Heavy character movement (not input lag)

Everything about latency. Tips, testing methods, mouse lag, display lag, game engine lag, network lag, whole input lag chain, VSYNC OFF vs VSYNC ON, and more! Input Lag Articles on Blur Busters.
Sirito
Posts: 55
Joined: 17 Jun 2022, 06:48

Re: Heavy character movement (not input lag)

Post by Sirito » 21 Jun 2022, 11:54

kokkatc wrote:
21 Jun 2022, 00:30
Sirito wrote:
20 Jun 2022, 17:52
kokkatc wrote:
20 Jun 2022, 17:40
Sirito wrote:
17 Jun 2022, 08:25
I have a strange problem and I believe it is not input lag, simply my character in Valorant moves heavily like the gravity is pulling me harder than it should, jumping also makes me float a little longer.
Why I believe it is not input lag is because all my inputs are registered correctly at the correct time, and I can flick enemies with my mouse too. Currently I'm immortal 1 which is considered high elo in the game but the game isn't enjoyable this way, I changed my playstyle to match this heavy movement so I stay back site and use my gamesense skill, because if I tried to move I will be heavy and the enemy will react on me before I even see him.
Does anyone face something like this? what is this problem called? what may be the reasons causing it? Thank you.

PC Specs:
Motherboard: Asrock B560M Pro4 (Bios Version 2.20)
Processor: Intel Core i5-10400F (2.9 GHz / 4.3 GHz) (Max Temp 55)
GPU: GTX 1060 6 GB Zotac OC (Max Temp 75C)
Monitor: AOC 24G2 23.8 IPS 144hz (Overdirve set to Strong) (Freesync set to off)
RAM: x2 HyperX Fury 8 GB DDR4 3200 MHz CL16 (Slot 2 and 4)
SSD 1: KINGSTONE SA2000 500 GB (M.2) (Windows Setup here)
SSD 2: PLEXTOR PX-128S3C 120 GB (SATA)
HDD: Samsung 7200 RPM 1TB
Power Supply: Cougar VTE 500W
Mouse: Logitech G305
Keyboard: HyperX Alloy Origins Core
Try these things:

Bios: Disable EIST, Speedstep, thermal monitor, C-states. Any and all power saving features in your BIOS.

XMP: In my experience, heavy character movement is often connected to memory. Whether your memory OC isn't stable or not compatible w/ your board or possibly memory training is failing giving you undesirable timings/setings, etc. Do a test run w/ XMP disabled and see if you still experience the same feeling. If you don't, may want to find a compatible kit for your board or look at into fixing the timings. I'd also experiment by disabling turbo and any OC's you may have and testing. An unstable OC can also cause this.

Windows: Enable Ultimate Performance Power Plan: Disable usb select suspend and pci power savings in the plan advanced options
Use timer resolution (Google for download link), set to .5ms and leave it running while you game. ISLC can also be used.

Chipset Drivers: Make sure your chipset drivers are the latest and installed. You can get them here for your board: https://www.asrock.com/mb/Intel/B560M%2 ... p#Download
Set a restore point before you install your drivers just in case the chipset drivers happen to make your issue worse, then you can revert back.

Game: Of course, remember to disable fullscreen optimizations by finding your game's .exe, right click for properties, go to compatibility, click on disable fullscreen optimizations, save, close.
Ty for detailed reply,

Bios:
I already disabled these settings.

XMP:
My ram was running at XMP 3200MHZ with auto timing (I think it was 15 17 17 36 69) .
But I read that maybe this is affected by ram timing so I made my ram run at 2133MHZ with auto timing(it is now 15 15 15 35 373) and didn't notice any difference and till now it is running on 2133MHZ no XMP

Windows:
Ultimate performance already enabled.
Usb power saving disabled.
Using intelligent standby list and timer is set to 0.5 and also latency mon values are very good during gaming.

Game:
Fullscreen optimization already disabled.

I tried a lot really a lot, I think the only thing left is the manual ram timing tuning but I tried once and my pc didn't boot up until I reset CMOS.
In your bios, make sure your memory is training properly. I'd disable fast boot in bios so your memory trains at every boot. This is quick and easy to do and test. You can download a program called timing configurator by ASRock for z690 from here https://www.overclock.net/threads/offic ... k.1794799/. This program will show all your timings in the windows gui. You want to take note of your rtls/iols and make sure they're training properly. The lower numbers are usually identical. The higher numbers shouldn't be more than two numbers apart (69 /71 for example). If they are more than 2 numbers apart then your memory won't be stable. You can also test running your ram in dimm slots 1 & 3 to see if there's any difference.

Some other things to try. Download DDU (display driver uninstaller). Boot into safe mode and run this and remove/clean your gpu drivers. Reboot into normal mode and reinstall gpu drivers.

If you're up for the exercise, consider reinstalling your os. Make sure your Ethernet is unplugged so windows update doesn't install driver's automatically. I've had several instances where windows update installs garbage drivers creating a lot of problems. After your new install and with your Ethernet unplugged, disable windows update from installing any drivers at all. You need to set searchorderconfig to a value of 0 in your registry, then reboot. This will prevent windows update from auto installing drivers for you. Here's a link for the registry key https://forums.guru3d.com/threads/windo ... sy.418033/. After you set this, reboot and plug your Ethernet cable back in. Update windows first, then install all of your chipset drivers and reboot after each one. Then configure all other essential settings. Use the audio and nic drivers that windows uses by default (these are already signed and tested, no need to update further unless there are problems). This is how I go about every os install to ensure I know what drivers are being used and to prevent drivers from being installed without me knowing. Also download gpuz and make sure your gpu is running in x16 4.0 or 3.0 mode, not 1 or 1.1. I once didn't seat my GPU all the way down into the slot where it was only registering 1.1.

I know this shit is frustrating as all hell. Start with a good baseline and go from there.
Sorry for the late reply,
I went to bios and found that fast boot is already disabled

I downloaded the program and here's the first screenshot
Image
and the second screensoht when I rebooted once
Image
FSB:DRAM and DRAM Frequency changed without me doing anything just a reboot.
Also if can you please check my rtls/iols cause I didn't fully understand what you said.

