LinuxDAW – Quality audio software for Linux (linuxdaw.org)
I made/released this exactly one year ago today.
It's a simple webpage giving an overview over quality software that runs on Linux. It has search and filter to narrow down what you're looking for. Default sorting is latest additions so it can be used as a "news site". There is also a RSS feed available.
I needed this myself, so I used it as a pet project to upgrade my knownledge from Vue2 to Vue3.
Source code is ofc open and contributions/feedback is always welcome. https://codeberg.org/fractalf/linuxdaw.org
Cheers!
64 comments
[ 2.7 ms ] story [ 110 ms ] thread[1]: https://linuxdaw.org/?t=daw
Nothing by falkTX / KXStudio?
https://kx.studio/
https://github.com/falkTX
EDIT: Oh I see there are some but they're not listed under Developer.
Also, perhaps add a CLI format?
I know some people use Glicol CLI on Linux:
https://github.com/glicol/glicol-cli
> NodeList(524)
That's a lot of quality audio software for Linux :) (Edit: I realize now there is a counter in the top right... Oh well)
Happy to see it, although I personally mostly use Windows for audio stuff... If Ableton worked on Linux I'd probably leave Windows behind.
I applaud your initiative and suggest you give more thought to categorization.
I agree, and moreover, the more complex the taxonomy, the more likely you'll find people who disagree with it. So perhaps you're right that a simple tags list is the best "simple" solution, and I'm not saying your list is deficient.
One thing I had in mind is allowing an "AND" operator, like searching for "vintage" AND "eq"
> a) usually a "thing" can only belong to 1 category and b) a category often belong in a hiarchy of categories and can only be 1 place
That's why I love hierarchical tags (where tags can have multiple parents) : You can put a "thing" under more than one category/tag while a tag can belong to more than one parent. However, it's true that this is rarely implemented.
There could possibly be a clearer name for this effort, since both Reaper and Bitwig run on Linux natively and are top notch DAWs (actual DAWs, not VST plugins). Plus VCV Rack, which is not technically a DAW but close.
Linux native VST plugins is a way more precise description. DAW is a very specific software category. It's like calling Prettier extension for VSCode an IDE.
Edit: seems like both linuxvst.com and linuxvst.org are available.
Yes, although I would recommend Yabridge rather than linvst.
https://github.com/robbert-vdh/yabridge
It should be already packaged in some Linux distros, but building and installation are straightforward. Basically it installs some libraries and an executable which once called will convert a Windows plugin in .dll form to a Linux+WINE loadable .so library. You install a windows plugin through WINE the usual way: setup.exe etc (or just drop the dll if that's the way it is distributed), then once it is installed, call yabridgectl adding the path of the new plugin dll, then call yabridgectl again but this time adding the option to convert plugins (going from memory, can't test it here), then you'll have the Linux loadable .so library in the same path of the windows one. now just add the same place to your Linux DAW plugins path and it will automatically see and load the converted plugins. If you install a new windows plugin, just repeat the two above steps.
One nice and often overlooked feature of loading plugins through the WINE compatibility layer is that many 15+ years old still great plugins that stopped working on Windows ages ago now are perfectly usable again on Linux, with the only problem of having to work with a much smaller GUI since they were created when 1024x768 or smaller desktops were a thing, so they can appear stamp sized on today's monitors, but size aside they work fine.
About Jack, let me repeat this one more time Jack is not necessary at all to obtain low latency, ALSA is more than enough; if you're writing audio software, having Jack support is nice, but please don't depend on it as it just complicates things.
I've been using linux audio for the last 15 years and I've been trough it all. However, things have much improved since I've switched to Pipewire. No more need for a bespoke implementation of the jack daemon, it's implemented natively right there in Pipewire, vastly simplifying setup and removing any annoyances that keeping the Jack daemon working entailed. As a bonus any jack client/jack graph visualizer, including carla, gets a direct view to all Pipewire clients, not just native jack clients. You can then use Carla to route audio directly from your hardware inputs and outputs trough, for example, pulseaudio client (blissfully unaware of whats happening) and then trough as many jack clients as you wish. It's truly a magical™ piece of software.
