I teach classes about audio recording, editing, and podcast production. I’ve been working in audio for over 25 years.
I often have to mention or demo Audacity because it is free, and because many people start with it.
However, I am continually disappointed in one feature of Audacity that makes the app difficult to recommend: The inability to independently record multi-channel audio, even if your interface supports it.
All other editors/DAWs allow me to independently select inputs. Not Audacity. Why is this?
I’ve even run tests against other apps, and I think the way Audacity is programmed to access audio interfaces is just plain wrong. We should have independent access to channels/inputs - not whatever Audacity is doing now, which appears to be merging them together when making a Mono recording, or panning hard L/R.
Anyone with a Scarlett Solo or similar device can test this for themselves. Go ahead, use the options Audacity gives you currently, and you will see that you have no independent control. A mic input in channel 2 recorded in Audacity is not going to give you the same waveform as you’ll get in Ableton or another DAW.
The feature that you are referring to is called “channel mapping”, and it’s a common feature in Digital Audio Workstations (DAW).
Audacity spent the first 20 years as an audio editor. It was never intended to be a full DAW.
Since Muse Group acquired Audacity, they’ve added some DAW-like features, including non-destructive time stretch and non-destructive effects. They have not yet channel mapping. Reports from Muse Group suggest that they intend to continue this direction to eventually turn Audacity into a simplified DAW.
Thanks for the background. To me, this is fundamental functionality, and I am looking forward to it being included. I watched the hour long video about the history of Audacity code, and understand it might be a little while, but is good that they are moving in the right direction.
When I teach audio recording and editing classes, I want to be able to recommend Audacity without reservation. But for now the lack of channel mapping throws a wrench in the gears of my lessons because the way Audacity talks to an audio interface does not make demonstrating how an audio interface works very intuitive. The current design adds several caveats to mono/stereo recording that muddy the waters for new learners. So yes, hoping for the best in 4.0 or later.
If you want to recommend a DAW, I’d suggest Reaper. Reaper isn’t free, but it is inexpensive and it has an unlimited free trial. It is probably the most popular full-featured DAW.
Not so good for Linux users. On Linux there’s only really one maintained audio editor, which is Audacity. As Audacity becomes less of an editor, it leaves a small hole in the Linux ecosystem. I’m sure something else will come along to fill it’s place, but in the short term it’s not good for Linux musicians.
I don’t see why everyone can’t benefit. For example, I use Adobe Audition almost entirely as a two-track editor, even though it can technically do multi-track and be considered a DAW. Audition properly allows discreet assignment of audio interface inputs, which is critical to podcasters who are working with a variety of wonky interfaces and setups.
When Audacity fixes the lack of discreet access to inputs, it will be helping so many users avoid unnecessary workarounds and troubleshooting. It will “just work” the way any other audio capable application does in 2026, whether you use it as a 2-track editor or DAW. You can’t use any app as an editor if you can’t properly record the audio in the first place.
The reason we’ve been holding off on this is that we have low confidence in the ability of our backend to handle this kind of thing. A particular big questionmark at the moment is PortAudio, which hasn’t seen a release in half a decade. Instead, switching the backend to the tracktion engine is being evaluated at the moment, though making this transition would be a similarly big transition as the new UI system in Audacity 4. So I’m afraid this feature will only make it into Audacity when Audacity 5 comes around, which is quite some ways away still.
Thanks for the additional context. I’m wondering if it would be possible to get an MVP of two or four discreet channels working (to capture all of those people using the ubiquitous Focusrite and entry-level devices). That’s enough for most entry level uses, such as podcasting and education. I’m thinking of the folks who have an interview based podcast with guest and host, or folks that record themselves at home (guitar/vocal).
Maybe after getting two or four channels to work properly you could expand to eight to capture the next level of user. But I’m guessing that the underlying technology issues and technical debt that you are referencing are there regardless of how many channels you are dealing with.
If/when this gets on the radar, I’d love to help test.
Actually, no. In 30 years of using various programs and interfaces, Audacity is the one app I’ve seen that seems to approach this area quite differently, giving you little to no control over mapping. For example, here’s how Adobe Audition displays the options:
Even with a basic two-input interface, Audacity will not allow you to use input 1 and input 2 as separate (discreet) inputs. You are forced to use them as stereo, which is a separate use case and not always what a user wants to do.
I’m going back with this too. The last time I worked a Windows based system it was with a digitec HD and Protools, but that was mapping channels from a mixing console via the digi HD. I’ve been on Linux since then, which usually takes a more modular approach.
I don’t know about other platforms but I built a Windows ASIO build script as the team could not redistribute builds due to Steinberg’s licencing (no longer an issue).
As I note in the introduction, whilst that adds multi channel support, it is basically useless. Audacity 3 is pretty much mono/stereo only. If you want more you’ll need a DAW.
Ah. yes. I totally realize that. But my point is that if the app truly will stay focused as an edit with only minor input support, it should at least provide that support correctly. I’d be totally fine with two channel access provided it is possible to use the inputs discreetly.
Input 1 (MONO) - Assignable to L or R
Input 2 (MONO) - Assignable to L or R
(AKA dual mono)
Input 1 (or) 2 - Assignable to L/ R (2-channel mono)
Inputs 1/2 Stereo assignable to L/R or R/L recording to a stereo track (ex. stereo tape deck)
All of these permutations have practical use cases I see every day for audio conservation and many of the reasons people use Audacity.
As things stand now, students learning within the confines of Audacity are learning about input selection in a non-standard way compared with all other apps on the market. I am not at all suggesting that Audacity become a 128 input beast. There are already plenty of options for that sort of thing! Maybe the most I would suggest would be 8, to compare with free Lite programs that some manufacturers offer. But that is secondary to the point of my post.
Point taken. Though, in 30 years, the only one of those I have heard of or used is WavePad. And just because those apps (which be honest, are on the fringe) don’t have basic channel assignment doesn’t mean Audacity should continue down that road. Audacity is much more visible in the audio community, the podcasting community and in the education sphere. But right now the lack of properly assignable inputs is holding it back from being what it could be for so many people getting started with audio.
Even though I pay for Adobe Audition, I am still forced to use Audacity for education and workshops because that’s what my audience can afford, and Audacity is honestly a great starting point for people. But during every class I teach, I have to highlight this “gotcha” about the channel mapping. Why? Because, if a student goes out and buys an interface to record a podcast, they will immediately encounter this issue, and get sidetracked.