I used DDU and NV Clean Install before, and also stopped windows from downloading drivers, also I switched my gpu to msi mode before and also clean installed windows, and tried some gaming windows which had many services removed, I actually tested these things a lot to eliminate placebo and I can say none of these did anything.

Downloaded GPU-Z and found that my gpu is running in PCIe x16 3.0 @x16 1.1, here's a screenshot about it too.
Image
Thanks for you time :).

kokkatc
Posts: 108
Joined: 23 Mar 2017, 13:49

Re: Heavy character movement (not input lag)

Post by kokkatc » 21 Jun 2022, 12:28

Sirito wrote:
21 Jun 2022, 11:54
kokkatc wrote:
21 Jun 2022, 00:30
Sirito wrote:
20 Jun 2022, 17:52
kokkatc wrote:
20 Jun 2022, 17:40


Try these things:

Bios: Disable EIST, Speedstep, thermal monitor, C-states. Any and all power saving features in your BIOS.

XMP: In my experience, heavy character movement is often connected to memory. Whether your memory OC isn't stable or not compatible w/ your board or possibly memory training is failing giving you undesirable timings/setings, etc. Do a test run w/ XMP disabled and see if you still experience the same feeling. If you don't, may want to find a compatible kit for your board or look at into fixing the timings. I'd also experiment by disabling turbo and any OC's you may have and testing. An unstable OC can also cause this.

Windows: Enable Ultimate Performance Power Plan: Disable usb select suspend and pci power savings in the plan advanced options
Use timer resolution (Google for download link), set to .5ms and leave it running while you game. ISLC can also be used.

Chipset Drivers: Make sure your chipset drivers are the latest and installed. You can get them here for your board: https://www.asrock.com/mb/Intel/B560M%2 ... p#Download
Set a restore point before you install your drivers just in case the chipset drivers happen to make your issue worse, then you can revert back.

Game: Of course, remember to disable fullscreen optimizations by finding your game's .exe, right click for properties, go to compatibility, click on disable fullscreen optimizations, save, close.
Ty for detailed reply,

Bios:
I already disabled these settings.

XMP:
My ram was running at XMP 3200MHZ with auto timing (I think it was 15 17 17 36 69) .
But I read that maybe this is affected by ram timing so I made my ram run at 2133MHZ with auto timing(it is now 15 15 15 35 373) and didn't notice any difference and till now it is running on 2133MHZ no XMP

Windows:
Ultimate performance already enabled.
Usb power saving disabled.
Using intelligent standby list and timer is set to 0.5 and also latency mon values are very good during gaming.

Game:
Fullscreen optimization already disabled.

I tried a lot really a lot, I think the only thing left is the manual ram timing tuning but I tried once and my pc didn't boot up until I reset CMOS.
In your bios, make sure your memory is training properly. I'd disable fast boot in bios so your memory trains at every boot. This is quick and easy to do and test. You can download a program called timing configurator by ASRock for z690 from here https://www.overclock.net/threads/offic ... k.1794799/. This program will show all your timings in the windows gui. You want to take note of your rtls/iols and make sure they're training properly. The lower numbers are usually identical. The higher numbers shouldn't be more than two numbers apart (69 /71 for example). If they are more than 2 numbers apart then your memory won't be stable. You can also test running your ram in dimm slots 1 & 3 to see if there's any difference.

Some other things to try. Download DDU (display driver uninstaller). Boot into safe mode and run this and remove/clean your gpu drivers. Reboot into normal mode and reinstall gpu drivers.

If you're up for the exercise, consider reinstalling your os. Make sure your Ethernet is unplugged so windows update doesn't install driver's automatically. I've had several instances where windows update installs garbage drivers creating a lot of problems. After your new install and with your Ethernet unplugged, disable windows update from installing any drivers at all. You need to set searchorderconfig to a value of 0 in your registry, then reboot. This will prevent windows update from auto installing drivers for you. Here's a link for the registry key https://forums.guru3d.com/threads/windo ... sy.418033/. After you set this, reboot and plug your Ethernet cable back in. Update windows first, then install all of your chipset drivers and reboot after each one. Then configure all other essential settings. Use the audio and nic drivers that windows uses by default (these are already signed and tested, no need to update further unless there are problems). This is how I go about every os install to ensure I know what drivers are being used and to prevent drivers from being installed without me knowing. Also download gpuz and make sure your gpu is running in x16 4.0 or 3.0 mode, not 1 or 1.1. I once didn't seat my GPU all the way down into the slot where it was only registering 1.1.

I know this shit is frustrating as all hell. Start with a good baseline and go from there.
Sorry for the late reply,
I went to bios and found that fast boot is already disabled

I downloaded the program and here's the first screenshot
Image
and the second screensoht when I rebooted once
Image
FSB:DRAM and DRAM Frequency changed without me doing anything just a reboot.
Also if can you please check my rtls/iols cause I didn't fully understand what you said.

I used DDU and NV Clean Install before, and also stopped windows from downloading drivers, also I switched my gpu to msi mode before and also clean installed windows, and tried some gaming windows which had many services removed, I actually tested these things a lot to eliminate placebo and I can say none of these did anything.

Downloaded GPU-Z and found that my gpu is running in PCIe x16 3.0 @x16 1.1, here's a screenshot about it too.
Image
Thanks for you time :).
Okay, first I'd focus on fixing your GPU. See attachments for a visual reference. Your GPU is operating at <x16 1.1> when it should be operating at <x16 3.0>. This is bad. Your GPU bandwidth is being severely limited which could easily explain the 'heavy movement' you're experiencing. The first attachment shows what your GPU is operating at and the 2nd attachment shows what IT SHOULD be running at.

