does anyone know what could have caused this, or is this simply a major bug that S397 would have to fix? I've done plenty of other driver changes with this car and this track. Has anyone else had this issue at other tracks or just at this track? I will post a clip of it once my live stream uploads.

 

as you can see I clicked the AI driver swap and it said changing in 3, 2, 1, and then yeah... that happened.

 

Okay, I've done some testing. This is what I have found:

- Occurs not just at Daytona in this race situation in this car, tested at Sebring and same thing occurred.

- Different cars, tested LMP2 and Porsche GT3 Cup, same things happened

- Race length does not matter, same thing occurred in a 1 hour race.

- Number of cars does not matter

- Type of session does not matter, same bug occurred in practice

- This DOES NOT occur when I switch from AI driver to me ANYWHERE BUT THE PITS ON THE JACKS. When I take control when the AI is on a straight, everything is fine. If I take control in the pitlane before or after stopping in the box, everything is fine. It is ONLY when the car is up on the jacks and I ask to change from AI driver to me that this occurs.

Hopefully this is helpful to anyone in the time being who is trying to race with AI driver swaps. Hopefully S397 will see this and take note of it .

 

I didn't know the AI would pit on their own. I'll have to check that out.

 

Yeah It's really cool when it works. For me my AI will come in to the pits normally once they are low on fuel. Just keep in mind the stuff I said earlier obviously.

 

OK, just did a two hour race at Sebring. I short pitted twice to hand off to the AI. The first time the AI came in, I was looking at the car from an external view. Didn't notice anything strange. I switched back to cockpit view & pulled away. I again only drove a few laps and came back in, making sure I was in the cockpit as the car went up on the jacks. No surprises. The AI driver drove away, I accelerated time again to use up his fuel. As he came in the 2nd time, I was in the cockpit view & waited til the car was up on the jacks before pressing the swap key and there was a decided THUNK/Crash sound as if the car fell off the jacks. But the timer kept rolling and I was able to drive away right on schedule.
Marc G has been preaching replacing the player.json file lately and since I had recently done a complete wipe and fresh install, my player.json was nearly clean.(i edited some engine start lines and fps changes. That may have been the difference. My wheel jolted, but my car did not explode up into space. Give that a try, rename or move your player.json and let Steam download a fresh install.

 

https://forum.studio-397.com/index....bang-when-ai-pit-and-player-takes-over.66763/

 

Yep, exact same thing.

 

Okay forgive me because I am a complete novice when it comes to this kind of thing. Where is the player.json file located? Would it make a difference to move it or rename it and reinstall the game if I have not made any changes to the player.json file in the past? If so, I assume this would mean that the reinstall would have the same values as the original player.json. I will see what you say and reinstall and verify game files anyway since its always a good idea to do that every once in a while.

 

C:\Program Files (x86)\Steam\steamapps\common\rFactor 2\UserData\player
When I rename something, I just change the extension. So edit player.json to player.jsXXXon. That way, your re-named file will be right next to the new file should you need to change back to the old file.
Everytime rF2 is closed out, there are several .json files that are written to by the program before they too are saved and closed. Any sort of hiccup in the shut down procedure could cause some sort of corruption inside the file. ss

 

Got it, makes sense. I'll give it a go.

 

Yeah I did this and the same thing still occurred. I will just take over from AI when driving down the pitlane!

 

There definitely is a thump there, but odd(not really, this IS rF2 after all) that the thump affects the car differently. I wonder if a help setting or some other option may be amplifying things.(like steering help or some other user selected option.

 

This sort of explosion/launch happens with too-hard bump rubbers or even springs, and certainly in the past happened when cars got inside each other and the game wanted to separate them (makes sense that the more they're overlapped the faster you want to move them apart - but only to a point...).

Seems like either the AI car and the player car are co-existing for a moment, or the lifted car isn't being taken into account and suspension positions are very extreme which leads to an overreaction (force squares with distance). The second could well vary between cars.

 

It very well could be. When I get time at some point I'll take the time to cycle through some of the assists and see what happens.

 

I

Besides the RSS DPi, I have just used default setups for the tests, but I will try softening the springs and playing with the setups later to see what comes of it.

 

I should have been clearer on that, was speaking from a modding viewpoint. If it's something along these lines it's probably the bump rubbers which you can't adjust, and really it's just a game bug. Best to just avoid it as you are.

 

I thought I would try this.

2nd and 3rd time I tried it crashed the computer.

 

I liked the idea that it was the suspension but that video (I have seen the same thing) shows the car moving instantaneously and that says to me it’s a pure software bug, something like inverting or zeroing the car's X co-ordinate for example.

 

Got it yeah that's what I thought. Thanks.