Absurd Problem That Somehow Makes Me Slower

S

SETHONMETH

Registered
I discovered this problem this January. In a nutshell it makes it impossible to post competitive times. I can still post very good laptimes, but my best laptime will max out at some arbitrary limit that falls short of the fastest lap that is possible. Also there is something going on with the physics that makes it near impossible to take steps forward developing a setup. It's like bizarro world when I develop a setup and nothing seems to make sense. I originally had an i5 6600k, ASRock Z170 Pro4, 16GB DDR4, GTX2070 and upgraded to an i7 7700k because I thought that this upgrade would solve the issue. I'm not sure it did. It makes the issue less extreme and I've noticed and increase in rear grip and setups that I developed while on the 6600k don't work at all on the 7700k. I forget where I read this, but somewhere I read that if you have this issue then your top speed will be reported as a different speed on Liveracers than the speed the game shows you on track on your speedo. Also I read in a thread to force 32bit to solve this issue which, even if it does, I can't do Nordschleife in 32bit.

So before I buy an entirely new Ryzen PC, I'd really like to do some fact checking and see if you guys at S397 have tests I could run to figure out what is going on. Maybe this is an issue with the way rFactor 2 is installed. But one of the first things I did on the 6600k was reinstall and test the fresh install; it made no difference. The game runs AMAZINGLY smooth at 144 fps on my triple screens. From my perspective, you would never guess that there is a problem. Even on the 6600k I would have never guessed this was the case. Sometimes certain cars, GT3 is maybe the worst offender, will give me an extremely rough time and I'll spin for no apparent reason and I'm left with no explanation.

My PC is amazing and I don't see any reason as to why I would have this problem when other people with significantly less powerful PCs have no issues. Could it be my motherboard? Whatever the case it seems like something I could never figure out on my own. I've tried all different combinations of CPU and GPU clock speeds, game mode on and off, disabled C-states and any such BIOS settings. I would really appreciate if somebody could maybe give me some troubleshooting steps to narrow down where the problem is coming from and maybe rule out buying an entirely new PC. I understand that this problem is subject to whataboutism and that this is an extremely rare problem in rFactor 2, but I'm fairly certain it is a real issue for me.
 
Last edited:
Indycar at St. Petersburg (InternationalSimRacing):
1st run rFactor 2 reports 158.84KPH, Liveracers reports 280.2KPH
2nd run rFactor 2 reports 167.60KPH, Liveracers reports 282.7KPH

The other lap I put in in a different server has apparently been removed but I can easily provide more if it's necessary.

(Made a mistake reporting my top speeds and edited it to the correct values, which are actually average speeds and not top speeds)
 
Last edited:
If you click "expand" under the time table in session, then right click your name or any driver's name it lists the sector times of every lap and the top speed recorded on that lap.
 
I was under the impression that it was not the peak speed (was told/saw something a looong time ago, can’t remember exactly what it was) a good way to check is do a full lap with pit limiter engaged and see if rF2 and liveracers report the same speed
 
This is a good idea. It could be a 'speed trap' that is set at a specific location near the end of the straight and not set to record the overall top speed. I'll post my results when I'm done.
 
OK I did one run with the limiter on but I couldn't see the results in Liveracers because only the top 3 were displayed. I'll be back shortly when I do a successful run. What I did find, however, is that I misreported above the speeds that rFactor 2 recorded. I thought it was giving them to me in MPH and I was converting to KPH by hand, but I've checked my settings and I definitely have KPH enabled and on my speed limiter run it was clearly in KPH.

So the actual results to my first test are as follows:
Indycar at St. Petersburg (InternationalSimRacing):
1st run rFactor 2 reports 158.84KPH, Liveracers reports 280.2KPH
2nd run rFactor 2 reports 167.60KPH, Liveracers reports 282.7KPH

Which is a much larger difference. If the results of my speed limiter test are inconclusive I'll use a telemetry plugin to check top speeds.
 
I reckon the rF2 speed might be the average...as a check divide the track length by the lap time and see if that comes close to 158kph.
 
Okay I was able to get a run in that registered on Liveracers.

OW F1 at Catalunya (RSL eSports):
rFactor 2 reports 80.89KPH, Liveracers reports 80.7KPH

And I couldn't help but think that maybe what I read about there being a discrepancy in the top speed reporting was written by somebody who didn't know what they were talking about. It seems odd that rFactor 2 would report a higher top speed if it was a speed trap and not an overall top speed being recorded. I can't remember what the results to my 6600k testing were, but in my testing today the speeds reported in rFactor 2 were always slower.

I have to say that I can't draw any conclusions from these results and it would be nice if somebody knew a different way to test if there is a problem. What is not in question is whether there was a problem on my 6600k. I just want to figure out if I still have the problem or not. There being a difference in top speed doesn't explain why cars behave strangely or why setups did not transfer from when I had the 6600k or why there was a difference in rear grip levels or why I was able to post such faster times with the 7700k.
 
