Multitrack recording - direct to Audacity or external device

G’day Folks,

What I’m reading indicates that it’s impossible to have Audacity record multiple sources to a multiple tracks. The best you can do is record two mono sources to a stereo track. Am I right?

Is this limitation in the OS (Vista Ultimate 64bit here) or is it a limitation in Audacity? I’d love to be able to plug in multiple USB soundcards and have them all record to their own tracks when recording in Audacity. I’m sure I’m not the only one.

So, assuming I’m right and I can’t do that, is there any external recording device I could look at that would let me record up to four microphone sources to four tracks that I can then ‘import’ into Audacity? Even if I had to manually import each one.

One important thing: portability. I want to be able to take this device on the road with me to record group interviews.

An external mixer recording down to a single track in Audacity wont really help as I don’t want to have to get my interviewees to do soundchecks (and then, as they’re not pros, get the sound levels right once they actually start to talk rather than use their ‘sound check voice’!) and then keep my eye on the recording levels to get it perfect first time. The point is to get it all recorded ‘reasonably’ balanced, then perfect it in post-prod.

Many many years ago I had a four-track cassette recorder. It used the stereo on both sides of the tape to end up with a single-sided tape with four tracks on it. Surely such a thing exists in the digital world?

Rick Measham

The Tascam DP004 [1] has the perfect form factor. The only problem is that it only records two sources at a time! Even the 8-track version [2] of this only records two tracks at a time.

The only thing I can find (not that I’ve searched for hours upon hours) that will do four at a time is the Fostex MR8HDCD [3] – but it’s an 8-track, it’s big and … well … it’s ugly.


No, some lucky users have managed to get Audacity to record multiple tracks simultaneously. This requires appropriate multi-channel hardware and suitable drivers. This has not been tested much by the developers due to a lack of multichannel hardware. If anyone has any spare multi-channel hardware to donate for testing, I would be delighted to test it and see if I can get it working with Audacity.

The main prerequisites are that the hardware must be multichannel hardware, and the driver must be one multichannel driver. Audacity does not ship with ASIO support, so for Windows machines it must be a multi-channel Windows driver.

Audacity does not support recording from multiple devices. The only way to get multiple simultaneous tracks with Audacity is with multi-channel hardware.

Fostex MR-8HD
Edirol R-44
Sound Devices 744T
Zoom R16

probably quite a few others.


I’m having an Ediroll m-16dx, and recored 8 tracks simultaneously:
-Front Stage Utmost Left
-Front Stage Left
-Front Stage Right
-Front Stage Utmost Right
-Piano 1
-Piano 2

And the result is pretty fine.

Actually I recorded 18 channels (16 channes + 2 live Mix), but the remaining channels were not fed.
But as you see, the recording of multitrack is working.

Unfortunately, I’m not that loaded, and cannot dedicate the device to the developer/testers, although I could be contacted, to perform some tests if needed.

One suggestion:
The interface towards recording multitrack could be “enhanced”. Meaning, As you can see in my description, I’m capable of recording 18 channels simultaneously, but in this case, I had only use for the 8 (actually 7) described. It would be swell if before recording, one could assign/prepare the channels to specific tracks

say something like:
assign channel 1 of m-16dx to stereo track 1 left named “live mix left”
assign channel 2 of m-16dx to stereo track 1 right named “live mix right”
assign channel 3 of m-16dx to mono track 2 named “lead mic”
assign channel 16 of m-16dx to mono track 3 named “bass”
assign channel 17 of m-16dx to mono track 4 named “synt”
assign channel 18 of m-16dx to mono track 5 named “FX”
assign channel 4 thru 15 of m-16dx to dummy track (don’t record)

This would reduce the data recorded (6GB of data for 1 hour, instead of the 13G I had now) (recorded in 48kHz).

Friendly Greetings,

Liselore Vermeulen:
Audacity 1.3.10 Bèta
Macbook 2.26GHz Ic2 2GB Mem
OsX 10.6.2

It’s always good to hear from someone that can tell us that Audacity has success with a multi-channel device.
+1 for your suggestion about “channel mapping”
I’ve added your vote here:

Ethersound seems like a natural fit for Audacity. Ethersound passes 16 channels (over a 64 channel backbone). If Audacity could either map those channels (as previously posted) or simply accept and work with any or all of them, it would be a great device for many live sound techs.

At the PC end, there’s no new equipment needed, as Ethersound works over standard Ethernet hardware, with a different driver.

At the other end, you need some hardware that converts between analog and Ethersound. Some digital boards have this ability as an add-on card. It’s not the cheapest answer, but it’s effective and many live sound engineers would find it easily within budget.

I too would love to have the capability to record say 4 tracks at the same time, eg a stereo keyboard + 2 vocals. If 2 persons could record simultaneously that would in most situations be more fun than having to overdub everything. But my recent conversation with focusrite (audiointerface) tech support indicates that WDM-drivers is not well suited for multi-channel, “it is somewhat limited by how the WDM driver is addressed in Windows” the guy says, and goes on to recommend asio supported software if I want to go multi-channel.

