Sound pops / distorted sound - nVidia driver issue

System specs:

i9 9900k / ASUS motherboard / 16GB RAM, EVGA 2080 / Windows 10 / 461.40 Drivers

This issues started in Autumn last year. I believed it to be a faulty sound card (SoundBlaster Z) so I purchased a SB AE5-Plus but sadly the issue was still present. All previous drivers had been uninstalled prior to the installation of the new sound card. I next moved onto the wiring for the speakers (Logitech Z906) just in case one of the cables was faulty. All was replaced but the issue was still present.

Rebooting the computer can alleviate the issue for a little while but it comes back at random times. My PSU decided to start playing up in November last year. Random hard resets every now and then so a new PSU (850 Corsair) was purchased and the hard resets problem was sorted.

I borrowed another set of speakers thinking that maybe it's my speaker system that was starting to play up. The problem was still there. I switched to onboard sound as a last resort and I thought I may had cracked it as my system went for more than 24 hours with no issues but it came back whilst watching something on YouTube.

I had a light bulb moment one afternoon. There is a sound driver installed within the nVidia drivers. I decided to uninstall my drivers with DDU and install the same drivers but disabling the sound. From that moment onwards my sound issue was solved until a couple of days ago when I installed the new 461.40 minus the sound drivers. It's back again. The sound pops / distortion only happens when there is sound coming out of the speakers, no other time.

Has anyone else had this issue or have heard anyone with something similar as it's really annoying. As far as I'm aware I have not had the issue when I've been using my SteelSeries Arctis Pro Wireless headphones gaming.

I would naturally lean towards it being a sound card driver issue but like I said above, once the sound driver was uninstalled from the nVidia drivers, the problem went away...

 

Is your PCIe Bus saturated, by any chance?

Are your cables (especially audio ones) properly shielded/is there any possible interference?

Do you have any IRQ conflicts? You might want to try to use MSI Mode Utility and enable or disable MSI mode for the video card and various sound devices you may have and see if that helps.

 

Sound popping/distorted sound is an issue that actually started with the 461.x drivers and so far hasn't been fixed! I mentioned this before, with proof and there's a real arrogant member here that need to be ignored when reporting this issue. In fact I did an update to this in the previous 461.33 driver thread where I said this :

I've gone over to try this on my 980 GTX and can confirm very high DPC rates as well with the latest 461.09 and 461.33 drivers, again noting sound popping and crackling in both Warzone and other gaming related software. I went did a DDU, went back to 460.79 and same as before the latency is much much lower, with no crackling and no DPC complaints. So that's both the RTX 2070 and GTX 980 suffering from high DPC Latency and sound popping/crackling issues with the latest drivers ( Both Intel systems, both windows 10 ). I even went ahead and updated network drivers with Nvidia HDMI sound disabled to see if that'd fix it, but no. The only change to fix is downgrading the drivers back to the 460's or below.

Other users are also reporting the same

1. I`m getting that in World of Warships I thought it was the game as its just been updated but interesting you say that.

461.06, 461.33, 461.40 - High Latency : Sound Popping/Cracking



Now compare with the 432.00 and 460.79 ( Low Latency ) No Issues

 

there is actually a depop setting on asus boards.

 

Thank you for the replies guys.

Your findings are very interesting Neil79. Have you shown them to nVidia ?

I'm glad it's a driver issue anyway. Hopefully someone from nVidia can shine some light on this subject. For now, I'll revert back to the older drivers until it's fixed.

Where abouts is this "Depop" setting in the BIOS ? That's a new one on me.

*Edit: I've just run the same program as you and I too am getting the same results. High latency.

 

Try the 460.79 drivers, doing a "clean" install option when installing them over the top, does your latency improve?

 

I have just run DDU and then installed the 460.79 drivers with a clean install. I've run the latency test and now it's reporting everything is fine. Good catch on this one Neil79. That looks to be the problem !

 

not gonna say dpc latency crackling is common on nvidia, but it pops every now and then.to the point it's probably best to do a quick check after every driver change.

seen it on my 1080fe when I bought it on launch.it's subtle but once you hear it,it's gonna drive your brain crazy

 

Whatever you do, do not go onto the GeForce forums. You'll be bombarded by a member called Sora, who is a very hostile member to anyone trying to report driver issues. Trying to belittle them as if they are clueless to computers.

The exact same person that caused a lot of hostility on the Auzentech forums back in the day and other forums who eventually got banned.

I think it may have also have been the same person that got banned here years ago. Squall Leonhart something...

 

Oh yes, it drives me nuts ! "Pops" is probably a more accurate representation of what I'm hearing.

 

what I heard was more like there was a bonfire somewhere in my room but I didn't smell sausage

and before you ask,no,pissing on the computer didn't put a stop to it.

 

Put the sausage away and the fire might go out

 

That's great! Wonder what's wrong with the 461.x and 465.x then

 

Sounds like Astyanax who I recently put on ignore for trolling members when they have issues with the drivers. Thankfully I fixed my issue ( bad drivers ), which he didn't!

 

Okay, I still have the issue with the 460.79 drivers. I will uninstall them again tomorrow but this time leave out the sound driver within the nVidia drivers as that worked for me last time. I'll update on my findings.

 

he refuses to do any form of tracing or testing, or investigation of his own machine for the actual cause, which is a requirement for any action to be taken on DPC reports.

He has never even provided sufficient system information to begin searching for and collecting other cases.

