Shutters while shooting and seeing enemy even in CS 1.6

[Uğur Uyar]

I have 10200-H and 1650Ti and easily get 200 fps (locked) on CS 1.6 and getting around 150-170 fps on CS2 but in CS2 there is spiking while i see enemy and shoot to him. Its like for just 1 seconds but its enough to make me dead. I just surprised when it happens on cs 1.6. Not happens so much in 1.6 but still annoying with that specs. How can i find the problem? With using xperf? Or any app?

 

[ir_cow]

Try with v-sync off?

 

[Uğur Uyar]

in CS2 its already off and there is no v-sync. Its just fps_max 200 command for limiting

 

[gord787]

I have 10200-H and 1650Ti and easily get 200 fps (locked) on CS 1.6 and getting around 150-170 fps on CS2 but in CS2 there is spiking while i see enemy and shoot to him. Its like for just 1 seconds but its enough to make me dead. I just surprised when it happens on cs 1.6. Not happens so much in 1.6 but still annoying with that specs. How can i find the problem? With using xperf? Or any app?

Hey! Try FRAPS to monitor FPS, GPU-Z for GPU usage, and CPU-Z for CPU usage. Also, check Task Manager for background processes. Tweaking CS2's graphics settings might help too!

 

[Uğur Uyar]

nothing works at background except steam and the game.

 

[LabRat 891]

H? Its a laptop?

Not using WiFi, are you?
Would check and make sure any power saving/allow windows to turn off this device to save power
are disabled on iNIC and WiFi.

Also, would try manually updating your drivers for both network adapters.

 

[Ruru]

H? Its a laptop?

Yeah as OP has also a GTX 1650 Ti, that's never released for desktops.

...unless it's some kind of a Frankenstein Chinese motherboard.

 

[RJARRRPCGP]

It may also be packet loss.

 

[Vincero]

Set up a Lan game with bots on your device - no need to actually get anyone to join - see if the behaviour happens in game regardless of player type / origin.

I was under the impression that the source engine loaded most needed assets at round / level start, so no additional loads should happen except in specific cases. Seeing an enemy or shooting at them shouldn't be triggering a data transfer load from hard drive.

Another option may be to fire up latencymon (https://www.resplendence.com/latencymon) and seeing when a spike happens if it's attached to a certain windows component or device driver.
Just for clarity, you are best off starting the game and getting into a map then starting latencymon - that way it will only track what happens during the play time and not any latency loading the game files / graphics mode switching, etc. (which in theory shouldn't be an issue unless you were loading data off of a failing hard drive, etc.)

 

[eidairaman1]

Old ass games arent guaranteed to work properly on the latest hardware. Deus Ex which was from 1999, ran in hyper mode on a Radeon 9700 Pro in 2003, had to max graphics out and run another task that accessed the gpu in background, just as an example

 

[LabRat 891]

Old ass games arent guaranteed to work properly on the latest hardware. Deus Ex which was from 1999, ran in hyper mode on a Radeon 9700 Pro in 2003, had to max graphics out and run another task that accessed the gpu in background, just as an example

They mention it happens in Counter Strike: Global Offensive (AKA 'CS2') too.

in CS2 there is spiking while i see enemy and shoot to him.

I assume it's network device/driver-related.
Maybe, a process sitting on a 'sleeping' core that's taking a moment to throttle back up when suddenly tasked? Anti-Cheat?

Another option may be to fire up latencymon (https://www.resplendence.com/latencymon) and seeing when a spike happens if it's attached to a certain windows component or device driver.

This.

 

[Uğur Uyar]

Set up a Lan game with bots on your device - no need to actually get anyone to join - see if the behaviour happens in game regardless of player type / origin.

I was under the impression that the source engine loaded most needed assets at round / level start, so no additional loads should happen except in specific cases. Seeing an enemy or shooting at them shouldn't be triggering a data transfer load from hard drive.

Another option may be to fire up latencymon (https://www.resplendence.com/latencymon) and seeing when a spike happens if it's attached to a certain windows component or device driver.
Just for clarity, you are best off starting the game and getting into a map then starting latencymon - that way it will only track what happens during the play time and not any latency loading the game files / graphics mode switching, etc. (which in theory shouldn't be an issue unless you were loading data off of a failing hard drive, etc.)

