j_e_f_f_g wrote: ↑Sun Mar 21, 2021 1:03 pm
If you're a musician who has done any pro music production on Windows/Apple, and you're now trying out Linux, and you just can't understand why you're having so much trouble getting even the most simple use of sound working on Linux... It's not you. It's JACK.
Woah! Seems I triggered some discussion with my statement.
Well, I'm not a "musician" at all. I do play the Saxophone for a few years now on amateur level in a wind band. Not the lead one. I don't do "music production" either.
But I have some basic needs for my everyday Computer use that does involve audio and that turned out to be more complicated than thought. I still have to translate everything here into "Windows Language" and I'm quite certain that I get it wrong more often than not.
Some time ago it all was very easy. I had a PC with an embedded sound card. A headphone jack on the front and a line-out on the back. The latter got a PC speaker system connected with a desktop dial, so that I can turn the speaker on and off and set the volume, parallel to the headphones. Easy cheesy, even with Windows - once you get the driver set so that it don't silences the back audio when a headphone is connected on the front.
Later I added an external USB audio interface (UR242) where I can connect a good microphone to make recordings of my exercises. The interface allows for an external audio source (iPad with AnyTune). The OUT however is wired to a Stereo with good speakers, but normally I will use the PC sound system still
wired to the internal sound card. And the headphones are connected to the external interfaces for the obvious reason of zero latency monitoring. And since I don't want to wire it all anew for every different application, I'd prefer to keep it that way (or another that works equally well). And at last I have a Piano that has it's own MIDI but with an USB cable, and not wired to the external audio MIDI DIN sockets.
In Windows I quickly found about VoiceMeeter that can easily take all audio inputs, no matter what interface (and the webcam too) and patch it to whatever output I like. And it keeps it there. So there is virtually ZERO hassle when I listen to something and get a "hey, lend me your ear" moment where I turn up the volume in the PC audio to play it loud. The audio is streamed synchronously to both interfaces in the background.
The same when I want the audio from the iPad playing on the PC speaker (into the USB interface and out to the internal sound card). Or if I want to have the Webcam Micro out to the conference, but the background music just into my Ears.
As you see, there is no DAW involved at all. Not normally. I do use one from time to time, but mostly I have just "daily need". Video conferencing, monitoring my Sax lessons together with the playalong, playing music from the network library or the external player. At times in the headphones, at times in the speaker, at times both.
That's where I was told to talk to "Jack" when switching to Linux. I do not know better, but it seems to be the only way to get more than one audio interface working together in a way that is as comfortable as it was in Windows. Or maybe even at all.
"Standard Linux" seems to work similar to "Standard Windows", i.e, choose ONE interface and finish, so I can use either the USB audio OR the internal one. As I have learned, "Jack" and Family should be the thing to use, to get all options back again.
Meanwhile I did cut off the Stereo to free the Line-Out from the USB Audio to connect it to my PC-speaker, as I didn't manage to get it done right.
But it still leaves me with the problem that the sound is crackling every other time I wake up my PC from sleep. "restart jack" does rectify that. But I cannot nail it down. It is NOT X-Runs. At least the Counter in "Ubuntu Studio Controls" does not show the crackles, although it still counts upward slowly over time, even with 512x3 buffers.
And then there ARE times when I fire up Reaper (that's what I am used to) to record my session where it is totally OK to just use ONE interface but preferable with low latency. ASIO as we used to say on Windows. A Ryzen 7 3800XT, 32GB RAM plus SSD should be plenty enough power to record a simple mono audio stream at 48kHz without glitches. Shouldn't it? Well, I guess it should.
If this all can be done with "standard Linux audio" (Pulse? ALSA?), I forget about Jak et al.
Edit:
I just saw that my
opening posts attracted some really promising suggestions getting this rectified, so I will try to keep my problems there instead spilling them all over the board.