pipewire issueshttps://gitlab.freedesktop.org/pipewire/pipewire/-/issues2024-03-16T16:54:33Zhttps://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3063wireplumber 0.4.12 crashes sporadically on Alpine Linux 3.172024-03-16T16:54:33ZEvolving The Futurewireplumber 0.4.12 crashes sporadically on Alpine Linux 3.17I have had a couple of crashes lately on Alpine Linux 3.17.
Unfortunately I don't have a way of reproducing it nor do I have any debug build at the moment.
If it would help if I rebuilt wireplumber with debugging enabled let me know an...I have had a couple of crashes lately on Alpine Linux 3.17.
Unfortunately I don't have a way of reproducing it nor do I have any debug build at the moment.
If it would help if I rebuilt wireplumber with debugging enabled let me know and also a hint on the best flags to use when building...
```
[84241.281378] Bluetooth: hci0: SCO packet for unknown connection handle 513
[84241.404437] wireplumber[21188]: segfault at 50 ip 00007fb504ade03d sp 00007fb505ae76b0 error 4 in libspa-bluez5.so[7fb504acf000+45000]
[84241.404452] Code: 5e 41 5f c3 41 57 41 56 41 55 41 54 55 53 48 89 fb 48 83 ec 38 64 48 8b 04 25 28 00 00 00 48 89 44 24 28 48 8b 87 30 01 00 00 <83> 78 50 02 75 0e 4c 8b af c8 19 00 00 bd f0 00 00 00 eb 0e 0f b7
```
I'm using the following versions on Alpine Linux 3.17 x86_64:
wireplumber-0.4.12
pipewire-0.3.60
bluez-5.66https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3056lineout device missing on login. killing pipewire restores all devices2023-03-02T20:42:29Zlachlanlineout device missing on login. killing pipewire restores all devicesVersion: 0.3.66-1
When i log into my desktop I should have 3 audio output devices but i'm only shown 2
```
lspci | grep -i audio
0b:00.1 Audio device: NVIDIA Corporation GP106 High Definition Audio Controller (rev a1)
0d:00.4 Audio dev...Version: 0.3.66-1
When i log into my desktop I should have 3 audio output devices but i'm only shown 2
```
lspci | grep -i audio
0b:00.1 Audio device: NVIDIA Corporation GP106 High Definition Audio Controller (rev a1)
0d:00.4 Audio device: Advanced Micro Devices, Inc. [AMD] Starship/Matisse HD Audio Controller
```
The Starship/Matisse device has an 'SPDIF' and 'line out' outputs.
To restore this 3rd device i need to kill pipewire after i've logged in and then let it restart.
This is what i see after killing the pipewire bin, the SPDIF is recreated and the line out is added.
![image](/uploads/307eecbab96791b97f4a590bc9b7e214/image.png)
I use the lineout for all audio and for my headset microphone so i don't have audio until killing the pipewire process.
I have pipewire-pulse running but don't need to kill that process.
This also affects the microphone lineout which i assume is the same device as the lineout output.
This has only happened after the transition from pulseaudio to pipewire.
I haven't changed any config to try and rectify this. initially i thought that as a user my desktop was starting audio processes before pipewire was fully started but it seems to never find this line out device until the process is killed while the desktop is running.
```
-- System Information:
Debian Release: bookworm/sid
APT prefers unstable
APT policy: (500, 'unstable'), (500, 'testing'), (1, 'experimental')
Architecture: amd64 (x86_64)
Foreign Architectures: i386
Kernel: Linux 6.1.0-5-amd64 (SMP w/16 CPU threads; PREEMPT)
Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=en_AU.UTF-8, LC_CTYPE=en_AU.UTF-8 (charmap=UTF-8), LANGUAGE=en_AU:en
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled
Versions of packages pipewire depends on:
ii adduser 3.131
ii init-system-helpers 1.65.2
ii libpipewire-0.3-modules 0.3.66-1
ii pipewire-bin 0.3.66-1
pipewire recommends no packages.
pipewire suggests no packages.
-- no debconf information
```https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3051Add DSD over PCM support detect or param for device2023-02-28T09:44:25ZAlexander MiroshnichenkoAdd DSD over PCM support detect or param for deviceSome DAC's USB interfaces does not support native DSD, only DoP. Pipewire can't distinguish a device with/whithout DoP support. Would be great detect HW capabilities or to manually guide pipewire for that, to not convert DSD stream to PCM.Some DAC's USB interfaces does not support native DSD, only DoP. Pipewire can't distinguish a device with/whithout DoP support. Would be great detect HW capabilities or to manually guide pipewire for that, to not convert DSD stream to PCM.https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3049No audio from Kali UNF speaker when connected via USB2023-02-24T21:29:11ZArthur SilvaNo audio from Kali UNF speaker when connected via USB<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version (`pipewire --version`): 0.3.66
- Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`...<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version (`pipewire --version`): 0.3.66
- Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`): EndeavourOS
- Desktop Environment: Cinnamon
- Kernel version (`uname -r`): 6.1.6-arch1-g14-1
## Description of Problem:
I can't get the [Kali UNF](https://www.kaliaudio.com/in-unf) speakers to work with Pipewire. They do appear as outputs but no audio comes out, I tried both options to no avail.
