TR Forums

demolition wrote:

notfred wrote:

Do you have a UPS? Does the issue happen if running direct from the wall power without the UPS? Does it happen if you unplug the UPS from the wall and run from the UPS battery?
I'm still suspecting something else stealing cycles like a graphics card or even a bad USB device where the processor has to keep retrying.

No, I do not have a UPS.

I would also consider the graphics card or a USB device if it hasn't been for the fact that I had already tried stripping the system down to the bare minimum without it resolving the issue. That means removing the gfx card and just using the onboard (Intel Gfx), only having a single SATA device connected with a fresh Win10 install, and no USB devices except keyboard and mouse. And also replacing the PSU with a brand new Seasonic one that I was able to borrow for the test.

I found posts in several forums about people having similar issues with the audio (mostly with the onboard Realtek audio) and some of them were able fix it by disabling some CPU power saving features. One method which seemed to help many was to create a new power scheme in Windows and set the minimum processor state to 100%. I tried this with no improvement. I also went through the BIOS to try and find and disable any power saving / throttling features there as well, but could not find much.

We're rapidly approaching the point of needing a young priest and an old priest.

ludi wrote:

We're rapidly approaching the point of needing a young priest and an old priest.

Well, I also considered whether it could be malware of some sorts, however I do not know how that would make it on to a freshly installed Windows which was done with all my other HDDs and storage devices disconnected.

This morning, I tried turning off all dynamic CPU stuff like speed stepping, Cx states, turbo core etc. so the CPU should run at a constant 3.7GHz with all cores active 100% and with a fixed voltage. Didn't help. I will say though that the issue is less prevalent with the new MB/CPU, maybe just 20% as bad as on the old system, but that is still not good enough.

demolition wrote:

ludi wrote:

We're rapidly approaching the point of needing a young priest and an old priest.

Well, I also considered whether it could be malware of some sorts, however I do not know how that would make it on to a freshly installed Windows which was done with all my other HDDs and storage devices disconnected.

This morning, I tried turning off all dynamic CPU stuff like speed stepping, Cx states, turbo core etc. so the CPU should run at a constant 3.7GHz with all cores active 100% and with a fixed voltage. Didn't help. I will say though that the issue is less prevalent with the new MB/CPU, maybe just 20% as bad as on the old system, but that is still not good enough.

Seems about the only thing left is to try another OS..........................................................

You guys haven't tried setting the PC up in another house yet.

I've had a case where a stereo was getting pops from the speakers after a few minutes and try as I might it wouldn't go away. But when we took the whole thing to the audio store to get it checked out, the same setup, amp, speakers, cables and all played perfectly for hours. Turned out power was just really dirty in the room where I had it set up originally.

The solution was a servo-motor AVR and two toroidal transformers. AVR first to try and smooth out the incoming power as much as possible, then into one transformer to step up, then into the second transformer to step back down. I think the trick really was the two toroidal transformers which completely isolates it from the mains and makes sure the power coming into the amp is all from inductance.

This is what I’ve for the AV rack. It has helped with the old wiring in our 50’s house.
PYLE PCO850 15 Amp Power Supply Conditioner with Rack Mountable Power Strip and Surge Protector with 9 Outlets https://www.amazon.com/dp/B00BQO5G4O/re ... eBbDP4DY79

jihadjoe wrote:

You guys haven't tried setting the PC up in another house yet.

No you are right, I haven't. I find it hard to think that should be the source of the problem though due to the nature of it. Mains noise would generally be added to the audio output, just causing missing samples. The PSU should also be able to take care of most of mains noise and I found no difference between two different PSUs.

I will however, build a PC using my old MB and CPU. I already have a spare case, PSU, SSD and gfx card, so I am just missing the cooler (I used the old one in the new build) and some memory to get it running. I just ordered a cooler and will borrow some memory from another PC and then I will do a fresh Win10 install. If the issue is present, then I will bring it to work and try it out there as well, just so I can also rule out mains noise.