First thing I'd do is to find out whether your GPU is in power savings mode or if your card isn't operating at the correct speed. Keep GPU-Z open, and run the render test in GPU-Z. Click the question mark next to the bus interface speed to run render test. See if the bus interface changes to <x16 3.0> during the benchmark/test. It could just be throttling down when idle but I doubt it. If it still shows your GPU is operating @ x16 1.1 during the benchmark, then you know you have a problem.

Three things can cause this. 1) Your GPU is installed in the incorrect PCIE slot. Make sure it's installed in the top slot. You can also refer to your motherboard's manual. 2) Your GPU isn't seated properly. Apply good force to make sure it fully seats and you should hear a click. Keep your case open and don't screw down the card all the way, just enough so the gpu is being held by the screw on the bracket. Sometimes the screw when fully tightened lifts the GPU out of it's socket. Boot back up and check GPU-Z and see if it changed to @ x16 3.0. 3) Your default PCIE slot settings are set to gen 1 or defaulting to gen 1. If it's still showing x16 1.1, go into your BIOS and find your PCIE port settings and change to the highest available gen. For your board in the BIOS, go to 'Advanced \ Chipset Configuration, change pcie1/pcie2/pcie3 link speed to highest gen for all pcie ports. Save, reboot, check GPU-Z.

Since we're on this subject, make sure your PSU GPU power cables are properly connected to the card as well. I would just disconnect and reconnect to be sure. What PSU do you have, how many watts?

Fix this first before looking further into memory.

Post back a screenshot of GPU-Z after you fix this.
Attachments
GPU Z Bus Interface.png
GPU Z Bus Interface.png (267.03 KiB) Viewed 2888 times
GPUZ Bus Interface GOOD.png
GPUZ Bus Interface GOOD.png (38.94 KiB) Viewed 2888 times
Last edited by kokkatc on 21 Jun 2022, 13:09, edited 4 times in total.

Sirito
Posts: 55
Joined: 17 Jun 2022, 06:48

Re: Heavy character movement (not input lag)

Post by Sirito » 21 Jun 2022, 12:37

kokkatc wrote:
21 Jun 2022, 12:28
Sirito wrote:
21 Jun 2022, 11:54
kokkatc wrote:
21 Jun 2022, 00:30
Sirito wrote:
20 Jun 2022, 17:52


Ty for detailed reply,

Bios:
I already disabled these settings.

XMP:
My ram was running at XMP 3200MHZ with auto timing (I think it was 15 17 17 36 69) .
But I read that maybe this is affected by ram timing so I made my ram run at 2133MHZ with auto timing(it is now 15 15 15 35 373) and didn't notice any difference and till now it is running on 2133MHZ no XMP

Windows:
Ultimate performance already enabled.
Usb power saving disabled.
Using intelligent standby list and timer is set to 0.5 and also latency mon values are very good during gaming.

Game:
Fullscreen optimization already disabled.

I tried a lot really a lot, I think the only thing left is the manual ram timing tuning but I tried once and my pc didn't boot up until I reset CMOS.
In your bios, make sure your memory is training properly. I'd disable fast boot in bios so your memory trains at every boot. This is quick and easy to do and test. You can download a program called timing configurator by ASRock for z690 from here https://www.overclock.net/threads/offic ... k.1794799/. This program will show all your timings in the windows gui. You want to take note of your rtls/iols and make sure they're training properly. The lower numbers are usually identical. The higher numbers shouldn't be more than two numbers apart (69 /71 for example). If they are more than 2 numbers apart then your memory won't be stable. You can also test running your ram in dimm slots 1 & 3 to see if there's any difference.

Some other things to try. Download DDU (display driver uninstaller). Boot into safe mode and run this and remove/clean your gpu drivers. Reboot into normal mode and reinstall gpu drivers.

If you're up for the exercise, consider reinstalling your os. Make sure your Ethernet is unplugged so windows update doesn't install driver's automatically. I've had several instances where windows update installs garbage drivers creating a lot of problems. After your new install and with your Ethernet unplugged, disable windows update from installing any drivers at all. You need to set searchorderconfig to a value of 0 in your registry, then reboot. This will prevent windows update from auto installing drivers for you. Here's a link for the registry key https://forums.guru3d.com/threads/windo ... sy.418033/. After you set this, reboot and plug your Ethernet cable back in. Update windows first, then install all of your chipset drivers and reboot after each one. Then configure all other essential settings. Use the audio and nic drivers that windows uses by default (these are already signed and tested, no need to update further unless there are problems). This is how I go about every os install to ensure I know what drivers are being used and to prevent drivers from being installed without me knowing. Also download gpuz and make sure your gpu is running in x16 4.0 or 3.0 mode, not 1 or 1.1. I once didn't seat my GPU all the way down into the slot where it was only registering 1.1.

I know this shit is frustrating as all hell. Start with a good baseline and go from there.
Sorry for the late reply,
I went to bios and found that fast boot is already disabled

I downloaded the program and here's the first screenshot
Image
and the second screensoht when I rebooted once
Image
FSB:DRAM and DRAM Frequency changed without me doing anything just a reboot.
Also if can you please check my rtls/iols cause I didn't fully understand what you said.

I used DDU and NV Clean Install before, and also stopped windows from downloading drivers, also I switched my gpu to msi mode before and also clean installed windows, and tried some gaming windows which had many services removed, I actually tested these things a lot to eliminate placebo and I can say none of these did anything.

Downloaded GPU-Z and found that my gpu is running in PCIe x16 3.0 @x16 1.1, here's a screenshot about it too.
Image
Thanks for you time :).
Okay, first I'd focus on fixing your GPU. See attachments for a visual reference. Your GPU is operating at <x16 1.1> when it should be operating at <x16 3.0>. This is bad. Your GPU bandwidth is being severely limited which could easily explain the 'heavy movement' you're experiencing. The first attachment shows what your GPU is operating at and the 2nd attachment shows what IT SHOULD be running at.

