Hi there,
I just wanted to give 96kHz audio a try but I cannot really get it working. All I can hear is silence - no amount of cranking up the volume helps. 44.1kHz works perfectly fine. aplay shows the same symptoms as all other audio players do. pw-top shows them trying to play at 96kHz too.
So far I’ve tried increasing the pipewire sample rate using pw-metadata -n settings 0 clock.force-rate 96000 as well as the wireplumber config (I can see it trying to configure audio devices with that new frequency in the logs now) aaand I’m out of ideas.
I’ve got a scarlett solo g3 which can go up to 196kHz.
I just can’t figure out what the problem might be… It was enough for me to edit, as @EOS advised, but only in the home directory. And everything up to 384,000 plays without problems.
My setup (Endevour ARM) can give me 48khz sound perfectly fine. 48khz is the standard for most linux systems nowadays i guess.
If you are a recording musician, i can understand the willingness to use higher sampling rates for the recording process, however, if you are just a listener, and have downloaded 96khz flac files and you think that those files will sound better than 44.1khz cd audio, well, according to the professionals in the business, it is highly unlikely that even the most sensitive human ears on earth can consistently discern and identify two identical recordings, one being 44.1khz and the other being 96khz.
There are so many articles and videos about the subject, you can dive into that rabbit hole if you want to. I am happy with the good old CD Audio standard. I think at least for stereo sound, it is the “ceiling” of our “underwhelming” hearing ability (20hz to 20khz “when you are young”). Most middle aged people on earth can “hardly” hear 17khz. 44.1khz sampling rate can capture up to 22khz. More than enough.
What could make and actually did make music recordings so much better sounding in my opinion are the advancing technologies in the instruments and software used in recording and mastering part of the job. These aspects could be done a lot better now with better microphones and better software. But the “digital signal” is still the same ones and zeros and our flimsy ears can not even hear the whole spectrum that 44.1khz provides.
As I understand it, we are not talking about some kind of “quality” in 96000. We are talking about a technical complexity that does not allow us to use the hardware potential.
P.S. In the vast majority of cases, everything above 48/24 is a game of human weaknesses, such as “more > better”. It is enough to make it a habit to consider hi-res files in terms of a spectrogram…
I’m pretty certain that changes via pw-metadata won’t be persistent.
The ArchWiki has pretty detailed instructions on changing the default sample rate and verifying the currently used ones (Section 5.1.7 to 5.1.8.2):
Furthermore:
Do not forget to restart the pipewire.service and pipewire-pulse.socket user units
(never forget pipewire-pulse.socket if you want your configuration changes
to be applied).
Or restart as an alternative, as @EOS already mentioned.
A real example, “caught” last week. Release 2024, 96/24. Everything circled in red not only has nothing to do with music, but also creates interference for sound and difficulties for equipment…
I don’t think that makes much of a difference if it doesn’t work either way.
I did follow those instructions
Just tried that as well, still no change.
qpwgraph doesn’t show anything useful, but I did just remember coppwr. That just shows me what I already know though. Sample rates are set, but no audio. I did however just notice that 48kHz also works.
Can you please share with us the hardware you use to play these high sampling rates? By the way, when you say over 44.1khz i think you also include 48khz. If you can’t play 48khz, then i agree that the issue is serious. Because most hardware nowadays are capable of playing 48khz. However, only a small number of devices are cabaple of playing more than that sample rate.
It could be a driver issue. Maybe this generic driver doesn’t recognise the full potential of your external sound card. Or maybe it is not the correct driver for it.
I’ve been thinking about this as well but I’ve found exactly zero people with this issue on linux yet, and I’d assume that there would be more people than just me using a scarlett solo g3 on linux.
The Scarlett Solo 3rd Gen does not need a separate driver on Arch Linux, as it is a class-compliant USB audio device that works with the native Linux kernel and audio systems like ALSA, PulseAudio, and PipeWire. You may need to disable Mass Storage Device (MSD) mode by loading the snd_usb_audio kernel module with a specific device_setup parameter to access all features. For advanced control of the interface, you can use the geoffreybennett/alsa-scarlett-gui tool, which provides a user-friendly interface to the ALSA controls.