Is there any chance this is just a result of the changes in the Windows audio "stack" in 1803? There are a lot of forum posts out there about this. I noticed when I went from 1709 to 1803 if I plug my headphones into the 3.5mm jack on the back of my "monitor" (a 2017 24" 1080p Vizio TV), which is driven by HDMI from a GT710, I get all kinds of pops and drops. But if I plug directly into the onboard (it's a Dell machine with an Athlon II and the 785G chipset), all is well.

Pagey wrote:

Is there any chance this is just a result of the changes in the Windows audio "stack" in 1803?

My old build was still running 1709 since it hadn't received 1803 yet, so it cannot be that. For some reason, it always took upwards of half a year before feature updates would appear on that build.
I have considered whether trying out an older build, like 1607, would help. The appearance of these problems could coinside with a feature update so it would be worth to try out. I don't remember exactly when the problems began, but it was some time last year. It could be the 1703 update.

demolition wrote:

Pagey wrote:

Is there any chance this is just a result of the changes in the Windows audio "stack" in 1803?

My old build was still running 1709 since it hadn't received 1803 yet, so it cannot be that. For some reason, it always took upwards of half a year before feature updates would appear on that build.
I have considered whether trying out an older build, like 1607, would help. The appearance of these problems could coinside with a feature update so it would be worth to try out. I don't remember exactly when the problems began, but it was some time last year. It could be the 1703 update.

With that said, I think the issue goes back as far as 1703 or 1709. For some, it was introduced with 1803, though. I know that I personally didn't have any issues until build 1803.

Finally some progress for me.

With the new motherboard and CPU, I found that the problem was gone with a fresh Win10 1803 installation. I then proceeded to installing things until eventually, the problem reappeared.

I found some forum threads with people having similar issues and they were able to diagnose the issue using LatencyMon: http://www.resplendence.com/latencymon
Indeed, this tool showed that when I had the problem, my kernel latency would spike to 3000-4000 us compared to the usual 120-140us, thus starving the audio driver so it would not be able to fill the audio buffers in due time.
On my new build I found that the high latency was caused by a combination of my BD-drive and AnyDVD, so if I simply disabled AnyDVD for my BD drive, the latency would stay nice and low and the audio would also be perfect. I still have AnyDVD running and enabled on my DVD-RW drive, so somehow it is just an incompatibility issue with my BD drive.

It is so nice to finally have proper audio again, and whenever I need to rip some BD discs, I can just temporarily enable AnyDVD.

Now with the old motherboard and CPU I still had the issue with a fresh Windows installation so perhaps something else caused the problem there. I just need a few components (DDR4 is still expensive), and then I will have a working PC with my old motherboard and CPU. Now that I have found LatencyMon it should be a lot easier to figure out what is causing the problem and then I could build a decent 2nd system (i7-6700k is still a nice CPU). Perhaps my parents could use an upgrade from their Core 2 Duo-based PC.

I always get popping when I listen to a HD YT video (example video: https://www.youtube.com/watch?v=AKrO7lzx-nk&t=2s) in a backgound tab in chrome. No clue why, it does not happen when the tab is the active one.
dgxkrnl.sys
ndis.sys
tcpip.sys

Are listed as the culprits with latencies spiking to 1.4K. Been trying to figure it out for a year now but no luck.

I'm glad you figured it out. I'll admit that I was wrong in my guess. Thanks for getting back to us.

tcpip.sys and my nvidia driver are also the worst on my PC in regards to DPC latency (nowhere near problematic though), but remember to run LatencyMon without any other programs running, otherwise they might create false high numbers.

I think that Chrome sets the priority of background tabs to low when they are in the background and that alone could count for the fact that you have those pops if it is simply not giving it enough CPU/GPU time (don't know if it still uses GPU resources for video decoding when it is in the background). But I guess it is not an uncommon usecase since many people will run YT in a background tab for music listening (I know I do sometimes), so it is probably not a general problem.