As for Carla, make sure you are in the "Multiple Clients" mode, activate the experimental features and check "Enable plugin bridges" and "Run plugins in bridge mode when possible". This seems to make it more stable and make it so that if a plugin crashes Carla will disable it and at worst you will lose the related plugin parameters.
I've been running Carla with Pipewire for more then a year as my system wide DSP stack (almost like a virutal AV Reciever) and its fairly stable.
I mean, technically JACK is built on top of ALSA, you can always get lower latency by going through ALSA directly than by involving JACK... but it's pretty hard to implement correctly.
They may have had a head start because of very popular and critical software that is un-portable, but the technical superiority is on our side.
AFAIK, there have been some professional successful DAWs for Linux and I can only think it will grow over time.
I saw people using OBS with a few USB cameras with reasonable quality making good quality webcasts and I'm sure that is also doable in the DAW-land. The future seems bright for musicians and recorders who can now be free from proprietary software and build a reasonable good studio at home using non-imorally-priced hardware.
It could also be a HDSPe AES (32 channels) or RayDAT (72 channels). These cards are quite widely in use.
Also the fact that the card supports 96 channels doesn’t mean you need to have as many A/D inputs necessarily.
Linux has a bit of a Thunderbolt audio problem though. Apple dragged the world into compatibility with class compliant USB; there's no equivalent for Thunderbolt though :-( we're back in the bad old days of individual (nonexistent) drivers per device.
Getting some Betamax vibes here ;)
But I also don’t really see the superiority versus Core Audio or ASIO-based systems to be honest.
The best site I've seen so far for this type of content.
https://codeberg.org/fractalf/linuxdaw.org/issues/44
I’m trying to de-Windows myself, but my pro-audio setup is the gap I can’t seem to fill.
I currently have an RME HDSPe MADI FX [1] which allows me to bring in 96 channels of 96khz audio via 3 MADI connections. They connect to my AD/DA convertors [2].
I found source for a driver [3], but couldn’t get the thing working, and it doesn’t seem particularly well supported/documented.
Moving toward Dante or AES67 is also an option (because my AD/DA convertors support those formats) if there’s any software solution that can support those formats, but there’s minimal online info. So if anyone’s gone this route I’d love to know how you got everything working!
[1] https://www.rme-audio.de/hdspe-madi-fx.html
[2] https://www.ferrofish.com/portfolio/a32pro-dante-converter-m...
[3] https://github.com/adiknoth/madifx
Because I don’t know of any USB device that can do 96 channels of pro IO. Also, it always seemed preferable to me to avoid the issue of ‘yet another wire’ and ‘yet another source of latency’.
Being attached to the PCI bus always felt like the most robust approach to getting audio in with the lowest latency (which it is on Windows).
I’d probably prefer to look at Dante/AES67 before USB, unless there’s a compelling reason not to. Mostly because it gives room for future expansion where a fixed soundcard doesn’t.
> yes, on Linux with pipewire and Bitwig you can use multiple cards
How are you clocking the cards? If they don’t share a single clock then you’re gonna get phase issues. The idea of going back to a word-clock does not appeal to me! I’d much prefer a single device with one clock. The fewer moving parts, the better.
Holy crap, that's a lot of input channels, didn't know it was possible. I also have a Behringer ADA8200 that connect to the 1820 (ADAT) which gives me 16 inputs. I'll admit it's not enough to run all of my gear in at once and this is just a home studio. I can imagine 96 channels is very useful. Maybe 6x my setup would work = 96 inputs ;)
> How are you clocking the cards?
You are obviously more advanced than me, not sure what you mean here. I'm using the NI card basically just for the output and the Behringer for inputs, 24/48. When recording, if there is a latency (which normally Bitwig can "fix") it's just a matter of moving the audio a couple of ms.