First thing I'd do is to find out whether your GPU is in power savings mode or if your card isn't operating at the correct speed. Keep GPU-Z open, and run the render test in GPU-Z. See if the bus interface changes to <x16 3.0> during the benchmark. It could just be throttling down when idle but I doubt it. If it still shows your GPU is operating @ x16 1.1 during the benchmark, then you know you have a problem.

A couple of things can cause this. Either your GPU isn't seated properly or your default BIOS settings for your PCIE ports are defaulting to gen 1. First, re-seat your GPU. Apply good force to make sure it fully seats and you should hear a click. Keep your case open and don't screw down the card all the way, just enough so the gpu is being held by the screw on the bracket. Sometimes the screw when fully tightened lifts the GPU out of it's socket. Boot back up and check GPU-Z and see if it changed to @ x16 3.0. If it's still showing x16 1.1, go into your BIOS and find your PCIE port settings and change to the highest available gen. So in your case, it's likely going to max out at gen 3 or gen 4. And of course make sure your GPU is in the top slot.

Fix this first before looking further into memory.
You'r right, I'm playing Valorant right now and found that it is now @x16 3.0
Image

kokkatc
Posts: 108
Joined: 23 Mar 2017, 13:49

Re: Heavy character movement (not input lag)

Post by kokkatc » 21 Jun 2022, 13:24

kokkatc wrote:
21 Jun 2022, 12:28
Sirito wrote:
21 Jun 2022, 11:54
kokkatc wrote:
21 Jun 2022, 00:30
Sirito wrote:
20 Jun 2022, 17:52


Ty for detailed reply,

Bios:
I already disabled these settings.

XMP:
My ram was running at XMP 3200MHZ with auto timing (I think it was 15 17 17 36 69) .
But I read that maybe this is affected by ram timing so I made my ram run at 2133MHZ with auto timing(it is now 15 15 15 35 373) and didn't notice any difference and till now it is running on 2133MHZ no XMP

Windows:
Ultimate performance already enabled.
Usb power saving disabled.
Using intelligent standby list and timer is set to 0.5 and also latency mon values are very good during gaming.

Game:
Fullscreen optimization already disabled.

I tried a lot really a lot, I think the only thing left is the manual ram timing tuning but I tried once and my pc didn't boot up until I reset CMOS.
In your bios, make sure your memory is training properly. I'd disable fast boot in bios so your memory trains at every boot. This is quick and easy to do and test. You can download a program called timing configurator by ASRock for z690 from here https://www.overclock.net/threads/offic ... k.1794799/. This program will show all your timings in the windows gui. You want to take note of your rtls/iols and make sure they're training properly. The lower numbers are usually identical. The higher numbers shouldn't be more than two numbers apart (69 /71 for example). If they are more than 2 numbers apart then your memory won't be stable. You can also test running your ram in dimm slots 1 & 3 to see if there's any difference.

Some other things to try. Download DDU (display driver uninstaller). Boot into safe mode and run this and remove/clean your gpu drivers. Reboot into normal mode and reinstall gpu drivers.

If you're up for the exercise, consider reinstalling your os. Make sure your Ethernet is unplugged so windows update doesn't install driver's automatically. I've had several instances where windows update installs garbage drivers creating a lot of problems. After your new install and with your Ethernet unplugged, disable windows update from installing any drivers at all. You need to set searchorderconfig to a value of 0 in your registry, then reboot. This will prevent windows update from auto installing drivers for you. Here's a link for the registry key https://forums.guru3d.com/threads/windo ... sy.418033/. After you set this, reboot and plug your Ethernet cable back in. Update windows first, then install all of your chipset drivers and reboot after each one. Then configure all other essential settings. Use the audio and nic drivers that windows uses by default (these are already signed and tested, no need to update further unless there are problems). This is how I go about every os install to ensure I know what drivers are being used and to prevent drivers from being installed without me knowing. Also download gpuz and make sure your gpu is running in x16 4.0 or 3.0 mode, not 1 or 1.1. I once didn't seat my GPU all the way down into the slot where it was only registering 1.1.

I know this shit is frustrating as all hell. Start with a good baseline and go from there.
Sorry for the late reply,
I went to bios and found that fast boot is already disabled

I downloaded the program and here's the first screenshot
Image
and the second screensoht when I rebooted once
Image
FSB:DRAM and DRAM Frequency changed without me doing anything just a reboot.
Also if can you please check my rtls/iols cause I didn't fully understand what you said.

I used DDU and NV Clean Install before, and also stopped windows from downloading drivers, also I switched my gpu to msi mode before and also clean installed windows, and tried some gaming windows which had many services removed, I actually tested these things a lot to eliminate placebo and I can say none of these did anything.

Downloaded GPU-Z and found that my gpu is running in PCIe x16 3.0 @x16 1.1, here's a screenshot about it too.
Image
Thanks for you time :).
Okay, first I'd focus on fixing your GPU. See attachments for a visual reference. Your GPU is operating at <x16 1.1> when it should be operating at <x16 3.0>. This is bad. Your GPU bandwidth is being severely limited which could easily explain the 'heavy movement' you're experiencing. The first attachment shows what your GPU is operating at and the 2nd attachment shows what IT SHOULD be running at.

First thing I'd do is to find out whether your GPU is in power savings mode or if your card isn't operating at the correct speed. Keep GPU-Z open, and run the render test in GPU-Z. Click the question mark next to the bus interface speed to run render test. See if the bus interface changes to <x16 3.0> during the benchmark/test. It could just be throttling down when idle but I doubt it. If it still shows your GPU is operating @ x16 1.1 during the benchmark, then you know you have a problem.

Three things can cause this. 1) Your GPU is installed in the incorrect PCIE slot. Make sure it's installed in the top slot. You can also refer to your motherboard's manual. 2) Your GPU isn't seated properly. Apply good force to make sure it fully seats and you should hear a click. Keep your case open and don't screw down the card all the way, just enough so the gpu is being held by the screw on the bracket. Sometimes the screw when fully tightened lifts the GPU out of it's socket. Boot back up and check GPU-Z and see if it changed to @ x16 3.0. 3) Your default PCIE slot settings are set to gen 1 or defaulting to gen 1. If it's still showing x16 1.1, go into your BIOS and find your PCIE port settings and change to the highest available gen. For your board in the BIOS, go to 'Advanced \ Chipset Configuration, change pcie1/pcie2/pcie3 link speed to highest gen for all pcie ports. Save, reboot, check GPU-Z.

