Does Mousetester/LatencyMon respond to emi?

Separate area for niche lag issues including unexpected causes and/or electromagnetic interference (ECC = retransmits = lag). Interference (EMI, EMF) of all kinds (wired, wireless, external, internal, environment, bad component) can cause error-correction latencies like a bad modem connection. Troubleshooting may require university degree. Your lag issue is likely not EMI. Please read this before entering sub-forum.
Forum rules
IMPORTANT:
This subforum is for advanced users only. This separate area is for niche or unexpected lag issues such as electromagnetic interference (EMI, EMF, electrical, radiofrequency, etc). Interference of all kinds (wired, wireless, external, internal, environment, bad component) can cause error-correction (ECC) latencies like a bad modem connection, except internally in a circuit. ECC = retransmits = lag. Troubleshooting may require university degree. Your lag issue is likely not EMI.
🠚 You Must Read This First Before Submit Post or Submit Reply
lizardpeter
Posts: 208
Joined: 01 Dec 2020, 14:41

Re: Does Mousetester/LatencyMon respond to emi?

Post by lizardpeter » 21 Sep 2021, 17:24

Eonds wrote: ↑
29 Jun 2021, 19:58
Playing in the AM is noticeably smoother/responsive though probably due to cleaner electricity.
I have thought I have felt this before, but honestly I am not 100% sure. I think this "feeling" could honestly have been from my mousepad (Zowie GSR-SE) being affected by humidity and causing it to feel muddy (it is a well-known issue), so I have just switched to the Artisan Zero Mid. Is there anything you could recommend to be able to alleviate this, if it is in fact a problem? I know there are UPS, line conditioners, and AC regenerators. I have heard very mixed reports about all of them. Also, are there any more user-friendly tools to test for this kind of issue? Maybe something that can be plugged into the socket?
i9 9900k | RTX 2080 Ti | 32 GB 4x8GB B-Die 3600 MT/s CL16 | XV252QF 390 Hz 1080p | AW2518H 240 Hz 1080p | PG279Q 144 Hz 1440p

Razer Viper 8K | Artisan Zero Mid XL | Apex Pro TKL | 1 gbps FiOS (Fiber)

Eonds
Posts: 262
Joined: 29 Oct 2020, 10:34

Re: Does Mousetester/LatencyMon respond to emi?

Post by Eonds » 21 Sep 2021, 18:50

lizardpeter wrote: ↑
21 Sep 2021, 17:24
Eonds wrote: ↑
29 Jun 2021, 19:58
Playing in the AM is noticeably smoother/responsive though probably due to cleaner electricity.
I have thought I have felt this before, but honestly I am not 100% sure. I think this "feeling" could honestly have been from my mousepad (Zowie GSR-SE) being affected by humidity and causing it to feel muddy (it is a well-known issue), so I have just switched to the Artisan Zero Mid. Is there anything you could recommend to be able to alleviate this, if it is in fact a problem? I know there are UPS, line conditioners, and AC regenerators. I have heard very mixed reports about all of them. Also, are there any more user-friendly tools to test for this kind of issue? Maybe something that can be plugged into the socket?
I mean most of the EMI stuff is more or less snake oil/Rare You should really focus on the more low hanging fruit. I think it's reasonable to try new things like a fiber optic display port cable, messing around with spread spectrum settings etc. There's more advanced things you can do but that requires the ability code in C & C++ & have a fundamental understanding of how windows/any OS works from top to bottom. Don't forget some games just act weird and have bugs that cause unintentional issues. Rocket league is one of those games where if i alt-tab, or restart the game at any point, my input becomes so delayed and awful. It makes no logical sense other than there's something seriously wrong with the game. Even if you're experiencing electrical issues, the best thing to do is to try EMI filters/spread spectrum settings. Or if you're a rich guy you can try using a large battery to power your PC & see if the issues are still present. I just find it laughable that people have fiber connections and good setups and have something to complain about. They run stock windows and don't even use basic things like msi mode, disabling c-states. The most basic fundamentals aren't even checked off the list. Seriously there's so many variables that could be the issue. It's best to start out with basics. Making sure there's no faulty parts, monitoring the OS, tracking down any DPC latency issues, disabling power saving features, accessing hidden bios settings, properly tuning overclocks, enabling msi/msi-x, setting PROPER priorities, disabling useless services, killing windows explorer, running exclusive full screen, use proper in-game settings for example, disabling anti-aliasing, multi threaded rendering, setting affinities so no singular core is being maxed out to the gills, buying proper low latency gear, turning off rgb, disabling buffers for M&K and disabling/lowering buffers for drivers which are not needed/can be lowered to improve latency, buying low latency hardware (AMD GPU's, Intel CPU's, DDR4-4400 (downclocking & tightening timings), tightening VRAM timings for your GPU, recording your screen with an ultra high FPS camera to see if there's micro stuttering (almost always will be present). Hardware desynchronization is usually the main issue these people have regardless of what's specifically is causing it. The lack of understanding they have about how computers function makes them believe that it's some mystical force randomly affecting them. If you've properly done all of these things, done PROPER benchmarks, making sure your cooling is 100% good (below 60C on all components at all times), Tuned your tile based rasterization etc. You can configure your system in any manner you want for example trade offs for latency, higher FPS, and or consistency. You have to understand computers fully to really be able to do these kinds of things and test it without it backfiring/making things much worse.