Their USB interface is supposed to be 24/48K and works well when I'm running on Windows 10 and also in Windows 11 (_if Windows11 system volume is 44% or higher_).
![image](/uploads/3b6790b3b050c1abcc7d265e3ca1f758/image.png)
![image](/uploads/c8da0861efa28629b2b3b5e9aac796fc/image.png)
## How Reproducible:
100% of the time
### Steps to Reproduce:
1. Connect speaker
2. Play anything
3. Observe no sound being played.
# Additional Info (as attachments):
[pactl_list_cards.log](/uploads/4c35cfd2698d80b8c2d03b58217f8362/pactl_list_cards.log)
[pactl_list_sinks.log](/uploads/6fa3722a3f098feabe426e86655a479b/pactl_list_sinks.log)
[pactl_info.log](/uploads/1df7e38e391639470155822770e17258/pactl_info.log)
[pw-dump.log](/uploads/49169edf3f764fa00939c8c974ccfde4/pw-dump.log)
[cat_proc_asound_INUNF_stream_0.log](/uploads/3d7ad44aaa6a25adb39b1d1a245c7b05/cat_proc_asound_INUNF_stream_0.log)
[aplay_-l.log](/uploads/a430451bc829bad0142f39e4ab261e02/aplay_-l.log)https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3048Crackling audio on V4L device over USB 3.02023-03-17T22:56:49ZMartmistsCrackling audio on V4L device over USB 3.0<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version: 0.3.66
- Distribution and distribution version: Arch Linux (all packages updated as of 1677070800)
- Deskto...<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version: 0.3.66
- Distribution and distribution version: Arch Linux (all packages updated as of 1677070800)
- Desktop Environment: KDE Plasma
- Kernel version: 6.1.12-arch1-1
## Description of Problem:
When trying to connect a UVC-enabled device over USB 3.0, the audio quality goes down pretty far. USB 2.0 is not an option, as v4l2 is unable to read from the device, although audio works fine.
## How Reproducible:
100%
### Steps to Reproduce:
1. Connect the device
2. Use any software (e.g. qpwgraph) to listen to the device.
### Actual Results:
![GBA Boot Sound, Nintendo Switch](/uploads/57b515602275ee73b3b1e55985d1f461/gba.wav)
This exact same device works fine on Windows using the same USB cable and port.
### Expected Results:
[GBA Boot Sound, GBA](https://www.youtube.com/watch?v=6_ZD3FxMcvQ)
# Additional Info (as attachments):
- `pw-dump > pw-dump.log`: [pw-dump.log](/uploads/d5459eeba1125bc15992591cc38276a7/pw-dump.log)https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3047No sound - no bluetooth2023-02-21T21:57:06ZRedbakanNo sound - no bluetooth When I start up, there is no sound.
I found that the command `systemctl --user stop pipewire.service pipewire.socket` fixes the problem (sound immediatly works). But I have to run it every time I start up. Also, the bluetooth "works", ... When I start up, there is no sound.
I found that the command `systemctl --user stop pipewire.service pipewire.socket` fixes the problem (sound immediatly works). But I have to run it every time I start up. Also, the bluetooth "works", but I can't connect my headphones (they are recognized, paired, but I can't connect - they are displayed as connected, but it doesn't work).
Fedora 37 - kernel 6.1.11 (updated from 6.0.7 where sound worked but not anymore when i tried from the grub older kernel version - same issues)
With the command systemctl status --user pipewire I get :
````
○ pipewire.service - PipeWire Multimedia Service
Loaded: loaded (/usr/lib/systemd/user/pipewire.service; disabled; preset: disabled)
Drop-In: /usr/lib/systemd/user/pipewire.service.d
└─00-uresourced.conf
Active: inactive (dead) since Tue 2023-02-21 21:14:33 CET; 1h 39min ago
Duration: 16.921s
TriggeredBy: ○ pipewire.socket
Process: 2195 ExecStart=/usr/bin/pipewire (code=exited, status=0/SUCCESS)
Main PID: 2195 (code=exited, status=0/SUCCESS)
CPU: 61ms
févr. 21 21:14:16 jl-why.home systemd[2047]: Started pipewire.service - PipeWire Multimedia Service.
févr. 21 21:14:16 jl-why.home pipewire[2195]: mod.rt: Can't find org.freedesktop.portal.Desktop. Is xdg-desktop-portal running?
févr. 21 21:14:16 jl-why.home pipewire[2195]: mod.rt: found session bus but no portal
févr. 21 21:14:33 jl-why.home systemd[2047]: Stopping pipewire.service - PipeWire Multimedia Service...
févr. 21 21:14:33 jl-why.home systemd[2047]: Stopped pipewire.service - PipeWire Multimedia Service.