Since we're on this subject, make sure your PSU GPU power cables are properly connected to the card as well. I would just disconnect and reconnect to be sure. What PSU do you have, how many watts?

Fix this first before looking further into memory.

Post back a screenshot of GPU-Z after you fix this.
Can you try a different version of timing configurator? Your DRAM frequency makes no sense and I wonder if this version doesn't work w/ your board. Uninstall timing configurator and try this one instead https://www.touslesdrivers.com/index.ph ... _langue=en Screenshot back like you did before w/ this version of timing configurator.

Sirito
Posts: 55
Joined: 17 Jun 2022, 06:48

Re: Heavy character movement (not input lag)

Post by Sirito » 21 Jun 2022, 15:21

kokkatc wrote:
21 Jun 2022, 13:24
kokkatc wrote:
21 Jun 2022, 12:28
Sirito wrote:
21 Jun 2022, 11:54
kokkatc wrote:
21 Jun 2022, 00:30


In your bios, make sure your memory is training properly. I'd disable fast boot in bios so your memory trains at every boot. This is quick and easy to do and test. You can download a program called timing configurator by ASRock for z690 from here https://www.overclock.net/threads/offic ... k.1794799/. This program will show all your timings in the windows gui. You want to take note of your rtls/iols and make sure they're training properly. The lower numbers are usually identical. The higher numbers shouldn't be more than two numbers apart (69 /71 for example). If they are more than 2 numbers apart then your memory won't be stable. You can also test running your ram in dimm slots 1 & 3 to see if there's any difference.

Some other things to try. Download DDU (display driver uninstaller). Boot into safe mode and run this and remove/clean your gpu drivers. Reboot into normal mode and reinstall gpu drivers.

If you're up for the exercise, consider reinstalling your os. Make sure your Ethernet is unplugged so windows update doesn't install driver's automatically. I've had several instances where windows update installs garbage drivers creating a lot of problems. After your new install and with your Ethernet unplugged, disable windows update from installing any drivers at all. You need to set searchorderconfig to a value of 0 in your registry, then reboot. This will prevent windows update from auto installing drivers for you. Here's a link for the registry key https://forums.guru3d.com/threads/windo ... sy.418033/. After you set this, reboot and plug your Ethernet cable back in. Update windows first, then install all of your chipset drivers and reboot after each one. Then configure all other essential settings. Use the audio and nic drivers that windows uses by default (these are already signed and tested, no need to update further unless there are problems). This is how I go about every os install to ensure I know what drivers are being used and to prevent drivers from being installed without me knowing. Also download gpuz and make sure your gpu is running in x16 4.0 or 3.0 mode, not 1 or 1.1. I once didn't seat my GPU all the way down into the slot where it was only registering 1.1.

I know this shit is frustrating as all hell. Start with a good baseline and go from there.
Sorry for the late reply,
I went to bios and found that fast boot is already disabled

I downloaded the program and here's the first screenshot
Image
and the second screensoht when I rebooted once
Image
FSB:DRAM and DRAM Frequency changed without me doing anything just a reboot.
Also if can you please check my rtls/iols cause I didn't fully understand what you said.

I used DDU and NV Clean Install before, and also stopped windows from downloading drivers, also I switched my gpu to msi mode before and also clean installed windows, and tried some gaming windows which had many services removed, I actually tested these things a lot to eliminate placebo and I can say none of these did anything.

Downloaded GPU-Z and found that my gpu is running in PCIe x16 3.0 @x16 1.1, here's a screenshot about it too.
Image
Thanks for you time :).
Okay, first I'd focus on fixing your GPU. See attachments for a visual reference. Your GPU is operating at <x16 1.1> when it should be operating at <x16 3.0>. This is bad. Your GPU bandwidth is being severely limited which could easily explain the 'heavy movement' you're experiencing. The first attachment shows what your GPU is operating at and the 2nd attachment shows what IT SHOULD be running at.

First thing I'd do is to find out whether your GPU is in power savings mode or if your card isn't operating at the correct speed. Keep GPU-Z open, and run the render test in GPU-Z. Click the question mark next to the bus interface speed to run render test. See if the bus interface changes to <x16 3.0> during the benchmark/test. It could just be throttling down when idle but I doubt it. If it still shows your GPU is operating @ x16 1.1 during the benchmark, then you know you have a problem.

Three things can cause this. 1) Your GPU is installed in the incorrect PCIE slot. Make sure it's installed in the top slot. You can also refer to your motherboard's manual. 2) Your GPU isn't seated properly. Apply good force to make sure it fully seats and you should hear a click. Keep your case open and don't screw down the card all the way, just enough so the gpu is being held by the screw on the bracket. Sometimes the screw when fully tightened lifts the GPU out of it's socket. Boot back up and check GPU-Z and see if it changed to @ x16 3.0. 3) Your default PCIE slot settings are set to gen 1 or defaulting to gen 1. If it's still showing x16 1.1, go into your BIOS and find your PCIE port settings and change to the highest available gen. For your board in the BIOS, go to 'Advanced \ Chipset Configuration, change pcie1/pcie2/pcie3 link speed to highest gen for all pcie ports. Save, reboot, check GPU-Z.

Since we're on this subject, make sure your PSU GPU power cables are properly connected to the card as well. I would just disconnect and reconnect to be sure. What PSU do you have, how many watts?

Fix this first before looking further into memory.

