• 0 Posts
  • 13 Comments
Joined 1 year ago
cake
Cake day: June 18th, 2023

help-circle



  • I didn’t get to spend too too much time doing pro audio on linux because as soon as I realized Wwise will not work, I didn’t spend much more time in Reaper after that. But it was good, especially with an audio interface (if you’re buying an audio interface, check to see how well it works with linux. Apparently some may have issues)

    By the way, the whole point of Debian is that it has older software, and in exchange you’re almost guaranteed to have a system that doesn’t break. But for some professional software you’ll want the newest version. I recommend using Flatpak for that stuff instead of Apt (like for Reaper)

    When installing Debian, when it asks “Allow login as root?” be sure to select “no”. This one step is why some people don’t recommend Debian, saying it complicates the install process. But if you get that right, then you’re all good. Or I think sometimes it will instead ask you to create and type in a root password, in which case you should leave it blank and click next. You only want to make a password for your user, not for the root.


  • First off, I want to make it clear that the distro doesn’t really matter. Different distros are just what it comes pre-packed with by default.

    TLDR:

    • For something easy to use: Linux Mint

    • For something that has pre-installed audio software (but maybe not the ones you want): Ubuntu Studio

    • If you want to build your system from scratch: Debian (or Arch if you want the latest and greatest software, and don’t mind the occasional update breaking your system around once a year or so, and needing to spend an hour fixing it)

    • Regardless of which distro you get, use JACK or PipeWire for your sound server. PulseAudio (on its own) has too much latency.

    More details:

    I first tried Ubuntu Studio. It comes with a lot of software related to audio production. But I found it to be insanely slow, and it didn’t even come with Reaper anyway.

    I tried OpenSUSE because I liked that it had the option to manually deselect the software you don’t want (and I was too much of a beginner to know how to pick my packages from the ground-up). It worked well.

    Eventually I moved to Debian. I didn’t want any of the extra fluff and found it was pretty easy to choose everything myself. One thing that’s important is that you don’t want to use PulseAudio. Either use JACK (which I think needs to be used in conjunction with PulseAudio actually) or use PipeWire, which is what I use.

    For any Windows software, use Bottles to emulate them on Linux. I actually ended up needing to go back to windows because of one audio software: Wwise. There was no way of running it in Linux. A VM probably would’ve worked, but that would’ve been a massive hassle for how I’d need to use it.

    Free Linux VSTs: https://vital.audio/ https://lsp-plug.in/ https://github.com/TukanStudios/TUKAN_STUDIOS_PLUGINS

    Paid Linux VSTs: https://www.acmt.co.uk/products/index.html https://librewave.com/ https://www.audiodamage.com/collections



  • You’ll want to decide on a desktop environment or window manager (or compositor). That’ll be the biggest determining factor of what things will look like. From there, you’ll want to either read the manual or arch wiki on how to customize the different aspects of it.

    If you decide you want a tiling window manager, Hyprland is nice since you mentioned you wanted animations. But it’s only recommended on rolling release distros at the moment. It also might not work well with Nvidia.

    What kind of “app behaviour” customizations are you wanting to do? That sounds like it would be app-specific. My main form of app customization is to find ways to change the colour scheme (to fit everything else), and also to change the keybindings (I like using vim-like key bindings whenever reasonable)


  • You haven’t given any info regarding what you want to try.

    If you just want to try something different, Arch Linux is an obvious one. A nice learning experience. However, I’d say rolling release is not as recommended on a machine that you’ll be using less than twice a month, since I hear people say you want to update your stuff no less frequently than once a week on a rolling release OS.

    So another idea is NixOS. I think it comes with a stable release option? I haven’t tried it, but it’s another option if you want to install something for the learning experience.

    If you want something easy to install but different, consider Fedora or OpenSUSE (either version)

    For desktop environments, if you want a learning experience for something potentially fun, try a tiling window manager. Sway is one that I’m trying right now (it’s just i3, but with Wayland). Or for something easy but different, any of the big DE will do, like Gnome. I haven’t tried anything other than KDE or Sway.



  • Pipewire is much better than Pulseaudio, especially for pro audio work because of its low latency. Another popular option is JACK, which must be used in conjunction with Pulseaudio. Harder to set up, but is also great for pro audio. Some audio engineers were having issues with Pipewire when it first came out so they went back to using JACK, but I think Pipewire has improved. Pipewire has been flawless on my end.

    If you’re not in pro audio or any kind of multimedia work, it doesn’t really matter and you can just stick with whatever comes pre-configured on your distro. But my vote goes to Pipewire as the best server for pretty much anyone.


  • So I work in game audio. I have a masters in music composition, and beginner knowledge of coding just to add a few lines of audio-related code to my client’s spaghetti code.

    What you’re asking for is not a simple task, and will certainly need AI or machine learning or something along those lines. You could more easily make something like this if its just a single melody with no other backing tracks (and there’s plenty of software that does this), but to understand and distinguish separate instruments, including drums, hear them all as separate voices, and turn them into a whole other medium, is on a whole other stratosphere. If someone’s going to make something like this, it’s going to be made by a corporation and sold as a $1000 plugin.

    But hey, at least it will most certainly be in C++, because that’s the language of choice for most audio plugins.