Last edited:
You're definitely right about the average speed. 2.9km/0.0175125hrs=165.596KPH
(from my 2nd run at St. Petersburg)
At least we cleared something up
 
I'm remembering now that when I first found the problem it was because I noticed that the top speed on my speedo while on track was different from Liveracers when I was tuning a GT40 in the Ford v Ferrari server. So while I botched my testing today it might still be worth looking into with a telemetry plugin. I'd really like to hear more info on what might or might not be the case before I do a bunch of testing. It will have to wait until tomorrow if I'm going to do more tests.
 
Ctrl-C while driving and look at the horizontal purple bar. If it's not touching the right hand edge (which will make the histogram below do a 'wavy' thing, and it will have a few small bars all over the place) you won't have any sort of slowdown or speed issue.

My DAMPlugin mentioned earlier also logs any realtime loss, can be easier to spot than the live graph (though when I had CPU issues, I could see it live). Finally, your previous CPU was fine and your new one is even more fine, so I doubt you have any issues.

As far as grip, if you're not doing skidpan tests and measuring g-forces (which then needs smoothing or perfectly smooth skidpan, plus someone else to do the same thing and find a measurable and consistent difference...) it comes down to feel. I would look at display and input lag first, that can give you a disconnected feeling with nothing else apparently wrong.
 
Last time I tried Ctrl-C it did not pull up the purple/green bar. I thought that maybe in the latest built it had been taken out. Maybe I need to enable something somewhere. I just pulled up a session for the sake of this post and Ctrl-C while on track did not pull up the purple/green bars so maybe there is a step I'm missing here. However the 6600k was not fine and pegged my CPU usage at 99% while in rFactor 2. The 7700k stays below 25%. There could have been some kind of problem with my individual processor, as it never overclocked well and I couldn't get it stable above 4.4GHz with a beefy af water cooler. At the very least I lost the silicon lottery. There still shouldn't have been a problem but somehow there was. The biggest difference was my laptimes in everything I tested improved by about a second when I ran the same car/track on the same day where the only difference in my PC was the processor being switched. Also before I bought the new one my laptimes improved, but not by as much, just by forcing 32bit. Everything else is a bit arbitrary, but I stand by my findings here with the rear grip and setup stuff. How do I know the problem is fixed and isn't just smaller? The problem would make sense if it was caused by the CPU being overloaded. Does the purple bar just show CPU usage or is it a more complicated diagnostic?

As far as input lag goes I'm pretty positive I have almost none. My monitors have an extremely quick response time and there are no issues with my graphics card. The game runs amazingly smooth apart from pretty dang small stutters on certain tracks. There could have been a problem with the way force feedback was generated from the physics and that would explain both why it felt like more rear grip and why setups didn't transfer and why setups were impossible to develop.

I'm probably going to be grabbing your DAMPlugin tomorrow and comparing the top speed in it with the one in my in session replay with Liveracers. That should at least clear up any questions about top speed discrepancies.
 
There is no difference in physics between different CPU's. This issue was solved in the late 90's back when some game physics were tied to CPU clock etc. Nowadays all games run independent of CPU clock and number of CPU threads, just imagine the nightmare in professional online esports competitions if one CPU produced better lap times than the other. Literally the only possible way rF2 physics get messed up is if the CPU usage goes to 100%. What happens then is that the game falls out of realtime, which you will notice immediately, as driving becomes impossible. Under any other conditions, the physics will work at the same rate.
 
Marcel Offermans said:
What wheel and pedals are you using? One common cause for high CPU usage is a wheel driver that is relatively slow receiving FFB commands. A common solution for that is to search your controller.json file for an entry called "Use Thread" and set its value to "true".
Click to expand...

Maybe you should try this? Could be your CPU is hoging for any weird reason?

Marko
 
stonec said:
just imagine the nightmare in professional online esports competitions if one CPU produced better lap times than the other
Click to expand...
I remember this exact thing happening in PCARS 1 when certain CPUs were overclocked. But that game is a bit of a nightmare in general. I don't think other processors are faster, just that mine was making me slower. I've never seen any evidence that a processor could make somebody faster than the norm in rFactor 2.

It could have been the "use thread" option that caused the problem, but I've had that set to true for a while now, since I went through my player.json and changed everything I could when I got my new processor. I use a Fanatec CSL Elite wheel and pedals. It's sounding like my CPU hitting 100% on the 6600k was causing the issue. Are you guys unable to pull up the purple bar? I'd be a lot more comfortable pulling that up and not looking at external CPU monitoring. Or is there another way in rFactor 2? I know there is some kind of performance recording that you can enable for a strangely short period of time when the Ctrl-F dialog is up, but I don't know how to view the results of this.
 
You must log in or register to reply here.
Back
Top