Post back a screenshot of GPU-Z after you fix this.
Can you try a different version of timing configurator? Your DRAM frequency makes no sense and I wonder if this version doesn't work w/ your board. Uninstall timing configurator and try this one instead https://www.touslesdrivers.com/index.ph ... _langue=en Screenshot back like you did before w/ this version of timing configurator.
It is the same thing unfortunately
Image

But it is working in CPUZ
Image

kokkatc
Posts: 108
Joined: 23 Mar 2017, 13:49

Re: Heavy character movement (not input lag)

Post by kokkatc » 21 Jun 2022, 16:53

Sirito wrote:
21 Jun 2022, 15:21
kokkatc wrote:
21 Jun 2022, 13:24
kokkatc wrote:
21 Jun 2022, 12:28
Sirito wrote:
21 Jun 2022, 11:54


Sorry for the late reply,
I went to bios and found that fast boot is already disabled

I downloaded the program and here's the first screenshot
Image
and the second screensoht when I rebooted once
Image
FSB:DRAM and DRAM Frequency changed without me doing anything just a reboot.
Also if can you please check my rtls/iols cause I didn't fully understand what you said.

I used DDU and NV Clean Install before, and also stopped windows from downloading drivers, also I switched my gpu to msi mode before and also clean installed windows, and tried some gaming windows which had many services removed, I actually tested these things a lot to eliminate placebo and I can say none of these did anything.

Downloaded GPU-Z and found that my gpu is running in PCIe x16 3.0 @x16 1.1, here's a screenshot about it too.
Image
Thanks for you time :).
Okay, first I'd focus on fixing your GPU. See attachments for a visual reference. Your GPU is operating at <x16 1.1> when it should be operating at <x16 3.0>. This is bad. Your GPU bandwidth is being severely limited which could easily explain the 'heavy movement' you're experiencing. The first attachment shows what your GPU is operating at and the 2nd attachment shows what IT SHOULD be running at.

First thing I'd do is to find out whether your GPU is in power savings mode or if your card isn't operating at the correct speed. Keep GPU-Z open, and run the render test in GPU-Z. Click the question mark next to the bus interface speed to run render test. See if the bus interface changes to <x16 3.0> during the benchmark/test. It could just be throttling down when idle but I doubt it. If it still shows your GPU is operating @ x16 1.1 during the benchmark, then you know you have a problem.

Three things can cause this. 1) Your GPU is installed in the incorrect PCIE slot. Make sure it's installed in the top slot. You can also refer to your motherboard's manual. 2) Your GPU isn't seated properly. Apply good force to make sure it fully seats and you should hear a click. Keep your case open and don't screw down the card all the way, just enough so the gpu is being held by the screw on the bracket. Sometimes the screw when fully tightened lifts the GPU out of it's socket. Boot back up and check GPU-Z and see if it changed to @ x16 3.0. 3) Your default PCIE slot settings are set to gen 1 or defaulting to gen 1. If it's still showing x16 1.1, go into your BIOS and find your PCIE port settings and change to the highest available gen. For your board in the BIOS, go to 'Advanced \ Chipset Configuration, change pcie1/pcie2/pcie3 link speed to highest gen for all pcie ports. Save, reboot, check GPU-Z.

Since we're on this subject, make sure your PSU GPU power cables are properly connected to the card as well. I would just disconnect and reconnect to be sure. What PSU do you have, how many watts?

Fix this first before looking further into memory.

Post back a screenshot of GPU-Z after you fix this.
Can you try a different version of timing configurator? Your DRAM frequency makes no sense and I wonder if this version doesn't work w/ your board. Uninstall timing configurator and try this one instead https://www.touslesdrivers.com/index.ph ... _langue=en Screenshot back like you did before w/ this version of timing configurator.
It is the same thing unfortunately
Image

But it is working in CPUZ
Image
In regards to your memory's RTL/IOLs, they look fine to me. The rule of thumb is that your RTLs shouldn't be more than 2 apart and your IOLs shouldn't be more than 1 apart. In your case, your RTLs are identical from Channel A Dimm 1 (slot 2) to Channel B Dimm 1 (slot 4). This is fine since it's within 2, in this case, identical. Your IOLs are only 1 apart from Channel A Dimm 1 (slot 2) to Channel B Dimm1 (slot 4). If you had your memory in slots 1 & 3, then it would be Dimm 0 for both RTL and IOL. This can be a little confusing to read and understand. I'm not a memory overclocking professional but I have an okay grasp on the basics. If you're interested on learning more, here's a legit source: https://github.com/integralfx/MemTestHe ... 20Guide.md

Also, I'd recommend downloading TM5 and running this for 15-30m to test your memory overclock's stability. Download link and info found in the above link I posted.
Attachments
RTL IOL.png
RTL IOL.png (59.04 KiB) Viewed 2806 times

Sirito
Posts: 55
Joined: 17 Jun 2022, 06:48

Re: Heavy character movement (not input lag)

Post by Sirito » 22 Jun 2022, 07:30

kokkatc wrote:
21 Jun 2022, 16:53
Sirito wrote:
21 Jun 2022, 15:21
kokkatc wrote:
21 Jun 2022, 13:24
kokkatc wrote:
21 Jun 2022, 12:28


Okay, first I'd focus on fixing your GPU. See attachments for a visual reference. Your GPU is operating at <x16 1.1> when it should be operating at <x16 3.0>. This is bad. Your GPU bandwidth is being severely limited which could easily explain the 'heavy movement' you're experiencing. The first attachment shows what your GPU is operating at and the 2nd attachment shows what IT SHOULD be running at.

First thing I'd do is to find out whether your GPU is in power savings mode or if your card isn't operating at the correct speed. Keep GPU-Z open, and run the render test in GPU-Z. Click the question mark next to the bus interface speed to run render test. See if the bus interface changes to <x16 3.0> during the benchmark/test. It could just be throttling down when idle but I doubt it. If it still shows your GPU is operating @ x16 1.1 during the benchmark, then you know you have a problem.