I assume that your PC is not some 8 year old rig that can only barely handle HD playback?

demolition wrote:

I think that Chrome sets the priority of background tabs to low when they are in the background and that alone could count for the fact that you have those pops if it is simply not giving it enough CPU/GPU time (don't know if it still uses GPU resources for video decoding when it is in the background). But I guess it is not an uncommon usecase since many people will run YT in a background tab for music listening (I know I do sometimes), so it is probably not a general problem.

I get occasional glitches running Spotify in a background tab too; can't be a video decode issue in this case since there's no video.

demolition wrote:

I assume that your PC is not some 8 year old rig that can only barely handle HD playback?

Well .. not 8 years but she did turn 6 this year.

3570K @ 4.6GHz, 16GB DDR3 @ 1600 (stock), 1070GTX, soundcard is a Titanium HD.

I don't have problems with latency or audio when I play my piano using pianoteq, so I guess the general system is fine, might be a chrome issue.

I changed a setting in Firefox and I have not had any audio clicking or popping since. Under the Performance section in the Options menu, I unclicked the box that says Use Recommended Performance Settings and checked the box to Use Hardware Acceleration When Available.

So far, I have not been able to reproduce the digital noise. And, I'm coming off of a sleep mode, so the sound should be Rice Crispies. Could this have been the issue?

The Swamp wrote:

I changed a setting in Firefox and I have not had any audio clicking or popping since. Under the Performance section in the Options menu, I unclicked the box that says Use Recommended Performance Settings and checked the box to Use Hardware Acceleration When Available.

So far, I have not been able to reproduce the digital noise. And, I'm coming off of a sleep mode, so the sound should be Rice Crispies. Could this have been the issue?

Huh. I wouldn't have thought of that, but it is certainly possible. Especially if the stuttering was only happening in Firefox before.

Interesting tidbit:

I read on Reddit that disabling fast boot (useless with SSD anyway) could resolve the issue.
So far I have background tabbed a 15 + 5 + 8Min clip w/o a single pop, deliberately doing many a tab browsing while listening.

Only thing to keep in mind: if the computer goes to standby, apparently it has the same effect as booting with fast boot, so the pops will return after wake-up.

Will observe further over the next few days.

just brew it! wrote:

The Swamp wrote:

I changed a setting in Firefox and I have not had any audio clicking or popping since. Under the Performance section in the Options menu, I unclicked the box that says Use Recommended Performance Settings and checked the box to Use Hardware Acceleration When Available.

So far, I have not been able to reproduce the digital noise. And, I'm coming off of a sleep mode, so the sound should be Rice Crispies. Could this have been the issue?

Huh. I wouldn't have thought of that, but it is certainly possible. Especially if the stuttering was only happening in Firefox before.

The popping and clicking is still happening in other apps, but no longer seems to be an issue in Firefox. Does Chrome has a similar setting I can try?

I'm not sure what to do about the other apps, unless there is some setting that will allow me to do what I did in FF.

The Swamp wrote:

just brew it! wrote:

The Swamp wrote:

I changed a setting in Firefox and I have not had any audio clicking or popping since. Under the Performance section in the Options menu, I unclicked the box that says Use Recommended Performance Settings and checked the box to Use Hardware Acceleration When Available.

So far, I have not been able to reproduce the digital noise. And, I'm coming off of a sleep mode, so the sound should be Rice Crispies. Could this have been the issue?

Huh. I wouldn't have thought of that, but it is certainly possible. Especially if the stuttering was only happening in Firefox before.

The popping and clicking is still happening in other apps, but no longer seems to be an issue in Firefox. Does Chrome has a similar setting I can try?
I'm not sure what to do about the other apps, unless there is some setting that will allow me to do what I did in FF.

Don't know about dedicated audio programs,but just about all software video players will have a tickbox/setting for the use of hardware acceleration for decoding.
Usually under-tools,settings or preferences........