Going to try this. Btw it happens in Bannerlord (its optimized so well). When i die game shutters for less than a second sometimes my mouse is freezing like its disconnecting and connecting again in seconds. But this mouse problem only happens in Bannerlord.

How can I disable power savings for wifi?

I

This is the results of playing with bots for a while. I didnt notice any shutter except the MVP screen in first round. But there is important thing. Normal players shooting while peeking or they are peeking fastly. Bots never peek and they dont react so much. Most of the time it happens when enemy rushing to me like crazy.

Btw the interrupt to process latency was green before I quit the game. It just became red when i quit game. I dont know how to stop testing without alt-tabbing or quitting game. But DPC routine execution was red unless interrupt latency

 

[Vincero]

Hmmm, 8042 port driver would only be used for PS/2 devices - it's possible the laptop keyboard / trackpad internally are using this interface (gets around any USB emulation issues at boot-time). I'm not expecting any recent laptops would have a PS/2 port on them, even with a docking station.
Latency for that shouldn't really be that high - it's not a particularly complex thing (<250 micro-sec will not be causing seconds of stuttering).

The high latency from the DX subsystem is common when switching between modes - so as you say when exiting game it spikes.

Do you have any special drivers / software loaded for keyboard/trackpad/mouse?

Also, with latencymon running, see if doing certain specific hardware intensive things causes an issue, e.g. copy a large file from one folder to another (copy not move).
Also, if this is more an online thing, instead of playing, spectate an online game instead and see if any spikes are triggered.

 

[LabRat 891]

Hmmm, 8042 port driver would only be used for PS/2 devices - it's possible the laptop keyboard / trackpad internally are using this interface (gets around any USB emulation issues at boot-time). I'm not expecting any recent laptops would have a PS/2 port on them, even with a docking station.
Latency for that shouldn't really be that high - it's not a particularly complex thing (<250 micro-sec will not be causing seconds of stuttering).

The vast majority of laptops I've seen, use PS/2 for the touchpad. I've also seen a corrupt / outdated touchpad driver cause some weird issues. (Also, dedicated hotkeys MIGHT be non-USB too)

The high latency from the DX subsystem is common when switching between modes - so as you say when exiting game it spikes.

Switchable graphics issue? IIRC, in Win11 I've had both positive and negative performance impacts from having rendering passed-thru to another display device(GPU).

Do you have any special drivers / software loaded for keyboard/trackpad/mouse?

I haven't ever seen a laptop that doesn't. We're gonna need the exact model of the laptop to point them @ the OEM/ODM-supplied driver(s).

Also, with latencymon running, see if doing certain specific hardware intensive things causes an issue, e.g. copy a large file from one folder to another (copy not move).
Also, if this is more an online thing, instead of playing, spectate an online game instead and see if any spikes are triggered.

Updating UEFI/Firmware/BIOS might be a good idea, too. My Latitude 5411 was not 'properly' clock shifting on the firmware it came with.
I also wonder if TPM might be related.

So, OP.
What's the model of this laptop?
Seems manufacturer/model-specific drivers and/or firmware might need updating.

 

[Vincero]

I haven't ever seen a laptop that doesn't. We're gonna need the exact model of the laptop to point them @ the OEM/ODM-supplied driver(s).

Depends if they've wiped it from the factory install and done fresh install. Few actually 'need' the manufacturer drivers for touchpad or keyboard functions (even including the extra buttons / extended functions such as display switching, volume, etc.), but I'll admit I'd install driver / support software if it improved the extended functions in some way. That said, if it was a detriment then it would be uninstalled just as quick.

TPM / firmware issue is possible... but I'd have expected a latency peak around that if a TPM device has triggered it specifically - unfortunately if it's a low-level firmware / interrupt action it may be harder to track to a specific thing.
Turning off speed stepping / boost and extended C-states would potentially rule out a frequency scaling / core sleeping state problem if firmware updates aren't available (or don't fix it - but that's not really a solution).