> PCI
I've had earlier problems with PCI cards, in that they pick up a lot of internal noise from the motherboard, so I've stayed away from that after firewire/usb cards became available
96 output too. I have a lot of outboard gear, my SSL Sigma Delta alone is 32 in/out for mixing, that's before I get onto all my outboard compressors, EQs, reverbs, synths, drum machines, and modular gear. I use the soundcard pretty much as a virtual patch-bay, I can send and return to any piece of gear in my studio like a plugin (get Bitwig to measure the latency and automagically everything syncs).
> You are obviously more advanced than me, not sure what you mean here.
Because digital audio is in discrete time-slices, you need to sync the slice starts so all devices are aligned. Otherwise you can have subframe issues where phasing becomes a problem - in the worst cases it can cause audio glitches (which I've had with older AD/DA convertors).
This can be solved by using a Word Clock [1], which is the 'old' way, or via AES/MADI/Dante - which all have their own clocking protocols. The devices all need to be clocked by the same source, which you can't do by pairing devices in software.
My current soundcard is connected, via MADI, to my 3xFerrofish-A32-Pro AD/DA units. The soundcard is the source of truth for timing. The AD/DA devices all sync to the soundcard and therefore have no phasing issues. This gives 96 channels of analogue audio in and 96 channels of analogue audio out.
> I've had earlier problems with PCI cards, in that they pick up a lot of internal noise from the motherboard
Sure, if you plug analogue sources directly into the device. But my analogue inputs/outputs are all going into my AD/DA convertors, which are separate units. They are digitising the audio and by the time it arrives (via MADI) into my soundcard it's already digital. That means no chance of any noise from the computer.
Probably what I need is something like this [2], it seems to have Linux support, but unfortunately isn't produced any more! I seem to find that a lot with these Dante devices. Maybe I'd be better with AES67 as that's an open-standard that is compatible with Dante, just I haven't found anything yet!
In theory (because Dante/AES67 is over IP) a virtual soundcard can be written, I certainly have a powerful enough machine to run it, but again I haven't found anything yet. I'd write it myself, but I have a million things going on, and that's a distraction I don't need!
It does seem like there's a bit of a gap when you get to the real pro end of audio, which is a shame because the stability of Linux would is a huuuuuge selling point when producing music.
[1] https://en.wikipedia.org/wiki/Word_clock
[2] https://www.audinate.com/products/dante-enabled/four-audio/f...
heh, sorry!
Looks like I probably only searched for "Dante virtual-soundcard linux" in the past, because when I search for "AES67 virtual soundcard linux" I got [1] and [2] which may be the answer to my problems (if my machine can handle that many audio streams, but it should be able to with 64 cores haha!).
Looks like that's going to be my project tomorrow.
[1] https://github.com/bondagit/aes67-linux-daemon
[2] https://gitlab.freedesktop.org/pipewire/pipewire/-/wikis/AES...
"The Fart Machine is not just any ordinary guitar pedal or plug-in. Using our unique rectal microphone technology, we have modeled every component of the human bowel system, and now the Fart Machine can turn any sound into flatulence. This technology provides an unprecedented level of control over your fart sounds, enabling you to create a range of unique and unforgettable sounds that will leave your audience in awe."
I miss music software and its sampling and remixing culture.
Basic use case is being able to take input from an old MIDI keyboard and feed it into a DAW so that a simple melody can be turned into something more exciting.
I would appreciate someone being able to point a complete beginner towards a path (if one exists) in which initial work getting familiar with an interface will not be wasted if more advanced functionality is desired later.
I don`t mind paying for tutorials and/or books or the software (and I see Bitwig has some tutorials) but I want a rich environment that is going to reward investment of time. I would prefer something without lock-in and strongly favor FL/OSS.
1. https://manual.ardour.org/introducing-ardour/understanding-b...
Linux Audio Wiki - https://wiki.linuxaudio.org/wiki/start
KXStudio : Documentation - https://kx.studio/Documentation