Question Continuous stutter in games when using mouse

[Centel Cesurus]

Hey all,

For a while now I have been suffering from what I can only describe as some form of ‘stutter’ or ‘micro-stutter’ across multiple different games (albeit with slight variations in how this issue appears visually) whenever using my mouse in-game, particularly when making rotations. This is even more noticeable when using both my mouse and keyboard to move and rotate around an in-game environment. Using only my keyboard, however, to either move: forwards, backwards, strafe left or right has not produced any problems and programs run very smoothly with none of the aforementioned issues occurring. The same applies for when using an Xbox one controller via physical connection, no issues occur.

This user posted that they were suffering from an issue, very similar to what I am currency experiencing, back in 2013: Games stuttering when moving mouse | Tom's Hardware Forum (tomshardware.com) (My apologies if I am not supposed to link back to previous posts, this is my first time actually posting on the Tom’s Hardware Forums)
The post above perfectly describes what I am facing. I can’t definitively say that this is technically ‘stutter’ or ‘micro-stutter’ but I can agree that “its this slight tick-tick-tick-tick as [the mouse] moves”, “no jerky uncontrollable movement, just a stuttering”
What makes this most frustrating is that it is subtle but to a degree where it is still definitely impactful on the gameplay experience.

Impacted Games

The issue does vary in how it appears across different titles, for certain games this issue is far more prevalent than for others. Just to provide some more in-depth information I’ll list below all Titles impacted:

• WoW: The issue is incredibly noticeable with a continuous micro-stutter with mouse inputs that is just noticeable enough to be incredibly frustrating. Practically a complete lack of smoothness.
• BeamNG.drive: Problem is highly noticeable
• Minecraft (Java): Problem is highly noticeable
• Call Of Duty Warzone: Stutter isn’t as noticeable but is certainly present
• Automation: Stutter isn’t as noticeable but is certainly present (Especially in relation to background scenery)
• Grand Theft Auto V: Stutter isn’t as noticeable but is certainly present

With CSGO when using the raw mouse input setting the issue is incredibly subtle and barely noticeable. However, quite interestingly, when this setting is turned off there is a bit of stuttering every couple of seconds. This sort of intermittent stutter is practically a separate issue from the problem I have been describing.

These titles don’t appear to have been affected:

• Rust
• Minecraft (Bedrock Edition)
• Valorant

Attempted resolutions

I want to include as much potentially helpful info as possible so I will clarify that I have already attempted all suggested fixes in the post above along with numerous others suggestion I have found:

• Tested multiple different DPI and Polling Rates across effected titles
• Ensured mouse is connected directly into the back panel connector
• Used different USB ports, even tried using a USB 2.0 over USB 3.0 to see if it would produce different results
• Tried using a different mouse (Older model Logitech G502 Proteus Spectrum)
• Uninstalled and Reinstalled both the mouse driver and Logitech G Hub software
• Tested using the mouse without Logitech G Hub software installed
• Disconnected any additional peripheral devices from computer
• Used different surfaces (Wooden Desk, Corsair MM350, Logitech 840XL)
• Ensured Xbox game bar and Nvidia In-game overlay were disabled
• Ensured that I am running the latest Nvidia GeForce driver (Version 471.41 as of writing this)
• Changed pointer scheme setting to “(None)”
• Deleted Realtek Audio Driver
• Checked that Cortana was still disabled
• Checked that Enhanced pointer precision was still disabled
• Checked all impacted games were updated to latest version

I was suffering originally from a few other computer performance related issues which led me to perform a clean installation of Windows 10 early this month on the 4th of July. Similar problems were present before the clean install, however, I wasn’t playing as many games at the time aside from CS:GO and GTA V and so I can’t really pin-point if this clean installation is a contributing factor to this issue.

Because of this clean install I do not have: RivaTuner, MSI Afterburner, CPU-Z, GPU-Z or any similar software monitoring my computer as I did previously.

I’ll also include a few other changes I have made in general to the system since the install (I have these all documented on a .txt):

CP: Control Panel
NCP: Nvidia Control Panel

• Installed the latest BIOS version for my motherboard (Version 1301), Model: Asus ROG STRIX Z270F (Pre windows clean install)
• Turned on GPU Scheduling (Win Settings): 05/07/2021
• Turned on “Game mode” (Win Settings)
• Turned off Background app refresh (Win settings)
• Selected “High performance” power option (CP)
• Turned on low latency settings (NCP)
• Set Preferred refresh rate to highest available (NCP)
• Power management mode set to Preferred maximum performance (NCP)

(In services)