Overall, I can sense a tendency of as you walk along towards more advanced features, audacity has problems to adress it, because of asio or other restraints, so people (who can afford it) tend to move on to commercial packages. And in general I guess that audio interface vendors have no particular interest to support freeware players, in particular if they also market their own recording software, and I guess there’s no denying that asio is the de facto standard if you move in the direction of professional equipment and multi-channel.

On Windows computers ASIO is required to provide the level of performance required by the professional audio market. Well designed WDM drivers are capable of pretty decent performance, but often don’t cut it for more advanced studio work.

ASIO is not used or required for Linux operating systems as there is already a native high performance sound system for Linux called “Jack”. The Jack audio system will often provide higher performance, lower latency and lower CPU usage than ASIO when used on equivalent hardware.

The dominant position of Microsoft Windows on desktop computers but their relatively poor sound system places ASIO as the de-facto standard. However for anyone setting up a computer based audio studio from scratch it would be well worth considering Linux as the operating system due to its high performance and low cost. Audacity 1.3.12 works well with Jack on Linux systems.

I think that ASIO is available for Mac computers, but is far less common than on Windows because “Core Audio” (the native sound system for Mac OSX) does not suffer from the same poor performance issues that are common to Windows PCs.

"Interface support is normally restricted to Microsoft Windows, since other operating systems (e.g. Apple’s Mac OS X or Linux) do not have such mixer latency problems "


It’s good to know this isn’t a Secret Aural Teaching – that this can actually be made to work with available materials.

The previous champ was a multiple-referenced internet message that someone on old hardware managed to get a multi-channel recording in Audacity using obsolete software drivers. The modern ones wouldn’t work and had to be removed. Fascinating on Reddit, but not useful for production.

It is pointed out that we only get the problems, not the successes, but you would have thought that somebody would have posted a problem where channel 6 of 8 is distorted and everything else seems to work. Clearly indicating a successful multi-channel production – aside from the distortion.


@Steve, sounds interesting about the Linux Jack thing, but don’t you still need driver support from audio interface vendors to be able to plug it into your system? I’ve recently looked around a lot on the web for audio interfaces, and right off the cuff I can’t recall one single instance of someone saying “our hardware supports win mac AND Linux”. (Reminds me of the problems I’ve heard in the past of Linux trail blazers having problems getting their printers to work because of driver issues)

Fortunately ALSA supports a huge range of sound cards - most will work out of the box, though some may require a bit of tweaking.
Jack is usually configured to use ALSA as the sound card driver and can therefore be used with most sound cards.

Linux has built in support for a much greater range of hardware than is the case for Windows, so there is much less reliance on drivers being provided by the manufacturer/vendor. As Linux is becoming a more popular desktop operating system, a number of hardware manufacturers/vendors are coming on board and providing Linux drivers, though Linux developers strive to provide native, open source support as well so as to avoid any possible licensing or other problems that could occur from reliance on closed source proprietary code. Notable manufacturers/vendors to directly support Linux and provide drivers include NVidea, AMD, Dell and HP. Other manufacturers, such as Intel, provide (as source code) “development drivers” to the open source community.

When Microsoft released Vista, they reversed the situation for many users by making it more likely that their existing hardware would work correctly with Linux than with Windows.

It is pretty easy to test if hardware is going to work out-of-the-box with Linux by running a “Live CD” version of Linux (a Live CD runs directly off the CD rather than needing to be installed). If hardware does not work with a Live CD, then that does not automatically mean that the hardware will not run under Linux, but it may mean that a bit of tweaking is required to get it to work.

NVidia provides graphics drivers for linux, but they’re not free (as in the free software definition).

The last two usb audio interfaces I’ve used on linux worked out-of-the-box… Many manufactures probably don’t explicitly specificate that their products work on linux… but despite that most of them do work on linux.

The device collection rejoices. A M-Audio FastTrack Ultra 8R (usb) device has joined the rank.
8 mic/line inputs. And the recording is swell.

The Edirol, only having 4 mic entrances (with mic pre-amp and phantom),… suggested a purchase of extra mic preamps to use the 12 other line-entrances for recording of mic signals. The price for these pre-amps was bigger than the price of the M-Audio 8-channel. Hence the choice.