There is no such thing as irq sharing on PCIE platforms, the interrupts are virtualized/ emulated via message lines and even two devices with the same legacy interrupt number run in parallel

This setting (should be an always on one) is made to ensure that HDA devices resume from a hybrid boot (fast startup) state in a fully functional state by ensuring ACPI compliance is met.

You'll actually be told to check your system first and foremost for physical / electrical (especially house wiring grounding and earth loops) concerns (especially faulty dp cables or monitors) that are the usual cause of DPC issues that are only occuring with less than 2 handfuls of users and then to submit the requisite trace files, system information and kernel dumps once the issue has been reproduced and still occurs at a minimized case reduced to just the components necessary to boot the machine.

99% of driver issues reported are system specific, never get investigated and usually don't occur in the next driver even though no changes were made at all and likely resolved themselves on the machine through the alignment of stars, sacrificing of a goat or simply not installing a new driver the same day as windows updates drop. (guess what happens when WU updates the runtimes and your driver has already started making new shader caches for the old runtime)

There was never any hostility on the auzen forums , I provided functional Prelude drivers where Auzen couldn't due to their reliance on Creative Labs not providing broken APO libraries and only ever criticized the Auzen engineers, I'm sure there were a couple of auzen fanboys that got shitty about it but i was really just a disgruntled product owner having too many issues with a product they sold me,

I think there was only ever one issue which was the coz of my ban here in 2010 where a user corrupted his window install through his own fault and started harassing me on all the forums i'd been posting the Remix packages to, that resulted in both of us being perma banned (throwing insults in the thread), I threw in the towel with maintaining updated auzen packages when i'd had enough of the continuing bsods in openal bsods (lotro) and the screeching issue and my account on the auzen site was abandoned, it wasn't even till just recently i stopped seeding the auzen remix (though the torrent site i posted them too was killed prior to that anyway).

Oh, and Hilbert and Unwinder are already are i'm here, so get over it, if these users want the issues fixed they need to prove, log, trace, and isolated hardware test before nvidia will touch them,

I'm not saying this as a tryhard or making crap up, i've been explicitly told this by Manuel and a number of engineers i've been in touch with in all the years i've been feeding obscure or overlooked user reports through to them.

You might actually be confusing me with one of the other various users that have an account named Squall Leonhart, whilst i was banned from here in my stupid teen years for persistently causing fights in the ATI section (and again for the auzen fight), i've not actually been banned anywhere else permanently (except for maybe a now defunct tech forum that was getting misleading performance information posted by a certain AMD marketting agent in the lead up to bulldozer)


Actually I'm astonished things i would have gotten warned for back then are occuring now in the AMD/Intel section by Intel/AMD fans and how much bs the AMD section talks about nvidia whilst nvidia's section hardly talks about AMD

Far from what many believe, I do push nvidia to get crap fixed when It turns out to be a widespread issue (and yes, even obscure ones such as a driver start failure on certain gigabyte mainboards because they never implemented 64bit ACPI tables properly tend to be wide spread enough to find multiple cases)

Now on to the thread....

Inadequate information supplied,

Several motherboards have fixable DPC issues because they have SMBUS attached devices that saturate the system with Interrupt Requests, or misconfigured Intel i series ethernet controllers (either configure the nic interrupt moderation or seek updates for the smbus device).

Someone on ryzen setup reported consistent dpc misbehavior across all tested drivers in 2019 and even throughout his nvidia thread, reddit (and i believe this forum?) people with same hardware could not replicate it. The guy refused to rma/replace the cpu and power supply which was all he hadn't changed at all, despite evidence that his particular one has had dpc problems previously (Yes, even inconsistent / unstable power can cause dpc spikes)

Kepler DPC issue came down to a long loop operating on core0 when the machine was Hyperthreaded) and had multiple confirmable / traceable sources.

I've seen no spikes in any of my scouring of the internet that do not involve the device going from P3-P0 when a game or browser is started, which are expected behavior reported anywhere, not even amongst the worst of the worst mainboards for RGB saturation.


The machine needs to be taken back to bare basics with a fresh windows install on an alternative drive (to rule out system drive issues), tested on an alternate power socket (to rule out wiring) and even at different house to be thoroughly sure the issue is actually drivers. (some of which is not practical right now, if ever)

The cmos needs to be wiped to rule out any persistent nvram entries, and hardware has to be swapped to rule out bus faults in the cpu pcie controller, the mainboard or the graphics card,
PSU swapped out because i know for damn sure that some active PFC units only work properly on 60hz(110v/USA) and not on 50hz(230v/Aus/UK/Other) grids *cough*Be Quiet*cough*

oh, just as a possibility, try an different tv or display, because of the programmable scaler in Turing and Ampere displays themselves can be a compatibility factor for misbehaving Interrupts, particularly on Adaptive/Variable panels.

 

Dropped back down to 461.40 again and all's right with the sound again. I tried every driver up to the present date yesterday and all had sound issues. 100% driver related.

 

And it'll stay that way until you provide the required traces and system info.

 

Thanks for the information but sadly I don't have time to sit down for hours and hours try to get to the bottom of the issue, I wish I had but normal life / running a business gets in the way. Leads had been changed previously to no avail. New PSU. New sound card and even new speakers...

System spec:

i9 9900k
ASUS Z390 Pro Gaming Motherboard (latest BIOS)
16GB RAM Corsair
EVGA 2080
Windows 10
(Now running 461.40)

 

I never install the Nvidia sound drivers annoys me when they keep being selected as the default device, yet even though I untick the box and perform a clean install every time, it still shows as a audio device in my system, I have a proper full on sound system with a sound card I dont need nor want the GPU to provide sound but you cant stop its installation it seems