• Fax disabled
• Windows insider service disabled
• Remote desktop configurator disabled
• Touch keyboard configurator disabled
• Application layer gateway service disabled
• Windows image acquisition disabled
• Connected user experience disabled
• Retail demo service disabled
• Windows media network sharing disabled
• AllJoyn router service disabled
• Windows biometric service disabled

(For G-sync purposes)

• Vertical sync turned on (NCP)
• Set max frame rate to 142FPS (NCP)

Additional steps taken / tests

• Automation Devs recommended I try disabling Anti-Aliasing, changing this made no difference.
• The WoW forums recommended two different console commands as well as “/script SetCVar("rawMouseEnable", 1)” to enable/disable raw mouse input. This did not resolve the issue (However I doubt this worked as checking the WoW Config.wtf did not show any additions or changes)

Hardware related tests:

Using 3D mark:

• Timespy showed average performance results relative to similar systems
• Stress test stability percentage exceeded 98% and didn’t throw up any concerns
• CPU Specific testing didn’t come up with anything particularly concerning
• PCIE Express feature test showed bandwidth to be 3.06GB/s

Using the windows memory diagnostic tool showed RAM to be running as expected
Samsung Magician showed my 870 EVO SSD to be operating as expected.

Pre-existing Issues:

• My GPU (An Asus ROG STRIX GTX1080) has had difficulty constantly outputting whenever connected to the top PCIE slot of my motherboard. Having the GPU connected to the bottom PCIE slot on my motherboard seems to have resolved the issue. Checking both the GPU and PCIE slot hasn’t revealed any apparent damage to points of connection. Perhaps this is somehow linked to my issue? I can’t be sure as, again, this only occurs when using my mouse.
• Despite specifically returning my GPU to default settings via MSI Afterburner, my GPU still appears to run 200 MHz over the standard 1733 MHz Boost clock specified on the Asus website.

Specs:

OS: Windows 10 Home
GPU: Asus ROG Strix GTX 1080. GPU boost clock: 1900 MHz
CPU: I7-7700K @ 4.5GHz using “Turbo mode”
Motherboard: Asus Rog Strix Z270F Motherboard
Ram: 16GB of DDR4. Ram speed: 2333MHz
1st SSD: Samsung 870 Evo 500GB
2nd SSD: Samsung 850 Evo 500GB
CPU cooler: Noctua NH-D15
PSU: 850W (80 Plus Gold rated)

Main mouse: Logitech G502 Hero
Second mouse: Logitech G502 Proteus Spectrum
Monitor: Asus Rog Swift PG279Q 27 (2560 x 1440 resolution @ 144Hz)(100% DPI Scaling)

I will note that the majority of these affected games are stored on my new 870 Evo. (I’ll uninstall WoW and install it onto my 850 Evo to see if that makes any difference)
I will also note that I do have McAfee anti-virus installed and running on my system, I am aware of the issues it can cause but over the 5 or so years I have used McAfee I haven't experienced this issue until now.

My apologies for the long read, just want to provide as much potentially helpful info as possible. I’ve already had two very Tech experienced friends look over this for me and we still haven’t found a solution.

Any and all help/suggestions would be greatly appreciated!

Cheers, guys

 

[Centel Cesurus]

Update:

Issue is still unresolved but I have tried the following recommended suggestions from members of the OCUK forums:

Try lowering resolution to 1080p
Try disabling McAfee antivirus
Tried using different brand of mouse (Dell) but I'll see if I can get my hands on another gaming mouse to test.

I'll keep this thread updated about anything else I can find.

 

[Centel Cesurus]

Still nothing yet, I'm afraid. If anyone is interested I'll link to the OCUK forum thread where you can see some of the suggestions. If you are suffering from this issue perhaps one of the recommended solutions might be able to help? Should also give you an insight into what I've already tried.

Overclockers (UK) thread: Microstuttering in games whenever using mouse | Overclockers UK Forums

 

[gamerk316]

The obvious thing to try (just for testing) is to use a normal USB mouse and see if the problems go away; that would at least isolate the problem somewhere within Logitech's software package. Which you should make sure is updated to the latest version, and update your mouse's onboard firmware if the Logitech software detects it's out of date.

Other then that, try setting a lower DPS polling rate on the mouse; for whatever reason it might be lagging the system (granted, it shouldn't, but you may as well test this out.)

 

[Centel Cesurus]

The obvious thing to try (just for testing) is to use a normal USB mouse and see if the problems go away; that would at least isolate the problem somewhere within Logitech's software package. Which you should make sure is updated to the latest version, and update your mouse's onboard firmware if the Logitech software detects it's out of date.

Other then that, try setting a lower DPS polling rate on the mouse; for whatever reason it might be lagging the system (granted, it shouldn't, but you may as well test this out.)