lizardpeter
Posts: 208
Joined: 01 Dec 2020, 14:41

Re: Does Mousetester/LatencyMon respond to emi?

Post by lizardpeter » 21 Sep 2021, 19:05

Eonds wrote: ↑
21 Sep 2021, 18:50
I mean most of the EMI stuff is more or less snake oil/Rare You should really focus on the more low hanging fruit. I think it's reasonable to try new things like a fiber optic display port cable, messing around with spread spectrum settings etc. There's more advanced things you can do but that requires the ability code in C & C++ & have a fundamental understanding of how windows/any OS works from top to bottom. Don't forget some games just act weird and have bugs that cause unintentional issues. Rocket league is one of those games where if i alt-tab, or restart the game at any point, my input becomes so delayed and awful. It makes no logical sense other than there's something seriously wrong with the game. Even if you're experiencing electrical issues, the best thing to do is to try EMI filters/spread spectrum settings. Or if you're a rich guy you can try using a large battery to power your PC & see if the issues are still present. I just find it laughable that people have fiber connections and good setups and have something to complain about. They run stock windows and don't even use basic things like msi mode, disabling c-states. The most basic fundamentals aren't even checked off the list. Seriously there's so many variables that could be the issue. It's best to start out with basics. Making sure there's no faulty parts, monitoring the OS, tracking down any DPC latency issues, disabling power saving features, accessing hidden bios settings, properly tuning overclocks, enabling msi/msi-x, setting PROPER priorities, disabling useless services, killing windows explorer, running exclusive full screen, use proper in-game settings for example, disabling anti-aliasing, multi threaded rendering, setting affinities so no singular core is being maxed out to the gills, buying proper low latency gear, turning off rgb, disabling buffers for M&K and disabling/lowering buffers for drivers which are not needed/can be lowered to improve latency, buying low latency hardware (AMD GPU's, Intel CPU's, DDR4-4400 (downclocking & tightening timings), tightening VRAM timings for your GPU, recording your screen with an ultra high FPS camera to see if there's micro stuttering (almost always will be present). Hardware desynchronization is usually the main issue these people have regardless of what's specifically is causing it. The lack of understanding they have about how computers function makes them believe that it's some mystical force randomly affecting them. If you've properly done all of these things, done PROPER benchmarks, making sure your cooling is 100% good (below 60C on all components at all times), Tuned your tile based rasterization etc. You can configure your system in any manner you want for example trade offs for latency, higher FPS, and or consistency. You have to understand computers fully to really be able to do these kinds of things and test it without it backfiring/making things much worse.
Thanks for the in-depth response. I already do most of these things (fiber connection, disabled C-states, MSI mode on, disabled power saving features in Windows and the BIOS, overclocked CPU and GPU and RAM, always low settings and no AA, boost frequency locked on the GPU, all excess services disabled, custom liquid cooling with very low temps, all RGB disabled on everything, etc.). I am interested in trying out more things, but I do not have access to great tools to test with. I have recently purchased the Viper 8K and the Acer 390 Hz monitor (will be here soon). I am going to try to start assigning core affinities for different drivers. A few things you mentioned were really interesting to me. How do I go about "disabling buffers for M&K and disabling/lowering buffers for drivers which are not needed/can be lowered to improve latency"? Also, what does killing Windows Explorer do? Is there any difference between MSI and MSI-X? Lastly, what kind of hidden BIOS settings are you talking about? I am on an ASUS Maximus XI Hero Z390 with an i9 9900k.
i9 9900k | RTX 2080 Ti | 32 GB 4x8GB B-Die 3600 MT/s CL16 | XV252QF 390 Hz 1080p | AW2518H 240 Hz 1080p | PG279Q 144 Hz 1440p