```
`https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3043RME HDSPe MADI playback not set into Pro Audio Mode2023-02-22T09:56:55ZSchroedingers CatRME HDSPe MADI playback not set into Pro Audio Mode<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version: Compiled with libpipewire 0.3.65
Linked with libpipewire 0.3.65
- Distribution and distribution version: Ma...<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version: Compiled with libpipewire 0.3.65
Linked with libpipewire 0.3.65
- Distribution and distribution version: Manjaro Linux
- Desktop Environment: Gnome 43.3
- Kernel version: 6.1.12-1-MANJARO
## Description of Problem:
I've set input and output of my RME HDSPe MADI into Pro Audio mode via Gnome's system settings dialog, yet the playback part doesn't seem to be in Pro Audio mode.
## How Reproducible:
Happens here all the time.
### Steps to Reproduce:
1. Go to Gnome's System Settings->Audio and set both input and output profiles of the MADI device to Pro Audio
2. Run `pw-cli ls` in a terminal and check the `audio.channel` values
3. The capture channels are named like `AUX01` indicating Pro Audio mode while the playback channels are named `UNK`, which suggests the playback channels are not in Pro Audio mode
### Actual Results:
The playback device isn't in Pro Audio mode.
### Expected Results:
The playback device should be in Pro Audio mode.
# Additional Info (as attachments):
- [pw-cli-ls.log](/uploads/d7fdf8b7db041954751d63bab03e8156/pw-cli-ls.log)
- [pw-dump.log](/uploads/0b6bf8b8153c25b2eb3da97d9fc2ebcc/pw-dump.log)https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3041Cracking Noise with external USB audio card and when in Discord Call (WEBRTC)2023-02-20T10:26:54ZcyberfroggCracking Noise with external USB audio card and when in Discord Call (WEBRTC)<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version (`pipewire --version`):
```
pipewire
Compiled with libpipewire 0.3.65
Linked with libpipewire 0.3.65
```
- ...<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version (`pipewire --version`):
```
pipewire
Compiled with libpipewire 0.3.65
Linked with libpipewire 0.3.65
```
- Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`): `Fedora Linux 37.20230216.0 (Kinoite)`
- Desktop Environment: `KDE Plasma`
- Kernel version (`uname -r`): `6.1.11-200.fc37.x86_64`
## Description of Problem:
Cracking sound when Discord is running. Even if I stress test my CPU it doesnt give any cracking/glitchy/freezes sounds. But when I start discord app (flatpak) I starting to get a lot of cracking sound.
> EDIT:
> I found some conditions for my issue
> - usb external audio - discord call is off - cracking
> - 3.5 jack audio - discord call is off - no cracking most of time, but when I alt-tab makes some tiny noise
> - usb external audio - in discord call - cracking
> - 3.5 jack audio - in discord call cracking
#### My specs:
- Motherboard: GIGABYTE B550 AORUS PRO AX
- CPU: AMD Ryzen 5 3600 OEM
- Headphones: Logitech G PRO X (It has both methods of plug in - 3.5mm jack and external its own USB audio card. I tryed on USB audiocard and 3.5 jack and on record it also audible)
## How Reproducible:
Always
### Steps to Reproduce:
1. Run Fedora Plasma X11
2. Run any youtube music (Tested on orchestra types and remixes musics)
3. Start anything that loads your pc, you should not get any cracking sounds
4. Join discord call (someone should talk)
5. Hear cracking and glitchy sounds
### Actual Results:
Cracking audio, skips, delays, freezes, glitched-like lags
### Expected Results:
No cracking audio, no skips, no delays, no freezes, no glitched-like lags
### My Tries to fix:
# Additional Info (as attachments):
Log dumps:
```bash
PIPEWIRE_LOG_SYSTEMD=false PIPEWIRE_DEBUG=5 pipewire 2>&1 | tee logs-pipewire.txt
```
```bash
PIPEWIRE_LOG_SYSTEMD=false PIPEWIRE_DEBUG=5 pipewire-pulse 2>&1 | tee logs-pipewire-pulse.txt
```
```bash
WIREPLUMBER_DEBUG=2 wireplumber 2>&1 | tee logs-pipewire-pulse.txt
```
- [pw-dump.log](/uploads/3101fe02ba99856aeaf509d4d2a82577/pw-dump.log)
- [logs-pipewire.txt](/uploads/8dfb042629b25d69c6dd95673bd312ef/logs-pipewire.txt)
- [logs-pipewire-pulse.txt](/uploads/8e8c43a6494dd24ac6e5ee82c7b0c69d/logs-pipewire-pulse.txt)
- WIREPLUMBER log was emptyhttps://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3039new-with-userdata may overflow size_t and access unallocated memory2023-02-19T05:04:18ZThomas Weißschuhnew-with-userdata may overflow size_t and access unallocated memoryMany objects in pipewire can be constructed with associated user_data.