Thanks for your response!

Although I have already tested with a normal Dell work mouse (with which the issue persisted) it was wireless. I agree that it would be worthwhile testing with a different wired USB mouse as the Dell mouse I did use didn't seem particularly good. I'll see if I can get my hands on something!

In regards to Logitech G-Hub: I have already checked for updates when I uninstalled and reinstalled the software alongside the mouse driver. The software is updated to the latest version as is the mouse's onboard firmware.

As documented in the original post I have already attempted lowering both the polling rate and DPI; sadly such changes didn't have any discernible impact.

 

[Centel Cesurus]

After being recommended to use Event viewer on the OCUK forums I came back with these results:

Critical: 1 in the last hour / 2 in the last 24 hours / 3 in the last 7 days
Error: 2 in the last hour / 5 in the last 24 hours / 34 in the last 7 days
Warning (This one looks quite high): 10 in the last hour / 34 in the last 24 hours / 1345 in the last 7 days

1345 warnings in the past 7 days seems quite high, especially relative to the fact I've only had 34 warnings in the last 24 hours. Could this be a red flag?

 

[gamerk316]

Possibly, but without knowing what's throwing the warnings, hard to say. Does seem absurdly high though. Seems to indicate something is wrong at the very least.

 

[Centel Cesurus]

Possibly, but without knowing what's throwing the warnings, hard to say. Does seem absurdly high though. Seems to indicate something is wrong at the very least.

I'll list what are causing the warnings now (My bad, probably should have done that beforehand)

I can go into more detail if needed but the following are the main culprits over the past 7 days:

Critical: 'Kernel-Power' (All 3 critical events caused by this source)
Error: 'Kernel-EventTracing' (7 Errors), 'DistrubutedCOM' (7 Errors)
Warning: 'DistributedCOM' (1030 warnings), 'Kernel-PnP' (231 warnings), 'CertificateServicesClient-AutoEnrollment' (15 warnings)

All other Errors and Warnings were caused by a small handful of sources.

 

[Centel Cesurus]

A quick google search revealed that this 'Distributed Com' Error 10016 is a somewhat common issues. Is it possible this could be causing an issue related to mouse input and microstutter?

 

[Centel Cesurus]

Apologies for not including the Event IDs, I can if needed. I've admittedly never used Event viewer before. Hell, didn't even know it existed!

 

[MMorris666]

Not a fix but you said you have v-sync on for g-sync. if you have a g-sync monitor you should turn v-sync off.

 

[Centel Cesurus]

Not a fix but you said you have v-sync on for g-sync. if you have a g-sync monitor you should turn v-sync off.

Thanks for the info.

 

[Centel Cesurus]

Quick update:

Just before going to ditch McAfee (as recommended by the OCUK forum) I went ahead and ran some more tests to see if changing the GPU to the top PCIE slot made any difference. It did. CS:GO and Valorant are now suffering from the stutter issue despite working relatively fun before moving the GPU. Moving the GPU back to the bottom PCIE slot did not reverse this. Any ideas what this could mean?

I'm still looking into the high number of warnings and errors I have observed on event viewer.
I'll also look into running a DPC latency test as also recommended by OCUK members.

Any information related to any of these would be greatly appreciated.

 

[Centel Cesurus]

Additional Update:

After deleting McAfee I can confirm that the AV is not the cause of the issue.

Running additional tests as recommended by OCUK forum member:

Checking what speed the PCIE slot is running at:
Using GPU-Z the 'Bus Interface' showed: "PCIe x16 3.0 @x2 3.0" at idle, and identical info: "PCIe x16 3.0 @x2 3.0" when at load (Running the GPU-Z render test). These values, as shown, did not change irrespective of GPU activity.

Additionally, the 3DMARK PCI Express feature test showed a Bandwidth of 1.53 GB/s.
This is half the Bandwidth speed of 3.06 GB/s mentioned in the original post, this speed was also achieved with the GPU in the second PCIE slot. I don't know why the Bandwidth has decreased.

Note: I have disabled HPET in device manager as recommended to me by a couple of people.

This is pretty big. From my understanding this could very well be related to the motherboard, GPU or, at very least, software interreference with the operation of these physical components. If anyone can offer any help or information with the newly provided information I have shown above it would be greatly appreciated. I recommend reading the post above as it is likely linked. Components are listed in original post.

Apologies If I don't respond to any posts immediately, I'm juggling this thread as well as one on OCUK all whilst independently researching to assist in solving this issue.