Razer Viper 8K | Artisan Zero Mid XL | Apex Pro TKL | 1 gbps FiOS (Fiber)

Eonds
Posts: 262
Joined: 29 Oct 2020, 10:34

Re: Does Mousetester/LatencyMon respond to emi?

Post by Eonds » 21 Sep 2021, 19:20

lizardpeter wrote: ↑
21 Sep 2021, 19:05
Eonds wrote: ↑
21 Sep 2021, 18:50
I mean most of the EMI stuff is more or less snake oil/Rare You should really focus on the more low hanging fruit. I think it's reasonable to try new things like a fiber optic display port cable, messing around with spread spectrum settings etc. There's more advanced things you can do but that requires the ability code in C & C++ & have a fundamental understanding of how windows/any OS works from top to bottom. Don't forget some games just act weird and have bugs that cause unintentional issues. Rocket league is one of those games where if i alt-tab, or restart the game at any point, my input becomes so delayed and awful. It makes no logical sense other than there's something seriously wrong with the game. Even if you're experiencing electrical issues, the best thing to do is to try EMI filters/spread spectrum settings. Or if you're a rich guy you can try using a large battery to power your PC & see if the issues are still present. I just find it laughable that people have fiber connections and good setups and have something to complain about. They run stock windows and don't even use basic things like msi mode, disabling c-states. The most basic fundamentals aren't even checked off the list. Seriously there's so many variables that could be the issue. It's best to start out with basics. Making sure there's no faulty parts, monitoring the OS, tracking down any DPC latency issues, disabling power saving features, accessing hidden bios settings, properly tuning overclocks, enabling msi/msi-x, setting PROPER priorities, disabling useless services, killing windows explorer, running exclusive full screen, use proper in-game settings for example, disabling anti-aliasing, multi threaded rendering, setting affinities so no singular core is being maxed out to the gills, buying proper low latency gear, turning off rgb, disabling buffers for M&K and disabling/lowering buffers for drivers which are not needed/can be lowered to improve latency, buying low latency hardware (AMD GPU's, Intel CPU's, DDR4-4400 (downclocking & tightening timings), tightening VRAM timings for your GPU, recording your screen with an ultra high FPS camera to see if there's micro stuttering (almost always will be present). Hardware desynchronization is usually the main issue these people have regardless of what's specifically is causing it. The lack of understanding they have about how computers function makes them believe that it's some mystical force randomly affecting them. If you've properly done all of these things, done PROPER benchmarks, making sure your cooling is 100% good (below 60C on all components at all times), Tuned your tile based rasterization etc. You can configure your system in any manner you want for example trade offs for latency, higher FPS, and or consistency. You have to understand computers fully to really be able to do these kinds of things and test it without it backfiring/making things much worse.
Thanks for the in-depth response. I already do most of these things (fiber connection, disabled C-states, MSI mode on, disabled power saving features in Windows and the BIOS, overclocked CPU and GPU and RAM, always low settings and no AA, boost frequency locked on the GPU, all excess services disabled, custom liquid cooling with very low temps, all RGB disabled on everything, etc.). I am interested in trying out more things, but I do not have access to great tools to test with. I have recently purchased the Viper 8K and the Acer 390 Hz monitor (will be here soon). I am going to try to start assigning core affinities for different drivers. A few things you mentioned were really interesting to me. How do I go about "disabling buffers for M&K and disabling/lowering buffers for drivers which are not needed/can be lowered to improve latency"? Also, what does killing Windows Explorer do? Is there any difference between MSI and MSI-X? Lastly, what kind of hidden BIOS settings are you talking about? I am on an ASUS Maximus XI Hero Z390 with an i9 9900k.
Most buffers can't be modified without reversing drivers or modifying them in some manner. What i mean by M&K buffers is the ones you can change through the registry (KeyboardDataQueueSize & MouseDataQueueSize 0 for both). Yes MSI is slower than MSI-X. Msi-X is much much faster and lower latency than MSI. Hidden settings like PCIE relaxed ordering, Other Spread spectrum settings, HPET, watchdog timers, payload sizes, etc. There's just a lot i can't remember atm. A huge overlooked part is having SET VOLTAGES for all components. The more deviation in voltage/frequency the worse it will be for consistency. You want latency to be less variable. Windows makes that really fucking difficult. Forcing your CPU into C0 (disabling idle features via power plans) also lowers latency because any power state switching causes latencies and that applies for any other component. Even the infinity fabric for ryzen has it's own P-states lol. They're there for a reason but they make it more than difficult to modify/change. Killing windows explorer helps a lot because as you can see it takes up a shit ton of system resources & really fucks up any polling related benchmarking. The less threads & processes the better.

