Share your experience running VCV on Linux

@David I also use carla to do everything outside of VCV Rack. For example: adding effects with plugins outside of VCV and so on.

Hello! Solus Linux user for 2 years. As main OS at home, don’t use Windows at all. As second OS at work.

1 Like

I’ve recently switched to using the SkJack 4x4 module. It makes quite a difference to the module load my patches can bear before xruns appear. However, I’ve noticed that its connections are not persistent in saved patches, and when Rack closes it reports a segfault and core dump. Maybe a Jack API problem ?

re: patchbay: Just what’s available in QJackCtl.

I used / attempted to use Jack (but not SKjack) until a few weeks ago, with suboptimal results. Since then, I’ve stopped using Jack at all, and relied on Alsa alone, and VCV performances improved very noticeably. I still have to be careful and avoid the most CPU-hungry modules like Valley, but I can add to my patches many more modules than before.

I haven’t used Skjack since I don’t expect it to allow radically better performances than just using Jack (am I wrong?)

1 Like

Aw yeah, Helm was the first VST synth I ever used and I still come back to it 5 years later or however long it’s been.

1 Like

6 posts were split to a new topic: Promotion on new Vital Wavetable Synth

That’s a bit problematic for performance, so you really want to get jack working if you can.

If you can keep your entire audio/MIDI chain in jack-land, you gain the ability of rendering your audio, as opposed to recording it. Rendering your audio has two benefits: ti’s faster, and there are no XRUNs, no matter how complex the patch. Everything will come out sample-perfect.

Of course this only works if there is no live/performance aspect of the patch, so it may not be applicable to what you do. But if you make music that could be rendered, you’ll benefit from creating a work-flow that allows you to render. (Of course, there can be hybrid approaches too, where you render your base track and layer human tracks over top.)

Another benefit of jack rendering is that you don’t have to run in real-time mode. When you’re working on a patch the occasional glitch won’t matter, and when you render there will not be any glitches.

I actually create my workflows with that as the primary requirement (I’m not a performer, so loosing the ability to perform live isn’t a big thing for me).


i’m mainly linux user too and i can confirm that jack is working perfectly fine once you get it setup correctly (which might take some time and effort) … even on absolute low end hardware like the sonaremin (i.e. arm cpu single board computers with much less cpu power than a normal pc) i got it working really well …

best wishes - hexdump


If you need low cpu modules, most of mine are.


Not to raise the dead, but I am considering running rack under Linux, but I am curious about performance. Realistically speaking, am I going to get better performance under Windows 10 or is Linux comparable for rack?

1 Like

It depends on a lot of factors, but as a general rule of thumb, Linux is faster than windows.

1 Like

As others note, it’s not easy to say which is “better”, but I think the answer you are looking for is that VCV works great on Linux. It you are happily using it on Linux there is no reason to switch to a different OS. That’s what I hear, anyway.


My perfomances suck. Whether linux or windows, doesn’t really matter :see_no_evil:

But VCV itself is running very smoothly on Linux, the rest depends on hardware and software configuration.


I can confirm for me VCV runs well on both Win 10, and ubuntu 20.4. It took me a bit of fiddling with the block sizes and Jack. The F3 cpu meters show about a 20% better performance under Linux, but I don’t notice that difference in practice. The big difference for me is the cooling fan noise under windows is greater, but the core temperatures are comparable.

This is all on a single pc, with dual boot

1 Like

Thank you! If I can get rack to work with OSS, since that’s the only driver support for my interface, then I’ll be set.

I generally route things to Audio-8 regardless of whether or not I’ve fired up jackd.

1 Like

I think it’s likely to be easier to shove unneeded things out of the way on Linux. I have a very minimal Slackware set-up, and I’ve put serious time into keeping the bloat down. I’m not sure how you’d do that with Windows, and I know on my one Mac mini that logging on seems to create a metric f’ton of processes.

I think you mean :hear_no_evil: :stuck_out_tongue_winking_eye:

I’ve been playing with VCV Rack for the past month or so on Linux and MacOS, on an old but able 2010 Mac Pro with dual Xeons, lots of memory, and an AudioEngine D1 USB DAC for sound. I gave up on using MacOS for it, because I was getting lots of artifacts, (fuzzy noises that don’t belong) even running fairly simple patches. I think the Plateau reverb was a culprit, because of the resources it uses, but I’m not sure. Mind you, because this is a vintage computer, I’m stuck with Mojave, an old OS.

Booting under Ubuntu Studio 20.10, and using Jack, the same patches generally run fine, without the artifacts, and seemingly lots of overhead to spare. VCV Rack runs smoothly and reliably, with some quirks, like pop-up windows when saving a patch drawing under the main window, hidden, instead of on top. This has fooled me into having to kill the app to exit it many times. I think I still haven’t learned my lesson.

Although I’m a long time Ubuntu user, I’m no expert. I found I have to run Jack, if I don’t, and use ALSA directly, bad things happen if I try to say switch to Firefox to refer to a tutorial video. I’m finding Jack and the whole audio ecosystem in Linux to be a bear to figure out. Every once in a while (daily) it will explode in loud static for no apparent reason – I’m worried about blowing out my speakers, headphones or eardrums.

As a non-expert, the whole Linux audio ecosystem seems to me to be in a very confusing state, with hundreds of ways to do things, combining overlapping tools and components. It’s full of jargon and abstract concepts that no one seems to be very good at explaining. There’s a big gap in online resources between the gurus who tell you to just compile from source goddammit, and regular users who happened to solve a problem for themselves by trial and error, but don’t know how they got there.

That said, I plan to stick with Linux for my music environment. This is a hobby for me, and figuring things out is half the fun. One of the (many) things that impresses me the most about VCV Rack is that it is platform agnostic. I hope it can stay that way!


While I think I’ve had better luck with Linux audio, I have to agree it’s pretty rough terrain. I’m here, more or less, because I’ve been messing with UNIX-oid systems since '89. I have next to no Windows skill and very little time on MacOS. I’m trying to figure out how best to control instruments and effects without touching the computer keyboard or mouse, but I’m stuck on a bunch of things. (Too many options and not clear on the best approach.)