 

[gamerk316]

Hmmm, we're getting outside of my area of expertise here. All those warnings in the Event Viewer is certainly an indication something is wrong. One other thing that concerns me:

Checking what speed the PCIE slot is running at:
Using GPU-Z the 'Bus Interface' showed: "PCIe x16 3.0 @x2 3.0" at idle, and identical info: "PCIe x16 3.0 @x2 3.0" when at load (Running the GPU-Z render test). These values, as shown, did not change irrespective of GPU activity.

I'm at work so I can't check my GPU-Z for reference, but unless I'm horribly misunderstanding what is listed here, the physical PCI-E slot is running at x2 instead of x16, which would SIGNIFICANTLY hit bandwidth. Not sure why that would even happen unless the GPU wasn't getting sufficient power for some reason. Your PSU should be more then sufficient, so make sure those PCI-E power cables are fully inserted into the GPU; one of them might be loose (you'd be surprised how often this could happen).

 

[Centel Cesurus]

I'm at work so I can't check my GPU-Z for reference, but unless I'm horribly misunderstanding what is listed here, the physical PCI-E slot is running at x2 instead of x16, which would SIGNIFICANTLY hit bandwidth. Not sure why that would even happen unless the GPU wasn't getting sufficient power for some reason. Your PSU should be more then sufficient, so make sure those PCI-E power cables are fully inserted into the GPU; one of them might be loose (you'd be surprised how often this could happen).

Thanks for the reply!

You certainly aren't horribly misunderstanding anything! Checking GPU-Z my Bus interface is currently running at "PCI Express 3.0 x2" ! I've heard of people having issue with their Bus interface running at x8 instead of x16 but never before x2! Certainly explains my considerable drop in Bandwidth from 3.06GB/s to 1.53GB/s.

I suspect that the PCI-E power cables may very well be the issue as you noted. In the process of moving my GPU from the bottom to the top PCI-E slot and back again it is very possible that something could have worked loose. Could also explain that performance hit.
I just don't know why the previously unaffected titles are now experiencing this stutter issue as a result of moving the GPU around...

Sad thing is this seemingly separate problem may not even be linked to my main stutter related issue; just an unwanted consequence of my endeavours in finding a solution :/

 

[Centel Cesurus]

Just updating the thread:

Alright, digging around I've identified that:
When in the second PCIe slot the GPU is running at x2 Native with the Bus interface running PCI Express x2 3.0. As mentioned above this value does not change, irrespective of what the computer is actually doing. Load or Idle, it doesn't make much difference.
When in the top slot the GPU is running at x4 Native with Bus interface running PCI Express x4 3.0. This value also does not change.

Ensuring that PCIe Power cable is correctly connected to both the GPU directly and PSU hasn't done much
I've cleaned the GPU contacts with compressed air and Isopropanol Alcohol, and cleaned both PCIe slots again.
Used the Nvidia DisplayPort Firmware updater and updated the GPU accordingly
I've also tested quite a few related settings in the Bios that haven't made much difference.

When trying to install my GPU back into the top slot after cleaning it contacts I didn't receive any output despite said slot working just fine an hour ago. Moving the GPU into the second slot temporarily resolved the issue of no output. The fact that both the PCIe slot and GPU were clean really makes me think there is definitely something wrong with that top slot. That, or the graphics card needed to be re-seated.

At this moment in time I'm planning on flashing my motherboard's bios as it was updated relatively recently to the latest version, although I doubt this is the cause.

 

[Centel Cesurus]

I'm going to see if I can get access to a system similar to mine. I would be able to test my Graphics card and identify if it is still running at x2 and x4 native even in an entirely different system. If the GPU is fine this is either firmware or hardware related. Motherboard may even be buggered.
In the meantime any suggestions or thoughts would be appreciated.

 

[Centel Cesurus]

Hey all,

My apologies for my lengthy hiatus.

I have played around a fair bit with LatencyMon over the past few days and wanted to share some of my results:

• When idle, LatencyMon reported everything to be running as expected with a "current measured interrupt to process latency" fluctuating, but remaining below 80(µs) . After leaving the programme running for about 20 minutes highest interrupts gradually rise to the point where the programme recognised it as a potential issue.
• When booting up a game , however, LatencyMon immediately recognised there to be multiple issues, reporting the "Highest measured interrupt to process latency" to be 2051.20 (µs) , and "Highest reported DPC routine execution time" to be 1763.11(µs). The latter was linked to (ndis.sys - Network Driver Interface Specification (NDIS), Microsoft Corporation).
• When actually under load the "Current measured interrupt to process latency" remains under 300(µs) from the testing I've done so far.

The programme concluded that at least one problem appears to be network related, it suggested that a WLAN adapter could be the cause, however I do not use such a device.

I am unfamiliar with both this software and what these readings actually mean. What do these results indicate?

Thanks,