lizardpeter
Posts: 208
Joined: 01 Dec 2020, 14:41

Re: Does Mousetester/LatencyMon respond to emi?

Post by lizardpeter » 21 Sep 2021, 20:15

Eonds wrote: ↑
21 Sep 2021, 19:20
Most buffers can't be modified without reversing drivers or modifying them in some manner. What i mean by M&K buffers is the ones you can change through the registry (KeyboardDataQueueSize & MouseDataQueueSize 0 for both). Yes MSI is slower than MSI-X. Msi-X is much much faster and lower latency than MSI.
Is 0 the best value? Does it really not need any buffer at all? I have read that 0 might be the default now anyway. According to some Microsoft documentation I found at https://docs.microsoft.com/en-us/previo ... 9(v=vs.85), they claim it is the default. However, that is from 2011, so I can totally see how it could have changed for Windows 10. I see others recommending values like 16 or 20, however. Have you found that 0 is the best?

Also, how does one go about enabling MSI-X? I know how to enable MSI and it works, but I cannot find anything about how to enable MSI-X.
i9 9900k | RTX 2080 Ti | 32 GB 4x8GB B-Die 3600 MT/s CL16 | XV252QF 390 Hz 1080p | AW2518H 240 Hz 1080p | PG279Q 144 Hz 1440p

Razer Viper 8K | Artisan Zero Mid XL | Apex Pro TKL | 1 gbps FiOS (Fiber)

MT_
Posts: 113
Joined: 17 Jan 2017, 15:39

Re: Does Mousetester/LatencyMon respond to emi?

Post by MT_ » 22 Sep 2021, 09:06

Latency monitors can show a whole lot of things, but they can never always be specific and even if the root cause appears to be what it shows, it could have an underlying issue that just excerbates the issue pointing you to the alleged culprit.

Imagine BIOS SMI's where the whole CPU basically stalls and goes into system management mode to handle certain things (Device ECC, fan control, etc)

The OS is not aware of such SMI's occuring but if this happens right as a driver IRQ/DPC is being executed it could show up as the driver being the root cause of dpc latency spike.

General rule of thumb in PC land optimization: Optimize BIOS, this is literally where it all begins. (sadly most manufacturers lock down bios settings severely and bios mod is required to unhide most aspects.

The less unneccesary functions enabled, the lower the interference in example. Now you just need the actual expertise to know what most advanced options do.
LTSC 21H2 Post-install Script
https://github.com/Marctraider/LiveScript-LTSC-21H2

System: MSI Z390 MEG Ace - 2080 Super (300W mod) - 9900K 5GHz Fixed Core (De-lid) - 32GB DDR3-3733-CL18 - Xonar Essence STX II

Eonds
Posts: 262
Joined: 29 Oct 2020, 10:34

Re: Does Mousetester/LatencyMon respond to emi?

Post by Eonds » 22 Sep 2021, 09:36

MT_ wrote: ↑
22 Sep 2021, 09:06
Latency monitors can show a whole lot of things, but they can never always be specific and even if the root cause appears to be what it shows, it could have an underlying issue that just excerbates the issue pointing you to the alleged culprit.

Imagine BIOS SMI's where the whole CPU basically stalls and goes into system management mode to handle certain things (Device ECC, fan control, etc)

The OS is not aware of such SMI's occuring but if this happens right as a driver IRQ/DPC is being executed it could show up as the driver being the root cause of dpc latency spike.

