;tldr Beginning to use a new OS, even using a distro as friendly as Mint, is harder than the overall community says it is. The second there is a problem expect hours of consuming, likely outdated, information. That said Iām happy I switched.
Iām not a programmer. If you are someone who is unfamiliar with GNU/Linux you probably arenāt either. Good news: a week after you start using Linux youāll feel like one! Here are some critical things I eventually learned while installing Ubuntu/Mint:
You should expect to use the terminal . Period. Something about your particular hardware or software setup may require special tweaks or install that requires typing. Anyone who even hints this isnāt the case is at best deluded. I know this is a deal-breaker for many people but Iād rather not waste your time.
Locations and commands are case-sensitive . -h means help -H Human-readable (or is it the other way around? More typing yay!). Itās in /etc/ X 11, not /etc/x11 (something almost impossible to see the difference of on a blurry 1080i resolution not being properly displayed).
While the basic user storage locations mimic what you are used to, the underlying system organization is completely impossible to navigate. Pertinent files can be scattered over several locations for whatever reason so donāt even bother trying to figure out a pattern and just follow guides. That said,
Guides helping you to navigate this jumbled mess are possibly outdated so check their dates or you may end up following directions and quite possibly break your installation when you add/remove/alter a file that used to be important but has been deprecated or relocated and now redundant. Speaking of which,
It is possible/probable your distro is effectively a skin of another older distro , so you should search the underlying distro directions too in case there arenāt any for the āskinā youāre using.
All said and done, I am very happy to say I now have my Mint OS on a portable USB keychain that I can use on any PC (assuming TPM permission). The actual OS is pleasantly unobtrusive, nimble, and supports 90% of what I want to do with it. Critical failings seem to be completely relegated to proprietary software (for me, 1080i support was abandoned by all the graphics card developers years ago and Iām unable to either find older working drivers like I can in Win10, or find/figure out the tweaking needed to force the issue). Check all your mission critical programs to see if they are Linux compatible , or āsimplyā learn to use the open-source competitor if they arenāt.
For example changing playback quality from 48kbps to 96kbps.
Worse is that setting this then results in automatic upscaling on everything which will have your CPU maxed out.
Linux in general is awful for audiophiles.
I think their tone is perfect, far too many leap to claim you can use LM or other distros without ever needing the terminal and in my experience itās 100% a lie.
This is true on all systems. Windows and MacOS both also use internal sound servers to mix all audio sources into the single PCM signal that goes to the sound card.
If incoming PCM signals are at some other samplerate than what is being fed to the audio device, they must be resampled.
You canāt feed two PCM signals at different samplerates into a single output any other way. Or even a single one, if the input samplerate is different from the selected output samplerate.
This is how it works everywhere. Not just linux. It is a fact of digital audio.
Just. No.
Your system is not going to be burning up the CPU doing audio resampling. It simply isnāt that demanding a task. Itās something all system do regularly anyway, whether itās upsampling or downsampling, one audio stream (or at least two for stereo) or a dozen, or more.
This to me is an admission you have no idea what youāre talking about. Linux is PREFERRED by tons of audio engineers because its audio handling and routing capabilities on a system level are unmatched, as well as the ultra-low audio processing latencies it can achieve when configured for it on a kernel level.
Well having my cpu maxed out by this simple change was my experience and thatās with a 7800x3d.
It was the last straw and forced me back to Windows for over a year. Changing this setting should not be so annoying and definitely shouldnāt be hammering a cpu so badly but thatās my experience.
I could not care less about the reasons behind any of it. I have a good sound card and want to be able to set the playback quality easily and without things breaking.
Thatās it. I donāt want to learn about PCM mixing sound channels or upsampling or latency. I want it to fucking work.
I will stand by my statement that Linux is awful for audiophiles, Mac beats it all day every day because it actually works without having to open a terminal.
I promise you. It was something else. Audio resampling simply cannot max out a modern CPU, and is something your system does regardless of what samplerate you are using.
Thatās fine. But then donāt make claims about how it works and why it has problems. Just say you had problems.
It was pipewire and I tested it thoroughly. It was putting a single core to 70-100% during audio playback that required upscaling.
No. I described exactly the problem I encountered and will describe it how I want.
Iām not saying it wasnāt.
Iām saying the problem wasnāt āall audio being upscaledā. Thatās nonsense.
Something obviously went wrong. It just canāt be what you think it was.
What you tried should be completely possible and not come with any significant difference in CPU usage.
Which means you ran into a bug. Not some inherent difference in how linux works that means what you attempted is impossible.
The ācauseā you describe (audio resampling to match the samplerate being used), is how all digital audio has worked on all devices since the inception of digital sound. Itās not something stupid that only linux/pipewire does.