I haven’t had this problem til now, although I’ve always found it extremely difficult to keep track of changes made ot my audio settings when I use different devices (eg, go from webcam to system default).
To record off system audio (eg, a running application) I configured audacity to use ALSA and hw:0,0, which is analog output from a Realtek on board sound card. This has worked for years, but now suddenly either no or an extremely minute audio signal shows up in the recording. The weird thing is that audio does appear on the Recording Level monitor.
As I said, I find ALSA, Pulse, etc extremely complex and difficult to configure. Also, I don’t actually think this is an Audacity issue, ans I get a similar problem using arecord directly.
I’d be grateful for any advice and tips how to debug this whole setup.
Ubuntu 18.04.6 LTS
Program build date:
Oct 5 2022
969e36 of Wed Oct 5 13:43:02 2022 +0300
CMake Release build (debug level 1), 64 bits
GCC 7.5.0 (Audacity)
(Cross-platform GUI library)
(Audio playback and recording)
(Sample rate conversion)
File Format Support
(Ogg Vorbis Import and Export)
(ID3 tag support)
(FLAC import and export)
(Import via QuickTime)
(Import via GStreamer)
(Sound card mixer support)
(Pitch and Tempo Change support)
(Extreme Pitch and Tempo Change support)
(There are still many old “tutorials” on the Internet with instructions for hacking “.asoundrc” for loopback recording, but with modern systems there’s a high risk that they will mess up the sound system. My advice would be to not touch .asoundrc unless you are an absolute expert).
I assumed the ‘Sound’ control under ‘Settings’ would use pavucontrol, but apparently not…
My (about to be updated) system has some differences from the tutorial. For example, no ‘capture from’ dropdown appears on the recording tab in pavucontrol.
The real source of confusion for me has been that pavucontrol has a level control on both the ‘Recording’ tab and the ‘Input Devices’ tab, and both of these need to be turned up to increase recording level, but the level control on the ‘Recording’ tab is only available when audacity is actually recording.
Another mystery is that when I finally identified these two controls, I found they’d both been turned right down. It’s possible I did this myself while trying to solve the recording issue, but it’s far more likely some other software messed with them, because this was the problem I was trying to solve in the first place.
I’d really like some guidance on how other applications might go about interfering with these levels, and what else might have caused them to change. I can’t think of anything I’ve done lately that involved the sound system, except maybe playing around with spd-say in a script (which I notice shows up as three separate instances of speech-dispatcher in the ‘Applications’ tab of the ‘Settings|Audio’ control, although the daemon only appears once in pstree).
What config files are modified by pavucontrol and audio settings? I’d like to inspect changes and backup my working settings.
could you tell me the difference between ‘sysdefault’ and ‘default’ devices?
I have an nVidia card with HDMI audio, but I haven’t been using it for audio because I couldn’t get it set up for loopback recording. Do you think it would be a better option? My understanding is that the recording device can use the digital output directly, rather than using analogue, but I’m not sure how this interacts with DRM services…
I was confused by the ‘monitor’ showing a normal signal, even though the recording signal seemed to be more or less absent. I think of a ‘monitor’ as a way of ‘monitoring’ the signal available for recording, but should I be thinking of it as a source device?
Well behaved applications don’t mess with the global sound settings, just their own “per-app” settings. The Desktop’s volume control app usually control PulseAudio’s main “Input” and “Output” levels, and may also control in/out settings for each app. Different Desktops use different volume control apps. I’m using Xfce desktop, which uses its own PulseAudio xfce panel widget. Gnome / KDE / Mate / … will use different widgets to control PulseAudio.
Usually they are the same, just two different names. There’s probably some technical difference but in practice they both select the system’s default device.
I’ve never had much success using nVidia HDMI audio.
Think of it as a recording input device that is permanently attached to, and takes its input from an output device.