What do you think can cause the "semi-freezing/jumpy" display

S

suryaprihadi

Registered
Hi All,

I am able to re-create this problem with certainty, be it in DX11 or DX9, but it takes lots of try.
This is not a micro-stuttering per se, its more like intermittent frame lost and at the same time intermittent steering input/response loss. Yes, whenever the image goes whacky/jumpy, so is my steering engagement to the game, its on-off-on-off too.

20th July test.... DX11. Single GPU.
This actual video is how the frames "jump" looks like:
The first 2 seconds of the video is the most obvious, see the blue GTR in front of me.
The last 10-12 seconds is more like stutter.

I just created the condition again 21st July 6:09 AM, DX9 . Single GPU.
This is like Star Trek Enterprise spaceship about to do Warp Speed :)
If you see at 18th seconds, the white Radical physically/visually is like moving back a few frames and jump forward a few frames..sort of. Due to Youtube limit of 60FPS max and my camera recording the display also at 60FPS, the warp speed effect in real time at 120FPS display is more comical ...LOL.


This weird condition I can fish it out, ONLY when under the following circumstances :
- Race Mode with AI, not online. I don't do online race yet.
- I keep restarting the race before completion or I keep changing cars on the same track. Say at the 10th time I restarted the race before I complete it or/and when I changed to a few different cars on the same track. The Star Trek Warp Speed, all I got to do is re-start the race +-7 times and then change cars 3 times and the final car Radical get the Warp Speed effect :) and all other choosen cars thereafter.
Yes, my steering connectivity to the game also got effected too, same on-off-on-off condition.

To remove this weird condition, all I have to do is restart my PC.
By quitting RF2 and getting back to the game again, does not remove this issue, only PC restart will solve it.

If from DX9 and then I switch to DX11 via Steam, clean all the Cbash and Shaders and DO NOT restart PC, this weird thing go away, but for some reason my PC becomes laggy taking other command like opening a folder or turning off certain windows, if RF2 launcher screen is ON. This is NOT normal, if I restarted the PC, no laggy-ness will occur.

Usually I will restart my PC when I switch from DX9 >>DX11 vice versa.

I did get this weird problem when I was testing DX9 vs DX11 during DX11 beta development, but I thought it was due to the Cbash and Shader. Only the past two days I realized how I can trigger it to happen with certainty even when I started with a clean/deleted Cbash and Shaders and not switching between DX9 or DX11.


The Radical car video was 6:09AM today 21st July, PC system time.
I recalled I got a window Aero warning about computer being slow and color will be set to basic after that.
I look for error code in Windows and I get this :
a3.jpg


The detail view for the 6:28AM one is below :
a2.JPG


I then googled on how to look for more error codes ( me IT dummy:mad: ), at 06:11:41 AM I got this logged as error, but I dont know what that means, except it is from Rfactor 2
b1a.JPG

6:11:41 AM is when I shut down the DX9 RF2 game, because after I was able to get the Radical to do Warp Speed at 06:08 to 06:09 AM, I change car to GTR500 at 06:10 AM and it does Warp Speed too :)

b1b.JPG



Few questions/possibility came to my mind :

01. Is there any temporary buffer RF2 uses when we play a game and do lots of re-start of race or change cars ? I have 32GB memory on this PC, so probably memory is not an issue. RF2 and the Window OS never takes more than 8GB of memory as can be seen on the RAM info of the OSD of the video at 7.45 GB max. How about the Shaders and CBash ? Whenever I do change from DX11 to DX9 or vice versa, I always clear these two data set , as adviced.

02. To what extend our equipment like steering via USB , sequential transmission and pedals all via USB, can cause this kind of weird condition ? For now I have :

AA. Bodnar v2, 1 USB >> to PCIE-USB card

BB. HE 3 Pedals Set, 1 USB3.0 of motherboard

CC. Steering Wheel with Pro-Race 2 module, supporting small Speed + RPM display + shift LED,
flappy pedal and 12 buttons, 1 USB 3.0 of motherboard

DD. HE sequential transmission, 1 USB 3.0 of motherboard

Note : I do not use USB port in front of PC casing, because it uses extension cable.
So all USB connection is using only motherboard ones, at the back of PC.