For example the `user_data_size` argument to `pw_context_new()`:
With the expression `sizeof(struct impl) + user_data_size` the actual size of memory to allocate is...Many objects in pipewire can be constructed with associated user_data.
For example the `user_data_size` argument to `pw_context_new()`:
With the expression `sizeof(struct impl) + user_data_size` the actual size of memory to allocate is calculated.
However if `user_data_size` is close to `SIZE_MAX` this addition overflows and much less memory is allocated.
This then will lead to out-of-bound accesses in both the user of pipewire and pipewire itself.
These calls should use an internal wrapper around `calloc()` the catches the overflow returns `NULL` in this case.https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3035[RFC] Spatial audio2023-07-21T08:35:02ZDmitry Sharshakov[RFC] Spatial audio!1497 brought up a 3D audio renderer for headphones. This can be a nice starter for a throughout spatial sound implementation in the ecosystem.
Here is the list of tasks I plan to work on for getting modern spatial sound support for sys...!1497 brought up a 3D audio renderer for headphones. This can be a nice starter for a throughout spatial sound implementation in the ecosystem.
Here is the list of tasks I plan to work on for getting modern spatial sound support for systems running PipeWire and WirePlumber. Some of the tasks need assistance, mainly from API consumers such as Wine/Proton/Steam Play, Valve, media player and browser developers.
| Task | Status | Implementer | References |
| ------ | ------ | ------ | ------ |
| Render spatial audio to stereo headsets, math model and processing implementation | Done | @d3dx12.xx | !1497 |
| Ensure SOFA-based plugin is available | TBD | @d3dx12.xx, maybe others | #3009. To load spatializer, WirePlumber must know it has been built, to avoid creating a bad filter chain that outputs nothing |
| Automatically spatialize surround audio streams from applications when conditions are met | Planned | @d3dx12.xx | Pending WP 0.5 work |
| Evaluate and provide out of the box SOFA files and configs for spatial sound support by default | TBD | Distro maintainers | Default `/usr/share/libmysofa/MIT_KEMAR_normal_pinna.sofa` might be sufficient, but needs evaluation. For me personally it feels a bit biased, as it changes the sound frequencies |
| Panner module for mapping a dynamic sound object to a surround system (classic speaker system); AV receiver support | TBD | Anyone willing to do this can get some mentoring | #2310, Needed to support sound objects, because we mustn't just silence them if the user isn't wearing headphones. This should be a fallback positioning sounds in a stereo or surround space |
| Sound object creation by applications (API and design considerations) | Active | Comments from people working on games or related to Wine are welcome to comment on the way applications transmit position metadata to PipeWire, how do they create extra channels. This is likely to require a PipeWire-native driver, so @davidebeatrici's comments will be valuable. Valve is also an important user of PipeWire and comments from Steam Deck audio developers are more than welcome | #2310, https://github.com/ValveSoftware/Proton/issues/6138#issuecomment-1435493791 https://gitlab.winehq.org/wine/wine/-/blob/master/dlls/mmdevapi/spatialaudio.c, #1938. *Note: this does not include Doppler, Reverb and other FX. That is audio library task that is not tied to hardware nor user preferences* |
| Dynamic audio source support in libpipewire | Blocked by the previous point | Not yet determined | A client API implementation is necessary for applications willing to use 3D objects alongside static channels |
| Suggestions | Always welcome | Anyone | The better we shape the API for the future, the better would maintenance be |
Some API ideas for dynamic sources:
1. Create sound sources on demand. Has an issue of latency, because WP Lua engine isn't realtime, and filter chain creation takes time.
2. Application should report desired number of channels during the client creation. It supplies a number of static well-known channels, such as FL, FR, RR etc, which are either directly mapped to corresponding channels or reproduced with prebuilt spatializer config, alongside *dynamic channels*. Dynamic channels are declared when the client creates its output and just play nothing until needed. When needed, client repositions them (how?) and plays audio. If an extra batch of those is needed, client has to create a new connection.