Disabling fast boot definitely solved the issue for me. No more clicks and pops on background tabs. *yay*

The problem with fast boot is that it basically blocks you from doing a proper reboot of the machine. From what I gather, rebooting with fast boot enabled is about the same as putting your PC to sleep mode and then waking it up again. Many drivers have issues with sleep mode and if they go into an unstable state, then they often need a proper reboot to get them back on track.

I've always had fast boot disabled because for some reason Windows actually boots faster if I disable it.. With an NVMe boot drive, it doesn't take many seconds anyway.

Ifalna wrote:

demolition wrote:

I assume that your PC is not some 8 year old rig that can only barely handle HD playback?

3570K @ 4.6GHz, 16GB DDR3 @ 1600 (stock), 1070GTX, soundcard is a Titanium HD.

Well, a 3570K should by far be capable enough of playing back an HD video stream. I am wondering if the Spectre patch could be your issue? On my older (than yours) machines, I have disabled that patch since they became noticably slower. I think you need a Skylake CPU or newer to avoid taking this major performance hit. Being exploited through Spectre is quite theoretical anyway as it is hard to utilize, so the chance of that happening is probably minute. There are far easier ways to get into people's systems..

demolition wrote:

The problem with fast boot is that it basically blocks you from doing a proper reboot of the machine. From what I gather, rebooting with fast boot enabled is about the same as putting your PC to sleep mode and then waking it up again. Many drivers have issues with sleep mode and if they go into an unstable state, then they often need a proper reboot to get them back on track.

More or less. It restarts the user session but puts the OS kernel into a sort of hibernation state. The OS is only reinitialized for software updates that require it, or if hardware changes are detected.

Most of the time it works, but for persistent driver troubles that disappear after a clean reboot, disabling it is something to try.

demolition wrote:

Mains noise would generally be added to the audio output, just causing missing samples.

Before it hits the DAC, mains noise could cause errors on any communications bus in the audio processing path. This will cause latency during retransmission (if detected) or popping sounds if samples are missed. Most sound cards have very small buffers, so it is not out of the question for audio to be more sensitive than most applications.

Issues with drivers and other seemingly-unrelated software can cause popping. It is really one of the most difficult problems to resolve unless your case is one of the 2-3 easy scenarios.

I had a similar problem with an old Creative card a long time ago, and it was resolved by moving it to a different PCI slot (back when bus mastering was a thing).

TheRazorsEdge wrote:

demolition wrote:

Mains noise would generally be added to the audio output, just causing missing samples.

Before it hits the DAC, mains noise could cause errors on any communications bus in the audio processing path. This will cause latency during retransmission (if detected) or popping sounds if samples are missed. Most sound cards have very small buffers, so it is not out of the question for audio to be more sensitive than most applications.

Issues with drivers and other seemingly-unrelated software can cause popping. It is really one of the most difficult problems to resolve unless your case is one of the 2-3 easy scenarios.

I had a similar problem with an old Creative card a long time ago, and it was resolved by moving it to a different PCI slot (back when bus mastering was a thing).

If enough mains noise is getting through the PSU to cause errors in the digital part of the signal path you're going to be having much worse issues than just audio glitches. Like random system crashes and reboots.

When you talk about fast boot, is that a Windows setting or a setting in the BIOS?

I'd like to try it and see if it helps.

UEFI/BIOS

The Swamp wrote:

When you talk about fast boot, is that a Windows setting or a setting in the BIOS?

I'd like to try it and see if it helps.

Captain Ned wrote:

UEFI/BIOS

I think he's still running Win7; if so, I don't think fast boot is a factor.

I'm still using Win7.

The Swamp wrote:

When you talk about fast boot, is that a Windows setting or a setting in the BIOS?
I'd like to try it and see if it helps.

It's a setting under "power management / what happens when I press the on/off button / show all options" in Windows 10 (old style UI).