**. I have not yet even install this, HE Handbrake, will take another 1 USB.
** I don't have any special button box yet.

Due to my Asus motherboard bug* ( *according to Bodnar ), I was given a PCIE-USB card by Bodnar.
https://www.scan.co.uk/products/2-p...superspeed-usb30-(nec)-5gbps-pci-express-card

Without this PCIE-USB, my PC will not boot if the Bodnar USB cable is connected to the regular USB port provided by the motherboard. It will stop/freeze at that Window Logo Start Up thingy.
I must first unplug the USB cable to the Bodnar controller and when my PC already boot up well, I then connect the Bodnar USB cable, all will be good. With the PCIE-USB card, and Bodnar USB cable connected to it, I can now boot up my PC even with Bodnar USB cable connected.


03. If any of my equipment via USB, is causing this weird frame "jump" , why not it happen under normal use as in say I do 1-3 hours practice ( not race ) on the same track and same car or when I race but not keep re-starting the race or keep changing cars.

I don't know if this is RF2 issue or my PC issue or my equipment via USB issue ?
During the weird frame jump condition, I did test further by allowing AI to drive my car, which I assume all of my USB based equipment will not be giving data input to the game ... yet, still there was a minor frame jump, more of a minor stutter, but totally not as bad as when I am driving or giving input.
Hence I kinda suspect one of my equipment via USB may also be the culprit. The pro race 2 display on my steering, even when AI is driving my car, still is giving the RPM, speed , gear # and shift light data, but this is data reading from the game and not data injection to the game....right ?

If I were to test my equipment one by one to troubleshoot USB connectivity issue, the steering and the HE pedal is what I can not disconnect for the troubleshooting as I need minimum these two to drive. How do I then dismissed these two items as a non issue ? I don't have any other steering wheel or pedal set to try.:(

Any troubleshooting suggestion guys ?
Thanks very much ....in advance.
 
The windows error you posted happened with the DX9 dll, have you checked the windows logs to see what gets logged when you reproduce the error in DX11? (Sorry if I missed the answer in your post, but it is very long to read again :p)

If I understood correctly, this only happens when you restart a race several times and / or switch cars? If so maybe try keeping the task manager open and checking the memory consumption from the time you open rF2 until you see the issue.

Compare that memory consumption to a race in the same track / cars where you don't reproduce it, if the difference is big, or if the memory consumption keeps increasing every time you restart to switch cars then that could point to a memory leak.
 
The warping is usually what happens when the CPU is maxed and falls out of real time. I would suggest checking out a few posts regarding the CPU falling out of real time and troubleshooting with those suggestions. I think there are quite a bit of things to try, I used to experience the warping when I had a high amount of AI and settings high but upgraded by CPU not too long ago and that cleared the issues. You also might want to check the time scaling, someone mentioned before that upping that could cause quite a bit of calculations which could potentially drive the CPU up as well. I think it's Ctrl+C by default to view the physics which is the purple bar and if that hits the max, it will fall out of real time. I think I mentioned everything correctly, if I misspoke, someone can correct me.
 
^^ CPU. You need a fairly strong single core performance from CPU to run rF2 with many AI cars. By fairly strong I mean Intel i3/i5/i7 or AMD Ryzen class. Any i3/i5/i7 will do even if it's 5 years old. If you already use one of these CPU's, then it's a bug elsewhere.
 
It is this one "Bodnar v2, 1 USB >> to PCIE-USB card " which is causing the problem . Just look how ms fly up in your monitor D3D 9 or D3D 11 on the right side of FPS counter. If you like to have something else then GFX cards in your PCI-e slots you must buy different MBO then you have.
 