Generic:
How do we pass positions to the PW nodes? Some variants: use some buffer metadata, use a control port connected between client and filter chain or a reference to the chain (client sets positions just as `spin(2)` examples in !1497 do).https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3034Support for v4l2 cameras with uvc extension unit controls similar to gstreame...2023-02-17T19:59:03ZJames HilliardSupport for v4l2 cameras with uvc extension unit controls similar to gstreamer uvch264src v4l2src wrapper in SPA plugin.The current v4l2 SPA plugin does not correctly handle video output formats from cameras that use extension unit controls.
For example a number of Logitech cameras such as the C930e require special handling and extension unit control sup...The current v4l2 SPA plugin does not correctly handle video output formats from cameras that use extension unit controls.
For example a number of Logitech cameras such as the C930e require special handling and extension unit control support for accessing their h264 video output.
In gstreamer this is implemented via a [v4l2src wrapper](https://gitlab.freedesktop.org/gstreamer/gstreamer/-/tree/main/subprojects/gst-plugins-bad/sys/uvch264).
It would be good if the pipewire v4l2 SPA plugin could support these extension unit controls/handling so h264 video and other camera formats requiring extension unit controls are usable with pipewire.https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3033Default Sample rate on Microphone Input - Difference between Pulseaudio and P...2023-12-11T16:25:37ZLozramADefault Sample rate on Microphone Input - Difference between Pulseaudio and Pipewire Pulse Default InstallHello,
Dont know if this is really a issue but a difference between pulseaudio and pipewire pulse default install.
I am opening a microphone stream in python with pyaudio ( 1 channel, 16k sample rate ).
With default alsa configuration...Hello,
Dont know if this is really a issue but a difference between pulseaudio and pipewire pulse default install.
I am opening a microphone stream in python with pyaudio ( 1 channel, 16k sample rate ).
With default alsa configuration on Raspberry OS ( Debian ) i need to specify device by index and change to 48k sample rate and resample myself
else i get an Invalid sample rate error.
After installing pulseaudio 16k sample rate worked default in pyaudio without setting a device by index. I did not need to resample anymore in my Code. The resampling of pulseaudio was better than my own ( with scipy signal.resample ) for the error rate of my Speech to Text Engine.
After removing Pulseaudio and installing pipewire all works great on raspberry os ( no stutters anymore, stable, no buffer problems.. great work )
but i need to set again device by index and sample rate to 48k in pyaudio.open else i get the error:
File "/home/kitt/.local/lib/python3.9/site-packages/pyaudio/__init__.py", line 441, in __init__
self._stream = pa.open(**arguments)
OSError: [Errno -9997] Invalid sample rate
I am new to linux, don't know if pulseaudio creates a default microphone device with 16k sample rate or something but is this default "feature" of pulseaudio somehow possible in pipewire too and perhaps missing and perhaps desired on default install ?
Tried setting clock.rate clock.allowed-rates clock.force-rate to 16k but didn't help in pyaudios mic Input and changes also audio output lo lower rates if I am correct.
Thankshttps://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3029Pipewire fails to set saved Volume for Devices with Analog & IEC output & more2024-03-11T19:05:59ZWaffshappenPipewire fails to set saved Volume for Devices with Analog & IEC output & more- PipeWire version (`pipewire --version`): pipewire Compiled with libpipewire 0.3.65 Linked with libpipewire 0.3.65
- Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`): Fedora Linux 37 (Thirty Seven)
- Desktop...- PipeWire version (`pipewire --version`): pipewire Compiled with libpipewire 0.3.65 Linked with libpipewire 0.3.65
- Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`): Fedora Linux 37 (Thirty Seven)
- Desktop Environment: Gnome (and KDE on another install)
- Kernel version (`uname -r`): 6.1.10-200.fc37.x86_64
## Description of Problem:
This Bug report consists of multiple Problems i have with multiple of my Devices that all stem from the same Source as i guess.
My Devices show up as two devices or with multiple Profiles in all applicable Sound Management Tools.
For Example, "Rode NT-USB Mini" here, or "Sound Blaster PLAY! 3" or "Fiio XXXXX" and many others i've encountered.
![multi_profile_or_device](/uploads/a0e579bccad88ecbcbb23af5f5ddbcec/multi_profile_or_device.png)
The following Behaviour is observed:
The "Analog" Profile/Device sets the __maximum__ output Volume.
The "Digital" Profile/Device seems to be at a higher quality and operates __within__ that maximum Volume.
Visualized as Example:
```
■■■▨▨▨▥□□□□□□ 100% Possible Device Volume (Analog)
■■■▨▨▨▥ "Analog" Current Volume (50%)
■■■▥□□□ Actual Volume Range in Digital Profile (only within Analog)
■■■▥ Current Volume Digital Profile (50% = 25% of possible max volume)
```
This makes it extremely annoying to switch to the higher quality (see pw-dump samplerate) digital profile and keep adjusting volumes in this range.