General rule of thumb in PC land optimization: Optimize BIOS, this is literally where it all begins. (sadly most manufacturers lock down bios settings severely and bios mod is required to unhide most aspects.

The less unneccesary functions enabled, the lower the interference in example. Now you just need the actual expertise to know what most advanced options do.

Thank you lol. I've been saying this but no one listens and they are die hard believers of a mystical force making them bad at CS:GO.

Eonds
Posts: 262
Joined: 29 Oct 2020, 10:34

Re: Does Mousetester/LatencyMon respond to emi?

Post by Eonds » 22 Sep 2021, 09:39

lizardpeter wrote: ↑
21 Sep 2021, 20:15
Eonds wrote: ↑
21 Sep 2021, 19:20
Most buffers can't be modified without reversing drivers or modifying them in some manner. What i mean by M&K buffers is the ones you can change through the registry (KeyboardDataQueueSize & MouseDataQueueSize 0 for both). Yes MSI is slower than MSI-X. Msi-X is much much faster and lower latency than MSI.
Is 0 the best value? Does it really not need any buffer at all? I have read that 0 might be the default now anyway. According to some Microsoft documentation I found at https://docs.microsoft.com/en-us/previo ... 9(v=vs.85), they claim it is the default. However, that is from 2011, so I can totally see how it could have changed for Windows 10. I see others recommending values like 16 or 20, however. Have you found that 0 is the best?

Also, how does one go about enabling MSI-X? I know how to enable MSI and it works, but I cannot find anything about how to enable MSI-X.
0 either disabled the buffer or sets it to default. Usually the value 16 (decimal) is the lowest safe limit for the buffers. I can't say for sure if 0 or 16 feels better as my PC has issues with interference.

Windows will automatically use MSI-X over MSI if you're driver supports it (and you enable msi mode via the utility or via regedit) (as far as i'm aware). linux > windows.

lizardpeter
Posts: 208
Joined: 01 Dec 2020, 14:41

Re: Does Mousetester/LatencyMon respond to emi?

Post by lizardpeter » 22 Sep 2021, 16:59

Eonds wrote: ↑
22 Sep 2021, 09:39
0 either disabled the buffer or sets it to default. Usually the value 16 (decimal) is the lowest safe limit for the buffers. I can't say for sure if 0 or 16 feels better as my PC has issues with interference.
I'm going to test this more later today and try to find the exact buffer sizes with various settings (namely default, 0, and 16). I'll report back.
Eonds wrote: ↑
22 Sep 2021, 09:39
Windows will automatically use MSI-X over MSI if your driver supports it (and you enable msi mode via the utility or via regedit) (as far as i'm aware). linux > windows.
I have MSI mode enabled in my registry manually. However, I cannot seem to get this MSI-X mode working. I also downloaded the utility and looked around online and found that some drivers for other devices support MSI-X using the utility. As for the NVIDIA GPUs, I am yet to find anyone talking about getting MSI-X mode to work with them. Do you have it working under Windows with an AMD GPU?
i9 9900k | RTX 2080 Ti | 32 GB 4x8GB B-Die 3600 MT/s CL16 | XV252QF 390 Hz 1080p | AW2518H 240 Hz 1080p | PG279Q 144 Hz 1440p

Razer Viper 8K | Artisan Zero Mid XL | Apex Pro TKL | 1 gbps FiOS (Fiber)

lizardpeter
Posts: 208
Joined: 01 Dec 2020, 14:41

Re: Does Mousetester/LatencyMon respond to emi?

Post by lizardpeter » 23 Sep 2021, 00:08

So I ran some tests with three different values for MouseDataQueueSize. It does in fact seem like 0 sets it at the default value. The non-paged memory pool for both no registry value and 0 is 9872. With 16, however, it is only 1744.
i9 9900k | RTX 2080 Ti | 32 GB 4x8GB B-Die 3600 MT/s CL16 | XV252QF 390 Hz 1080p | AW2518H 240 Hz 1080p | PG279Q 144 Hz 1440p

Razer Viper 8K | Artisan Zero Mid XL | Apex Pro TKL | 1 gbps FiOS (Fiber)

Post Reply