One other out there suggestion is to either turn off or reduce Windows virtual memory to a minimum (usually it can't be smaller than ~200-300MB if enabled) - this will prevent applications from paging to disk (even an SSD is way slower than RAM and if the SSD is doing a hold for some reason, maybe trimming or other GC action, it can pause the system) and rule that out. Most likely it is at auto which will allow it to scale usage as it sees fit including paging information to virtual memory even when not needed (i.e. nowhere near out of RAM) - e.g. my work system has 32GB RAM and even though 13GB is free, Windows has still has a 2% pagefile use - there are good reasons for this behaviour but isn't needed just to test / fire-up counterstrike.

 

[Uğur Uyar]

My laptop brand is Monster from Turkey but originally its Clevo NH58DEQ. Bios is updated, and yes there is a control center but its totally sucks and makes system much more unstable so i uninstalled it with powershell. My C-States are closed should i enable it again? And yes it doesnt have MUX switch so display coming from integrated gpu and that a huge problem too but i dont think its related with it. Anyways how can i disable power savings for WIFI? I wanna try it.

Do you have any special drivers / software loaded for keyboard/trackpad/mouse?

I dont know anything except Clevo control center.

 

[LabRat 891]

My laptop brand is Monster from Turkey but originally its Clevo NH58DEQ. Bios is updated, and yes there is a control center but its totally sucks and makes system much more unstable so i uninstalled it with powershell. My C-States are closed should i enable it again? And yes it doesnt have MUX switch so display coming from integrated gpu and that a huge problem too but i dont think its related with it. Anyways how can i disable power savings for WIFI? I wanna try it.

View attachment 358123


I dont know anything except Clevo control center.

Holy crap, Brave AI Search is getting better!

Disable power savings wifi ethernet for gaming​

To ensure a stable and high-performance gaming experience, it’s recommended to disable power savings for both WiFi and Ethernet connections. Here’s a step-by-step guide:
For WiFi:

1. Device Manager: Open Device Manager (Press the Windows key + X and select Device Manager).
2. Network Adapters: Expand the Network Adapters section and find your WiFi adapter (e.g., Intel WiFi 6).
3. Properties: Right-click on the adapter and select Properties.
4. Power Management: In the Properties window, uncheck the box next to “Allow the computer to turn off this device to save power”.

For Ethernet:

1. Device Manager: Open Device Manager (Press the Windows key + X and select Device Manager).
2. Network Adapters: Expand the Network Adapters section and find your Ethernet adapter (e.g., Realtek PCIe GBE Family Controller).
3. Properties: Right-click on the adapter and select Properties.
4. Power Management: In the Properties window, uncheck the box next to “Allow the computer to turn off this device to save power”.

Additional Tips:

• Ensure your network drivers are up-to-date, as outdated drivers may cause issues with power management.
• If you’re using a Realtek Ethernet adapter, you may need to update the driver to access the Power Management tab.
• For some systems, you might need to modify the registry to disable power savings. However, this is not recommended unless you’re experienced with registry edits, as it can potentially cause system instability.

By disabling power savings for both WiFi and Ethernet, you’ll ensure a consistent and high-performance connection for gaming.
SOURCE

(not trying to be rude. I'm just lazy )

 

[Vincero]

My laptop brand is Monster from Turkey but originally its Clevo NH58DEQ. Bios is updated, and yes there is a control center but its totally sucks and makes system much more unstable so i uninstalled it with powershell. My C-States are closed should i enable it again? And yes it doesnt have MUX switch so display coming from integrated gpu and that a huge problem too but i dont think its related with it. Anyways how can i disable power savings for WIFI? I wanna try it.

View attachment 358123


I dont know anything except Clevo control center.

C-states being enabled will not help stuttering issues - you can try switching the option in BIOS (just because the option is there doesn't mean it works properly) and seeing if that makes a difference.

Wifi power settings are normally either in the control software for it OR you can find it in the windows power plan settings (i.e. within balanced / high performance / etc. - whichever you have active) if you go in to the plan settings there may be an option for it, or lastly the device manager additional properties will have the generic power option (doesn't always make a difference). Assuming you're using an SSD in the laptop, you can disabled the hard drive power-down time / set it to a really high number just in case that is somehow the issue (although really unlikely if its the only drive in the system).

Try setting paging file to 'no paging file' temporarily. You've got 16GB RAM so there shouldn't be an issue booting and running Steam + Source Game. Also, if you set a specific file size choose a single initial/max size - allowing size scaling doesn't really help unless you are tight on drive space in which case running out of space for the page file to fill in to will crash program, etc.