Since the Amplification on the Device is based on the analog profile all Amp Noise will also be at the Analog volume level, not the adjusted Digital one.
Additionally since ~1-2 Months ago Pipewire stopped saving the Volume on the Analog Profile.
I usually set that at 100%, default to Digital and then adjust from there. I kept raising the Volume
until i realized that after a Reboot its now set much lower every time. (Or on user switch).
I took a pw-dump from [before_adjusting.json](/uploads/6e0c4c161b27eb2b15c1c2198de3a41d/before_adjusting.json) and [after_adjusting.json](/uploads/b541bc94cb711f7ae423f17682e98845/after_adjusting.json) the "Analog" profile from 100 to 90% and back and then switching back to Digital.
Analog was saved at 100% before this reboot.
Why dont i just use the Analog Interface then all the Time?
Outside of the listed higher Samplerate quite simply because of another issue:
If i set the Volume of the analog profile smaller than 18% it cuts off audio output entirely in Hardware (Sound Blaster PLAY! 2 / 3) - yet 18% is too loud in many Applications for me on my Headphones.
### So, as a condensed a List of observed bugs that all have to do with this behaviour in my Observation:
1. Pipewire does not restore the Volume to the Analog Profile after a Reboot / User switch correctly
2. Pipewire does not shut down the Device when no Audio plays and it is in "Digital" Profile (Both RODE NT-USB Mini, Fiio E10 and my Sound Blaster PLAY! 2 and 3 exhibit this). It suspends correctly with "Analog"
3. Pipewire does not default to the (higher Quality?) "Digital" Profile for me
4. And finally, in my personal View - Pipewire listing these at all. It feels like this should be abstracted away by Pipewire, automated behind the Scenes. Especially the Volume handling. I have seen tons of USB devices that do exactly this and i have to jump this hoop everytime. I imagine finding this out and managing it could be a serious Hurdle for lots of Users.
## How Reproducible:
Alwayshttps://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3027Audio Cracking on voice chat (specially on Screenshare) on Discord2023-03-10T22:54:40Zapache aluminumAudio Cracking on voice chat (specially on Screenshare) on Discord<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
### PipeWire version (`pipewire --version`):
- Compiled with libpipewire 0.3.65
- Linked with libpipewire 0.3.65
### Distrib...<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
### PipeWire version (`pipewire --version`):
- Compiled with libpipewire 0.3.65
- Linked with libpipewire 0.3.65
### Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`):
- Fedora Linux 37 (KDE Plasma)
### Desktop Environment:
- KDE Plasma Version: 5.26.5
- KDE Frameworks Version: 5.102.0
- Qt Version: 5.15.8
### Kernel version (`uname -r`):
- 6.1.10-200.fc37.x86_64
## More Hardware Info
- Operating System: Fedora Linux 37
- Graphics Platform: Wayland
- Processors: 4 × Intel® Core™ i3-5020U CPU @ 2.20GHz
- Memory: 7.2 GiB of RAM
- Graphics Processor: Mesa Intel® HD Graphics 5500
- Manufacturer: ASUSTeK COMPUTER INC.
- Product Name: X555LAB
- System Version: 1.0
## Description of Problem:
When I'm trying to draw on Krita while I hear a voice chat or watch a stream on discord the audio cracks and depending the way I use Discord it can be from a basic audio crackling to a completely unhearable noise at the point I can't understand what anyone says.
## How Reproducible:
It always happen
### Steps to Reproduce:
1. Open Discord and Krita
2. Join a voice chat and create a new project on Krita. You can get the same effect with a music bot.
3. Draw or move the stylus over the canvas.
### Actual Results:
The audio cracks and stutters when I draw while I'm on the voice chat.
### Expected Results:
Being able to hear a voice chat without any voice.
# Additional Info (as attachments):
I tried this configuration: https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/1773 but it did not work. Actually the official client of Discord lose the audio when I draw after I setup that configuration on .config/pipewire/pipewire.config/
I have a graphic display so technically I use a double monitor setting.
This is the result on pw-top:
![pw-top-discord](/uploads/3069b8148e769cd208cc93b6838ac96e/pw-top-discord.png)
Another result on tw-top:
![Screenshot_20230203_220415](/uploads/fd42828af481b1eb75f39d4d4f9ffa55/Screenshot_20230203_220415.png)
- `pw-dump > pw-dump.log`:
(This one is based on the custom configs based on the other issue, but generally speaking on the default there is cracking too)
[pw-dump.txt](/uploads/331d47f9ab86327f1db318b77939b393/pw-dump.txt)
This is a recording I did to show the issue. What I did is I open a discord music bot on a voice chat an I started drawing on krita. The sound starts to break the moment I put the pencil on the canvas.