This is definitely CPU related.
In both videos you can see that your CPU is only 15-20% use. Even the 1000$ I7's and top end Rhyzen's use more than that.

What is your CPU?

32 gb of ram is way overkill... unless you render 3d movies while designing game-physics while watching a movie while playing a game and surfing the web all at the same time. :D
Even I only have 8gb and an FX-8320e @ 3.4ghz, and I just ran a 15 lap, 40 AI race at Silverstone with Apex GT, Simtek WS, and the URD C7.
I did notice the same stuttering you have, but only very briefly, and at the start, and my CPU use was 100%. :cool:

Cheers
rocket
 
rocketjockeyr6 said:
This is definitely CPU related.
In both videos you can see that your CPU is only 15-20% use. Even the 1000$ I7's and top end Rhyzen's use more than that.

What is your CPU?

32 gb of ram is way overkill... unless you render 3d movies while designing game-physics while watching a movie while playing a game and surfing the web all at the same time. :D
Even I only have 8gb and an FX-8320e @ 3.4ghz, and I just ran a 15 lap, 40 AI race at Silverstone with Apex GT, Simtek WS, and the URD C7.
I did notice the same stuttering you have, but only very briefly, and at the start, and my CPU use was 100%. :cool:

Cheers
rocket
Click to expand...

It is not the CPU he has i7-5960X. My i7-5960X@4.4Ghz usage is about 15-20% also.
 
Ctrl-C would show the CPU running out of time in game terms, so the question is why is the CPU 'suddenly' struggling. Though maybe Ctrl-C will indicate something else when the game is running fine (it may get close to full more often than this behaviour would indicate)?

'per-second' cpu occupancy is fairly irrelevant in rF2.
 
If you look ms on the right side in hes FPS counter you can see that it jumps up to 100 ms. I bet its that Bodnar v2, 1 USB >> to PCIE-USB card which is causing the problem.
 
Was windows downloading one of its updates during this time? That same question can apply to much of the rest of your software. Drive speed and fragmentation could also be a factor. These things are not CPU dependent. Also make sure there is no dust, loose wires/ram/cards. Make sure you have not burned up the thermal paste on the processor. Check the power supply, make sure it is clean and does not smell like burning silicone. If you turn your computer off between uses, that could also cause this as the computer has to get "ready" for rF2 (or any other use) and that might not be at the same speed as you think it is ready. What I mean by this is if you do not allow the computer to do everything it needs to do (wait for all drives to stop spinning and/or seeking, wait for all background processes to load) before you start steam and rFactor, you can have this happen. This does not happen often if your pc is always on but that is not always an option for everyone. Also steam itself needs to be restarted periodically and that can also do some of what you described if you leave steam on all day. There is also the Microsoft Indexing Service which keeps track of every file and other things. This process alone can cause these problems if you have a lot of files or you move/copy/delete files often. This process requires the drive or drives be accessed even if you are doing something important as rF2. None of this is CPU or GPU or ram or os version dependent. None of this may be your issue but hopefully something in here is helpful to you or someone else.
 
Last edited:
Thanks everyone, I did some more test and I think I am getting closer to the culprit...maybe.

I am on 8 core i7 5960X, clocked to 3.5GHZ using Asus mobo setting which call it 16% overclock.

This PC used to be a video editor so 24GB is borderline, hence 32GB was installed.

I read thru the memory leak as suggested by Camilo, googled microsoft guidance , to keep it short, it was not the cause nor a memory leak is occurring. The Afterburner logging of RAM is equal to what Window Task Manager showing. So I was able to see the video clips of normal play vs warp-speed/jerky play RAM use.

I use DX11 again this time. Tried hard to re-create the problem but failed. Keep restarting race and changing cars, problem did not surface.

So I raced a 10 lap instead, but when the race completed I did not enter the pit, I went for a fun lap.
By the extra 2nd fun lap , I noticed very minor stutter image wise, but steering works well. I then restarted the race and this time I get super jerky stop and go image, but no Star Trek warp speed :) Even AI driving my car its the exact same jerky stop and go...dang, this is unique and maybe a totally different problem ?