Three things can cause this. 1) Your GPU is installed in the incorrect PCIE slot. Make sure it's installed in the top slot. You can also refer to your motherboard's manual. 2) Your GPU isn't seated properly. Apply good force to make sure it fully seats and you should hear a click. Keep your case open and don't screw down the card all the way, just enough so the gpu is being held by the screw on the bracket. Sometimes the screw when fully tightened lifts the GPU out of it's socket. Boot back up and check GPU-Z and see if it changed to @ x16 3.0. 3) Your default PCIE slot settings are set to gen 1 or defaulting to gen 1. If it's still showing x16 1.1, go into your BIOS and find your PCIE port settings and change to the highest available gen. For your board in the BIOS, go to 'Advanced \ Chipset Configuration, change pcie1/pcie2/pcie3 link speed to highest gen for all pcie ports. Save, reboot, check GPU-Z.

Since we're on this subject, make sure your PSU GPU power cables are properly connected to the card as well. I would just disconnect and reconnect to be sure. What PSU do you have, how many watts?

Fix this first before looking further into memory.

Post back a screenshot of GPU-Z after you fix this.
Can you try a different version of timing configurator? Your DRAM frequency makes no sense and I wonder if this version doesn't work w/ your board. Uninstall timing configurator and try this one instead https://www.touslesdrivers.com/index.ph ... _langue=en Screenshot back like you did before w/ this version of timing configurator.
It is the same thing unfortunately
Image

But it is working in CPUZ
Image
In regards to your memory's RTL/IOLs, they look fine to me. The rule of thumb is that your RTLs shouldn't be more than 2 apart and your IOLs shouldn't be more than 1 apart. In your case, your RTLs are identical from Channel A Dimm 1 (slot 2) to Channel B Dimm 1 (slot 4). This is fine since it's within 2, in this case, identical. Your IOLs are only 1 apart from Channel A Dimm 1 (slot 2) to Channel B Dimm1 (slot 4). If you had your memory in slots 1 & 3, then it would be Dimm 0 for both RTL and IOL. This can be a little confusing to read and understand. I'm not a memory overclocking professional but I have an okay grasp on the basics. If you're interested on learning more, here's a legit source: https://github.com/integralfx/MemTestHe ... 20Guide.md

Also, I'd recommend downloading TM5 and running this for 15-30m to test your memory overclock's stability. Download link and info found in the above link I posted.
So I suppose there's no any problems with my ram currently, I also completed the test and it closed automatically after like 5 minutes and said no error detected, I have taken photos of some settings in my bios about the ram can you look at them if they'r important
In the last photo I found another fast boot option, I also have fastboot in boot section, I think this is the one in the ram section you meant me to disable it?
Image
Image
Image

kokkatc
Posts: 108
Joined: 23 Mar 2017, 13:49

Re: Heavy character movement (not input lag)

Post by kokkatc » 22 Jun 2022, 11:16

Sirito wrote:
22 Jun 2022, 07:30
kokkatc wrote:
21 Jun 2022, 16:53
Sirito wrote:
21 Jun 2022, 15:21
kokkatc wrote:
21 Jun 2022, 13:24


Can you try a different version of timing configurator? Your DRAM frequency makes no sense and I wonder if this version doesn't work w/ your board. Uninstall timing configurator and try this one instead https://www.touslesdrivers.com/index.ph ... _langue=en Screenshot back like you did before w/ this version of timing configurator.
It is the same thing unfortunately
Image

But it is working in CPUZ
Image
In regards to your memory's RTL/IOLs, they look fine to me. The rule of thumb is that your RTLs shouldn't be more than 2 apart and your IOLs shouldn't be more than 1 apart. In your case, your RTLs are identical from Channel A Dimm 1 (slot 2) to Channel B Dimm 1 (slot 4). This is fine since it's within 2, in this case, identical. Your IOLs are only 1 apart from Channel A Dimm 1 (slot 2) to Channel B Dimm1 (slot 4). If you had your memory in slots 1 & 3, then it would be Dimm 0 for both RTL and IOL. This can be a little confusing to read and understand. I'm not a memory overclocking professional but I have an okay grasp on the basics. If you're interested on learning more, here's a legit source: https://github.com/integralfx/MemTestHe ... 20Guide.md

Also, I'd recommend downloading TM5 and running this for 15-30m to test your memory overclock's stability. Download link and info found in the above link I posted.
So I suppose there's no any problems with my ram currently, I also completed the test and it closed automatically after like 5 minutes and said no error detected, I have taken photos of some settings in my bios about the ram can you look at them if they'r important
In the last photo I found another fast boot option, I also have fastboot in boot section, I think this is the one in the ram section you meant me to disable it?
Image
Image
Image
I'd suggest trying to change command rate to 1N as shown below in attachment. Just keep in mind that not all ram kits will be stable at 1N. If your KIT can't run 1N at all, your system may not boot and you'll have to hard reset and reconfigure your bios settings, FYI.

So for FAST BOOT, basically the way this works is that if FAST BOOT is enabled, it will bypass memory training at boot and use the settings/values already configured in your BIOS. If you DISABLE FAST BOOT, your memory will train at every boot which also means your system will take a little bit longer to boot since it's training. You don't always want this to be enabled because then memory settings could change without you even knowing and suddenly your game/mouse/etc might feel different and you won't know why. Once you find a spot where you're satisfied with your memory settings, you can ENABLE FAST BOOT so your settings remain locked in after every boot.
Attachments
COMMAND RATE.png
COMMAND RATE.png (1.48 MiB) Viewed 2687 times

Sirito
Posts: 55
Joined: 17 Jun 2022, 06:48

Re: Heavy character movement (not input lag)

Post by Sirito » 22 Jun 2022, 11:31

kokkatc wrote:
22 Jun 2022, 11:16
Sirito wrote:
22 Jun 2022, 07:30
kokkatc wrote:
21 Jun 2022, 16:53
Sirito wrote:
21 Jun 2022, 15:21