![2023-02-11_00-16-15](/uploads/1e481e2ae6d545102c7f0b99272605ed/2023-02-11_00-16-15.mp4)https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3024Degraded microphone output when gaming2023-12-06T01:29:18ZJakub NarolewskiDegraded microphone output when gaming- PipeWire version (`pipewire --version`): Compiled with libpipewire 0.3.65 / Linked with libpipewire 0.3.65
- Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`): openSUSE Tumbleweed
- Desktop Environment: KDE
-...- PipeWire version (`pipewire --version`): Compiled with libpipewire 0.3.65 / Linked with libpipewire 0.3.65
- Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`): openSUSE Tumbleweed
- Desktop Environment: KDE
- Kernel version (`uname -r`): 6.1.10-1-default
## Description of Problem:
For some time - about two'ish months - I have an annoying problem with a degraded microphone output in some situations - mostly when playing games. At first I thought that this was an issue with Discord only as its quality is not so great on Linux but recording my mic during that time with Audacity yields the same results.
Reproduction scenario itself is a bit ... unstable but it generally goes like this. I meet up with my friends on Discord and we start playing together. After some time - sometimes almost immediately, sometimes after an hour - friends tell me that my voice quality goes significantly worse and I'm sounding robot'ish. This effect worsens over time to the point that I'm barely audible.
This do not affect audio quality on my side as games or system sounds are fine. The issue tends to go away when I close the game I'm currently playing but this is not a rule as sometimes the effect itself seems to linger up to the point of pipewire restart.
Now, this does not seem to be connected to the specific game I'm playing nor is affecting only Discord as using qpwgraph and routing my mic output directly into my headphones setup I can hear the issue. To be clear about what I'm experiencing I'm attaching a sound sample recorded with Audacity with both bad and good variants of the output and also a pw-dump output from during the observed issue.
Pretty weird stuff as my hardware configuration have not changed and I'm just keeping my openSUSE installation up-to-date.
I feel like this is somewhat connected to the load on my system but simulating different scenarios with stress-ng I was not able to reproduce this.
I have tried to tweak pipewire and wireplumber configurations in some ways but I was not able to come with something that remedies this as all my changes are more or less trial-and-mostly-errors approach :P To be specific I have tried:
- forcing different quantum values in pipewire.conf (default.clock.quantum = 1024 and default.clock.min-quantum = 512 and some variants of them)
- enabling mem.allow-mlock = true in pipewire.conf and "memlock unlimited" / "rtprio 95" in my limits.conf
- setting ["api.alsa.period-size"] = 1024 and ["api.alsa.headroom"] = 2048 in wireplumber/main.lua.d/50-alsa-config.lua
- switching between Pro Audio and "not pro audio" card profiles which only leads to extensive, system wide crackling with pro audio similar to the situation described in: https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/2992
I think I can somewhat rule out the hw of USB interface itself as I do not experience this on a Windows system.
## How Reproducible:
Constantly.
### Steps to Reproduce:
Launch a game and monitor microphone quality.
### Actual Results:
Microphone output quality is degraded.
### Expected Results:
Microphone output quality is fine and stable.
# Additional Info (as attachments):
![good](/uploads/c3d503390473179592f6828e58cba991/good.wav)
![bad](/uploads/e24917bb0d4e91a05ace423b939dad34/bad.wav)
![pw-top.txt](/uploads/18fabaaa5fd666dc1087d739d21768e9/pw-top.txt)
[specs.txt](/uploads/dfa4302b65b8fc527d3382cce0e65f9a/specs.txt)https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3022RFC: Advertise rtp-source2023-02-10T13:48:59ZDmitry SharshakovRFC: Advertise rtp-sourceIt would be nice to have another computer being an RTP speaker. You set its rtp-source up properly, it advertises as a unicast receiver, and then your main computer sees that as a sink (like RAOP, for example).
It's an RFC to know what ...It would be nice to have another computer being an RTP speaker. You set its rtp-source up properly, it advertises as a unicast receiver, and then your main computer sees that as a sink (like RAOP, for example).