 

[Uğur Uyar]

Holy crap, Brave AI Search is getting better!

(not trying to be rude. I'm just lazy )

Lol its okay thank you

Try setting paging file to 'no paging file' temporarily. You've got 16GB RAM so there shouldn't be an issue booting and running Steam + Source Game. Also, if you set a specific file size choose a single initial/max size - allowing size scaling doesn't really help unless you are tight on drive space in which case running out of space for the page file to fill in to will crash program, etc.

Okay.

 

[Vincero]

View attachment 358124

If there's no specific option for it then (as most WIFI cards are a PCI device normally) the PCIe power savings option will dictate how aggressive power saving may be for it.
The device manager 'Allow the computer to turn off this device to save power' option may help, but generally that's usually not applied to an active device - within the device properties (where it will list things such as radio bands/frequencies, offloading, etc., there may be a power saving option.
e.g.
View attachment 1723211554126.webp

To be honest I don't think the Wifi is likely to be the culprit - Source games usually have a constant stream of data when in an online/LAN game (how do you think the ping time constantly being updated is measured) - it's unlikely it would have a chance to go to a reduced power state or sleep if already in a session as it will be constantly sending and receiving within a <100ms window.

 

[Uğur Uyar]

Damn I think that explains a lot of thing.

To be honest I don't think the Wifi is likely to be the culprit - Source games usually have a constant stream of data when in an online/LAN game (how do you think the ping time constantly being updated is measured) - it's unlikely it would have a chance to go to a reduced power state or sleep if already in a session as it will be constantly sending and receiving within a <100ms window.

Should I use Xperf for detecting the issue?

Is that much HID complimant things normal?

 

[Vincero]

Should I use Xperf for detecting the issue?

You can try, although keep in mind anything that does logging will add extra load - also keep in mind sometimes they may force the need to enable the page file depending on how much they generate.

 

[Uğur Uyar]

I disabled page file disabled usb suspending and disabled wifi power savings but still same.

 

[Vincero]

Is that much HID complimant things normal?
View attachment 358132

It's a laptop and the USB-HID class basically allows the manufacturer an easy interface / driver class to use for any additional switches/functions that are not standard. Can't say I've seen many with that much going on.

If you have an alternate keyboard and mouse you could try disabling the laptop ones (there is usually a button to kill the trackpad, and sometimes the keyboard also - maybe only in BIOS options) and seeing if the glitch still happens - that at least will rule them out.

I disabled page file disabled usb suspending and disabled wifi power savings but still same.

Well, it makes no odds really enabling / disabling the wifi adapter power saving options (unless you want to get the best economy you can from them - or use sleep mode often), so up to you if you want to put them back to default. Same goes for the swap file options, although I wouldn't necessarily set it how you had it set before. How Windows 10/11 use the page file isn't the same as how some older versions of Windows managed things and is a lot less aggressive, e.g.

32GB RAM, 20GB memory load = 7GB swap file, only 2% used. That's with system managed settings - really I could fix the page file at a fraction (maybe 1/2) of that size and have no issue... probably get slightly better performance and boot SSD life... too lazy.

Normally Paging RAM pages out to disk is only done if required by the app or if Windows thinks some memory pages are infrequently used enough that it would have little impact to performance and frees up actual RAM for applications. Unless you are in a low RAM situation this then falls back to using the page file to store RAM contents and switching things in and out of the page file as needed.

In reality with your 16GB RAM capacity I don't think you'd even use up a 16GB page file unless you are working with some big datasets (and even then, the performance hit would be painful).
Some people set it to a size just big enough to allow Windows crash dumps to work properly, or based on their expected workloads. The idea of setting it to .5x-1.5x your memory size or some such rule is flawed for most peoples use.

One other thing to look at - assuming Xperf doesn't yield anything useful - check your Event Log for any events around the time a glitch happens, you may find something such as a bad disk access, GPU reset or other issue, e.g.:

This would be under the 'Windows Logs'->'System' log area

 

[Uğur Uyar]

Will check it thanks.

There is a lot of warnings and errors. How can i send all of it like a minidump file?