It is the same thing unfortunately
Image

But it is working in CPUZ
Image
In regards to your memory's RTL/IOLs, they look fine to me. The rule of thumb is that your RTLs shouldn't be more than 2 apart and your IOLs shouldn't be more than 1 apart. In your case, your RTLs are identical from Channel A Dimm 1 (slot 2) to Channel B Dimm 1 (slot 4). This is fine since it's within 2, in this case, identical. Your IOLs are only 1 apart from Channel A Dimm 1 (slot 2) to Channel B Dimm1 (slot 4). If you had your memory in slots 1 & 3, then it would be Dimm 0 for both RTL and IOL. This can be a little confusing to read and understand. I'm not a memory overclocking professional but I have an okay grasp on the basics. If you're interested on learning more, here's a legit source: https://github.com/integralfx/MemTestHe ... 20Guide.md

Also, I'd recommend downloading TM5 and running this for 15-30m to test your memory overclock's stability. Download link and info found in the above link I posted.
So I suppose there's no any problems with my ram currently, I also completed the test and it closed automatically after like 5 minutes and said no error detected, I have taken photos of some settings in my bios about the ram can you look at them if they'r important
In the last photo I found another fast boot option, I also have fastboot in boot section, I think this is the one in the ram section you meant me to disable it?
Image
Image
Image
I'd suggest trying to change command rate to 1N as shown below in attachment. Just keep in mind that not all ram kits will be stable at 1N. If your KIT can't run 1N at all, your system may not boot and you'll have to hard reset and reconfigure your bios settings, FYI.

So for FAST BOOT, basically the way this works is that if FAST BOOT is enabled, it will bypass memory training at boot and use the settings/values already configured in your BIOS. If you DISABLE FAST BOOT, your memory will train at every boot which also means your system will take a little bit longer to boot since it's training. You don't always want this to be enabled because then memory settings could change without you even knowing and suddenly your game/mouse/etc might feel different and you won't know why. Once you find a spot where you're satisfied with your memory settings, you can ENABLE FAST BOOT so your settings remain locked in after every boot.
I noticed that boot is slower when I fully unplug my pc (from the wall outlet) so maybe the training is done each time I power off the powersupply completely? and that's why I feel different each time it boots (because like each day 50% a 1 power cut happens in my country).
But actually the "feel" we talking about isn't huge so it's not like day and night, it is just a small amount and it is connected only to the heavy character, my mouse is always in a good state, I can always flick and because my muscle memory always works I think the mouse always works too.
Maybe I change fast boot to enabled not auto, but I'm thinking are the trained values currently assigned good or not?
About the 1N I will try tomorrow after my exam because I'm using the PC rn and can't risk a cmos reset.

kokkatc
Posts: 108
Joined: 23 Mar 2017, 13:49

Re: Heavy character movement (not input lag)

Post by kokkatc » 23 Jun 2022, 13:39

Sirito wrote:
22 Jun 2022, 11:31
kokkatc wrote:
22 Jun 2022, 11:16
Sirito wrote:
22 Jun 2022, 07:30
kokkatc wrote:
21 Jun 2022, 16:53


In regards to your memory's RTL/IOLs, they look fine to me. The rule of thumb is that your RTLs shouldn't be more than 2 apart and your IOLs shouldn't be more than 1 apart. In your case, your RTLs are identical from Channel A Dimm 1 (slot 2) to Channel B Dimm 1 (slot 4). This is fine since it's within 2, in this case, identical. Your IOLs are only 1 apart from Channel A Dimm 1 (slot 2) to Channel B Dimm1 (slot 4). If you had your memory in slots 1 & 3, then it would be Dimm 0 for both RTL and IOL. This can be a little confusing to read and understand. I'm not a memory overclocking professional but I have an okay grasp on the basics. If you're interested on learning more, here's a legit source: https://github.com/integralfx/MemTestHe ... 20Guide.md

Also, I'd recommend downloading TM5 and running this for 15-30m to test your memory overclock's stability. Download link and info found in the above link I posted.
So I suppose there's no any problems with my ram currently, I also completed the test and it closed automatically after like 5 minutes and said no error detected, I have taken photos of some settings in my bios about the ram can you look at them if they'r important
In the last photo I found another fast boot option, I also have fastboot in boot section, I think this is the one in the ram section you meant me to disable it?
Image
Image
Image
I'd suggest trying to change command rate to 1N as shown below in attachment. Just keep in mind that not all ram kits will be stable at 1N. If your KIT can't run 1N at all, your system may not boot and you'll have to hard reset and reconfigure your bios settings, FYI.

So for FAST BOOT, basically the way this works is that if FAST BOOT is enabled, it will bypass memory training at boot and use the settings/values already configured in your BIOS. If you DISABLE FAST BOOT, your memory will train at every boot which also means your system will take a little bit longer to boot since it's training. You don't always want this to be enabled because then memory settings could change without you even knowing and suddenly your game/mouse/etc might feel different and you won't know why. Once you find a spot where you're satisfied with your memory settings, you can ENABLE FAST BOOT so your settings remain locked in after every boot.
I noticed that boot is slower when I fully unplug my pc (from the wall outlet) so maybe the training is done each time I power off the powersupply completely? and that's why I feel different each time it boots (because like each day 50% a 1 power cut happens in my country).
But actually the "feel" we talking about isn't huge so it's not like day and night, it is just a small amount and it is connected only to the heavy character, my mouse is always in a good state, I can always flick and because my muscle memory always works I think the mouse always works too.
Maybe I change fast boot to enabled not auto, but I'm thinking are the trained values currently assigned good or not?
About the 1N I will try tomorrow after my exam because I'm using the PC rn and can't risk a cmos reset.
If you unplug your power cable completely from the power supply and leave it unplugged for a 'period' of time, it could reset your bios settings. That 'period of time' can differ from mobo to mobo so keep that in mind.

To lock in your memory settings, first set fast boot to disabled, then reboot, your memory will train. Once you boot into windows, take a look at your RTLs/IOLs as previously discussed and make sure your RTLs are no more than 2 apart, your IOLs no more than 1 apart. If you confirm this, then reboot, go back into your bios, change to fast boot ENABLED. Your settings will remain and memory won't retrain at every reboot.

Post Reply