It's an RFC to know what format better fits this advertisement. Most rtp devices are only advertising what they send, not what they can accept, so you always have to accept a connection from the receiver side.https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3017Put MIDI ports on nodes (not MIDI bridge)2023-04-06T14:35:30ZDmitry SharshakovPut MIDI ports on nodes (not MIDI bridge)It'd be nicer to have MIDI grouped by device or process ID and organised into new nodes, not added to MIDI bridge. Unsure whether that change doesn't break any clients, so requesting comments.It'd be nicer to have MIDI grouped by device or process ID and organised into new nodes, not added to MIDI bridge. Unsure whether that change doesn't break any clients, so requesting comments.https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3014After every boot I have to switch output ports to Speakers then back to Headp...2023-03-15T09:36:59ZcgAfter every boot I have to switch output ports to Speakers then back to Headphones to listen to anythingThis happens after reversing stereo: https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3005
After every boot I have to switch output ports to Speakers then back to Headphones to listen to anything -see video.
https://video.h...This happens after reversing stereo: https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3005
After every boot I have to switch output ports to Speakers then back to Headphones to listen to anything -see video.
https://video.hardlimit.com/w/epftXwZBbtoQkSmQp4sfT4
I only have to do it once, so, are there any pactl or wpctl or other commands I could use in my autostart to switch to Speakers, then Headphones?https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3012TASCAM US122L not supported.2023-09-08T13:57:01ZstephematicianTASCAM US122L not supported.<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version: 0.3.63
- Distribution and distribution version: Ubuntu 22.04.1 LTS
- Desktop Environment: MATE
- Kernel ver...<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version: 0.3.63
- Distribution and distribution version: Ubuntu 22.04.1 LTS
- Desktop Environment: MATE
- Kernel version: 6.1.0-1004-oem
## Description of Problem:
I plug in the TASCAM US122L (an older USB audio interface), however pipewire does not identify it as a recording device.
## How Reproducible:
__I discovered that a custom `~/.asoundrc` is not needed - see https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3012#note_1755398__
To set this device up for recording with `arecord` I had to perform two steps. First, get the card number via `cat /proc/asound/cards`
```
0 [NVidia ]: HDA-Intel - HDA NVidia
HDA NVidia at 0xfc080000 irq 95
1 [US122L ]: USB US-122L - TASCAM US-122L
TASCAM US-122L (644:800e if 0 at 005/002)
2 [Generic ]: HDA-Intel - HD-Audio Generic
```
And then replace CARDNR with the value 1 in the following `~/.asoundrc` file:
```
pcm.!usb_stream {
@args [ CARD ]
@args.CARD {
type string
default "CARDNR"
}
type usb_stream
card $CARD
}
ctl.!usb_stream {
@args [ CARD ]
@args.CARD {
type string
default "CARDNR"
}
type hw
card $CARD
}
```
These instructions are as per <https://www.alsa-project.org/main/index.php/Matrix:Module-usb-us122l>
To record a simple test, I can use:
```
arecord --period-size=640 -c2 -fS24_3LE -r44100 -MDusb_stream:CARD=US122L > test.wav
```
![test](/uploads/14f346870ef76f284ec9af02088f77af/test.wav)
### Steps to Reproduce:
1. Use same `~/.asoundrc` as in the past __Actually not needed__
2. Plug in device
3. Look at `qpwgraph`
### Actual Results:
The view from `qpwgraph`:
![image](/uploads/a57d2c3cd4a2fbc4b1670f533802c5c5/image.png)
### Expected Results:
To see the device listed as an input in `qpwgraph`.
# Additional Info (as attachments):
- `pw-dump > pw-dump.json`: [pw_dump.json](/uploads/21a5333bd9dab8bf68c18f507dbe1390/pw_dump.json)https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/3011No sound in Unity3D-based games due to missing pulseaudio binary2023-02-06T10:08:32ZDominik MierzejewskiNo sound in Unity3D-based games due to missing pulseaudio binary<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version (`pipewire --version`): 0.3.65
- Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`...<!-- If you are filing this issue with a regular release please try master as it might already be fixed. -->
- PipeWire version (`pipewire --version`): 0.3.65
- Distribution and distribution version (`PRETTY_NAME` from `/etc/os-release`): Fedora Linux 37 (MATE-Compiz)
- Desktop Environment: MATE
- Kernel version (`uname -r`): 6.1.8-200.fc37.x86_64
## Description of Problem:
Games using the Unity3D engine fail to produce any sound. The following message can be seen in `~/.config/unity3d/<vendor>/<gamename>/Player.log`:
```
AudioManager: Using Unknown: NoSound Driver
```
## How Reproducible:
Always.
### Steps to Reproduce:
1. Run any Unity3D-based game (tested with: Tetrobot and Co, AER, A Short Hike, The First Tree, The ABC Murders, PAN-PAN)
### Actual Results:
No sound.
### Expected Results:
Sound plays normally.
# Additional Info (as attachments):
`strace` shows that the games try to run `pulseaudio --check` which is not provided even by pipewire-pulse.
`pipewire-pulse` doesn't support the `--check` option, but if it did, one could simply add a symlink:
```
ln -s pipewire-pulse /usr/bin/pulseaudio
```
as a work-around. Adding a symlink to `/bin/true` works well as a work-around. The check succeeds and the games initialize audio using the pulseaudio backend of pipewire.