I shut down RF2 game and went to Resource Monitor.
I then use Atube screen recorder to record what Resource Monitor showing, I can't log Res-Mon :)
The highest CPU use at 6 was System Interrupts and a screen recording name Atube ( eWorker.exe ) which was only using 1 to 4.

I am seeing CPU Total showing blue line which is the CPU frequency being maxed out ,but it is hill and valley and not a constant hill peak blue line. Odd, Rfactor 2 is already off. I saw CPU 14 hill and valley equal to the blue line.

a1.JPG


.
a2.JPG


I been suspecting the Bodnar might be the culprit. It seems so for now. When I terminated
the Bodnar Configuration software, the System Interrupt gone too
a3.JPG


However, I just remembered that the Bodnar Configuration software is only needed when adjusting steering center and angle, and no need to run it during a game as the game and Windows7 recognize it as Steering Wheel, as long as the Bodnar controller box is powered ON.
a5.JPG



I made a mistake, I should not have terminated the SimDash software so soon until I see CPU 14, and at least try playing the RF2 again. I also suspect SimDash software may be a potential culprit because I need to turn it ON for it to work feeding RPM & etc data to my steering wheel.
a4.JPG


For now, suspect #1 is the Bodnar Configuration software, I need to test again RF2 without the Bodnar software running.

By the time I terminated Bodnar Config software and the SimDash software, without PC reset, the game resume normal smooth image/play.

I think this is somehow a CPU related jerky/warp speed, but caused by that Bodnar Configuration software. Maybe the Bodnar Config software, under certain condition, does not allow Rfactor 2 to use a particular CPU thread, while RF2 itself did not seek to use other CPU threads and these two softwares fought each other for the same CPU thread.... LOL.


Ari,
On the PCIE-USB I am using for Bodnar as potential suspect, let me run more test without Bodnar Configuration software running.... if jumpy/warp-speed/jerky problem still persist, I will narrow it down to this small board. However this small board is on a baby PCIEX1_2 and it does not interfere with my currently single GPU set up or 2-way SLI ( PCIEX16_1 and 16_2 ) when S397 is SLI ready.

Again thank you very much to all who gave suggestions.
 
Will Mazeo said:
Btw did you set the use thread to true in controller.json?
Click to expand...
I will check, I never mess with that one though.

UPDATE : That setting is no more available in my JSON file.
I read it used to be called "Sim Processor Thresh".
Did S397 removed it or renamed it ?
I tried CPU, threading, hyperthreading,multithreading, processor... all NIL.
Only CPU search yield this :
"Inactive Sleep Time#":"Number of milliseconds to sleep each message loop if the game is not the active application (-1 to disable). will give more CPU to other apps when minimized, etc.",

Old info I got on multithreaded

Update 31 (Build 946) Changelog:
FIXES:
————-
– By default, the multithreaded physics are turned off for now. Set the PLR file variable “Sim Processor Thresh” to 3 to turn on, or 255 to turn off (note that we don’t allow them to be turned on for dual cores). Please do NOT use +procmask for the purpose of disabling multithreaded physics anymore.
 
Last edited:
Dang, Sorry Will, you meant controller Json and not Player..Json... LOL. I hought you were speaking of multithreading on CPU.

It is set as TRUE. Below :

"Use thread":true,

"Use thread#":"Use a separate thread to issue FFB commands which may block with some drivers"
 
Have you removed/disabled plugins you don't need, to test if they have any effect? Those repeated race restarts could potentially screw something up, depending what it's doing and how it's doing it. I wouldn't rule out my own plugins doing something strange in that scenario.
 
Hi Laz,
I removed all plug in since trying often DX11.
Some DX9 plug in crash DX11 :) , but yours does not.
So I have removed them all for sometime now.
Thanks for the tip
 
You must log in or register to reply here.
Back
Top