Now REALLY swell would be if I could record the tracks of both devices simultaneously.

now stevethefiddle states:

I was wondering whether this limitation is carved in stone…
Wouldn’t it be suggestible to allow recording of multiple tracks of multiple devices (keeping and extending my previous suggestion for the “Channel Mapping” feature)…
Now since we’re suggesting anyway… This multi-track, and multi-device feature… should not be limited to input only…
Having that M-Audio FastTrack Ultra 8R,… which supports 8 different audio-channels OUT tooo (I’m believing)… But the device has also a FireWave 5.1 ieee1394 adapter (6 extra channels)…

Now immagine having a “jackboard” in audacity… allowing the setting of
for recording: input device+input channel → record track (with support to individual choise of “create new track” and “append to existing track”).
permitting to name to-be-created tracks beforehand.
for playback: record track → output device + output channel…

Thanks for the info. Feel free to write a review if you wish: Sound Card Reviews

There is a practical difficulty there in that unless the devices share a common clock signal they will gradually drift out of synch.
Having said that, for short recordings they would probably be close enough for that to not be too much of a problem. Apparently on Mac and Linux it is possible to create a single virtual sound card (called an aggregate device) that unifies multiple sound cards with a single sound card driver. It does not totally overcome the synchronisation problem, and for it to work properly requires either using high end sound cards that give access to their clock signals, or physically modifying the sound cards so that they share a common clock signal. I’ve not done this myself, but you can Google it if you’re interested.

On Windows it may be possible to do a similar thing using “Virtual Audio Cable” Virtual Audio Cable (VAC) - connect audio applications, route and mix sounds
If you try this method and get it to work, please post - I’m sure a lot of people would be interested.

Hi liselorev!

You wrote some days back:

My Fast Track Ulta 8r arrived yesterday and I cannot get it to work satisfactorily. I only get 2 in and 2 out. Could you please post some details what your settings are? I mean stuff like OS and driver/firmware version, the Audiacity device settings and M-Audio Contol Panel settings (ASIO or no ASIO)?

Any help would be appreciated!

Thanks a lot!


Unless you have built Audacity from the source code and manually added ASIO support, ASIO will not be supported by Audacity.

Most multi-channel sound cards in Audacity require that you record using either 1 channel, 2 channels, or ALL of the channels (which I would assume is 8 for your device).
If you get it working with 8 but you only want, say, 6 channels, you will probably need to record 8 channels and delete the 2 empty channels after recording.

Some devices have a “multi” driver listed in the Audacity “Devices” Preferences (Edit menu > Preferences > Devices).

Make sure that the device is switched on and connected to the computer before you open Audacity.

For playback Audacity currently only supports mono or stereo.

If you get this working correctly I’m sure that you could make many other people very happy by posting detailed step-by-step instructions on out “Recording Equipment” section.


Then I fear I will have to build Audacity myself… could be complicated for me as I have no experience with compilng stuff on Windows.
To record all channels even if I don’t need them all is ok for me… there is plenty of harddisk space! :slight_smile:

Is it planned to add support for multi-channel output to Audacity? That would be a feature I REALLY would like to see… if I were a better software developer I might try to do it myself, but programming is only a hobby of mine…

Anyway, I will report back if/when and how I got multitrack recording working with the M-Audio Fast Track Ultra 8R and Audacity.


It an often requested feature that the developers are well aware of, but I guess it’s also a pretty major development project, so I doubt that anyone will be keen to start messing with it before Audacity 2.0 comes out.


I must admit I did not yet check the output channels yet,
using the device only for recording 8 microphone channels in,
using Audacity 3 beta (1.3.11 beta I believe it was)
OS: Mac Os X 10.6.4

After installing the m-audio (propriatary unfortunately) driver… and rebooting, the device was recognized.
Some problems were encountered when the mac had gone in standby modus and coming back…
the M-audio was not recognized anymore untill after reboot.

The site of m-audio (somewhere) also states that the driver is not capable of connecting two devices simultaneously…
However disconnecting and reconnecting the device is no prob,…
so I immagine 2 devices could be “configured” using 1 laptop,… but recording would not go so well (I fear).

The matrix in the device has to be configured (using the m-audio tool) to use the different outputs.
I believe the “startup config” is NO mic-in to any out channel.
Switching the device off and on,… looses the configuration (bummer).

edit d.d. 2010-12-27:

That last remark seems to be incorrect (about the default patch, and/or the saving of configuration).
I repeatedly started my m-audio ultra 8r, (without connecting it to my macbook), and each time, I discovered the configuration:
mic 1 → out 1
→ out 7
mic 2 → out 2
→ out 8
mic 3 → out 3
mic 4 → out 4
mic 5 → out 5
mic 6 → out 6
mic 7 → …
mic 8 → …

However if the device is connected to the macbook (driver), the patch is set to
mic n → … nowhere (for n in 1…8).

There is another problem waiting at the output end of the process. There are very few sound file formats that will manage uncompressed, multi-channel shows. So File > Export may be an adventure.

We tell people that Audacity Projects are useful for saving multi-channel shows. We also tell people that Audacity Projects are a terrible way to archive work. Projects are brittle and easily damaged.

Did you understand the problem with ASIO? Audacity can either include ASIO in the program or distribute the program. Not both. If you produce an Audacity with ASIO support, you can’t distribute that product.

It’s a production dead end.