Just plugged in headphones for the first time after getting my batch 3 laptop, lots of static popping only when audio is playing – mostly out of the left channel, it’s hard to tell. I’m using the standard microsoft drivers currently for the realtek controller to try and sidestep the sleep drain problems from another thread.
For what it’s worth, I’m running without power plugged in, 48000hz 16bit, Windows 10.
Changing to 24bit and/or disabling signal enhancements doesn’t have an effect.
Oh, and the headphones I’m using have a TRRS connector with an inline mic that isn’t picked up by the laptop (pretty sure that’s a driver thing, but might be relevant).
Static also seems to occur pretty consistently given repeated audio, while not exactly the same, it’s in about the same place timing-wise every time.
The discord sound effect seems to cause it every single time.
Just an observation, may or may not be relevant to the existing audio side crackle.
I have been plugged into the jack for an audio line out to listen to radio stations in my area.
If my amplified speakers are plugged in and on before the laptop is, I get a Lot of static noise from the speakers. As the laptop powers on, the sound then shifts to different static noise… And finally to quiet. Like power is applied by the EC, and then the driver kicks in.
SDR is not plugged in at the time I boot btw.
I have to wonder if it is an actual intermittent power problem or the headset jack driver going back to an open circuit condition… Or switching modes / levels when it should not.
Just chiming in that I’m also experiencing this with my laptop, which is running Windows 10 Pro. If I switch the same headphones over to a jack on a connected Thunderbolt dock the crackling goes away, so it’s not just my headphones.
I wanted to mention that I upgraded my Framework to Windows 11 and the crackle and static noise on the headphone jack is gone. So this is certainly a software issue, which is great.
But to repeat the noise and static are gone under Windows 11. To be fair under Windows 10 they were there, but they didn’t really bother me, because it only could be heard occasionally during times when it was basically quiet.
I seem to be able to reliably reproduce this and eliminate it (albeit via less-than-ideal means). Best I can tell it is a result of the CPU being in a low-power state. Here is my debugging and cause isolation log:
I am running NixOS with Kernel 5.14.15 and am using TLP for power management and Pipewire for audio. No peripherals were attached during this test except the expansion cards (2 USB-C + 2 USB-A). Tests performed while connected to WiFi 802.11ac network.
I notice this both with speakers and headphones. It is somewhat more pronounced with headphones.
This audio is my go-to for reproducing, however it is important to, via some means, play the audio only, not the video, in order to minimize CPU usage (I play with Spotify); we’ll see why later.
Rebooting has no effect, neither does restarting Pipewire.
Switching to charger removes crackle immediately, switching to battery nearly immediately causes crackle. However, this is perhaps a red-herring, see next bullet.
Exiting powersave config with tlp ac, even when on battery, immediately removes the issue. Entering powersave config with tlp bat, even when on charger, nearly immediately causes crackle.
Setting TLP config SOUND_POWER_SAVE_ON_BAT=0 and RUNTIME_PM_DRIVER_DENYLIST=snd_hda_intel had no effect.
When in powersave config, stressing the CPU (e.g. with stress-ng --cpu 1) immediately removes the crackle; stopping the stress nearly immediately restores the crackle.
I did some additional debugging with PowerTOP tunables, and have isolated the static and crackle I experience to occurring when power management is enabled for “PCI Device Intel Corporation Tiger Lake-LP Management Engine Interface”. This device has the driver mei_me, and is in TLP’s default RUNTIME_PM_DRIVER_DENYLIST, however, it seems that that setting doesn’t matter if the Kernel sets the power control for the device to auto.
To resolve, I added a custom udev rule to force disable PCIe runtime power management for that device (using this doc as a reference; I don’t use Arch, but its wiki is great):
Alternatively, one could set RUNTIME_PM_DISABLE in TLP, but that requires specifying the PCIe device path rather than ID, and the latter is typically more stable (or at least portable).
After a reboot to load the new udev rules, the ME device is successfully excluded from power management, and so is always on even on battery, and, albeit just with some short testing, I no longer have any crackle.
Note of course that it’s possible that the cause of my crackle is different than others, as mine seemed more reliable to reproduce and not tied to a channel as it was for others. But hopefully this helps someone.
I have persistent subtle crackling too, and unfortunately this didn’t quite do it. But I have a feeling it is still power management related. I’m using power-profiles-daemon and the crackling is much less noticeable when running in a higher power state (Balanced/Performance) than when in Power Saver mode.
I am experiencing a similar issue on my laptop running Ubuntu 21.10. I have the newer IDT 92HD95 chip. There are two issues:
Whenever the sound card is started, it produces a “click” sound. I’ve had this issue before on other laptops, and the solution has always been to disable snd_hda_intel.power_save so that the card is only started once on boot.
More importantly, there is a constant, quiet white noise played from the headphone jack at all times when the card is active.
A few things I’ve tried:
Upgrading kernels. Both 5.13 and 5.15.4 have this issue.
Fedora 35, which uses Pipewire instead of PulseAudio, also has this issue.
Setting snd_hda_intel.power_save=0 as a kernel param. In some ways, this makes it worse, since the white noise is only produced when the card is active, and this makes sure that the card is active all the time.
Different headphones. I’ve tried 2 different sets of headphones, with no difference.
Power profiles, peripherals, or whether I’m on battery or AC make no difference.
Whether something is actually playing makes no difference. It’s easier to hear the white noise without something playing, but I can still hear it while listening to music.
Messing around with hdajackretask didn’t seem to do anything, although I don’t know what I’m doing.
TLP stuff doesn’t apply, I don’t have TLP running right now.
I’m kinda at a loss for how to proceed from here. Is this potentially a hardware issue?
I’ve not much to add, either, some sound issue on November batch DIY i5, Fedora 35 KDE (on both Wayland and Xorg). Kind of similar to what @ImaxinarDM mentioned.
Whenever I’m going from silence to audio playing (even system sounds or notifications), there’s a sharp beep or crackle. Only happens if something is plugged in the audio jack. Laptop speakers themselves and anything attached via HDMI don’t produce this sound.
@popcornbushwhacker That problem has a mitigation. You can try running echo 0 | sudo tee /sys/module/snd_hda_intel/parameters/power_save
and it should stop clicking every time. If it works, you can set snd_hda_intel.power_save=0 as a kernel boot param to make it permanent.
I am on popos 21.04 and encountering very similar or the same issue when using the 3.5mm audio.
Static/crackling in the left channel, primarily when audio contains voices. Especially so if the voices are in the right channel. Opening the sound settings and testing the right channel (which plays a voice saying ‘front right’) causes a very noticeable amount of static in the left channel.
I’ve tried some of the power saving mode suggestions mentioned in this thread and so far no luck. It’s pretty consistent except with volume. Lowering my volume to about 25% or lower it stops occurring or may just be too low for me to notice.
I tried the udev rule, setting performance to high, and using tcp. Nothing works so far. Given so many people have a similar issue I don’t think it’s just a loose plug but maybe I’ll open up the laptop and try and reseat the module? Out of ideas otherwise.