Multi-track audio recording in the field

Multi-track digital audio recording has become progressively more accessible and more affordable over the years. Some truly excellent software, such as Ardour and Audacity, are available for free and with excellent community support. However, multi-track recording hardware still tends to be rather expensive and has become more and more complex. Most of the professional solutions tend to come pre-configured as turn-key hardware-software packages. Pro Tools from Digidesign is one such product, but there are others, of course. Lower-end gear can be mixed and matched "à la carte" but typically requires more effort (but lots more fun!) to make it work reliably.

Despite advances in digital signal processing and miniaturization, multi-track field recording hardware still remains prohibitively expensive, and makes sense only for business entities, such as news organizations, motion picture studios, etc. For all others, renting is still an option, especially if you live in a large metropolitan area or have access to a good university media department and/or independent film makers' studio.

What is multi-track recording?

Most of the readers probably know what multi-track means but just to clarify the concept, let me define it as a technique of capturing more than two simultaneous audio streams and storing each data stream from each analog channel as a separate (i.e., unmixed) digital audio file. In other words, recording a conversation with four microphones onto a two-channel stereo recorder does not meet the definition, hence such scenarios will be left out of the present discussion.

Why use multi-track recording?

The answer is simple: each time we want to capture more than two audio streams simultaneously, we should use a multi-track recorder. If our research goals call for a detailed analysis of these audio streams, we do need to keep those channels separate. Multiple voices, jumbled up in a stereo mix simply would not work.

By a very similar rationale, digital archivists and language preservation scholars are interested in storing individual voices in their unaltered, pristine form. Most scholarly archiving standards and best practices for audio digitization (e.g., think of those millions of hours of radio and TV shows that still exist on analog tape, stored in the basements of university libraries) strongly recommend using multiple tracks of audio (e.g., NINCH, EMELD, the Open Language Archives Community, Historical Voices, and others).

Other reasons for multi-track recording include the ability to do "proper" mixes in post production. It is particularly useful if we want to publish the recording as a podcast, or as part of a motion picture, a radio show, or even a conference presentation. Many modern 5.1 surround mixing standards for "new" media such as high-definition TV, the Blu-ray Disc standard, etc., are easiest to implement if the material originated as multiple audio files.

Scenario #1: Expensive but good

Scenario #1 is probably the right way to do multi-track recording in the field. Because I typically discuss one or two-channel location recording scenarios, simply take all the information included in the "Field recording" section of this website, and multiply it by the number of channels you are going to need. It's that simple! Right away, you will notice that your cost has gone up, as have size, weight, complexity, and power requirements. Even the sheer physical effort needed has gone through the roof. Do not despair, though. There are some great tools that are sure to improve the quality of your work, provided, of course, that you can afford them. Some of the most popular brands and models include:

If you have the budget to afford one of these, I would also recommend that you use exactly the same microphones and cables, so that you can truly take advantage of the system's capabilities and get a consistent performance out of it. Figure 1 shows a schematic of a typical signal routing schemes.

Multi-track recording

Figure 1. Multi-track signal routing

I am exaggerating a bit

Well, I am guilty of exaggerating a bit. You should be able to find an affordable four-track digital recorder, just don't expect it to work as well as a Nagra or a Diva. Edirol has a respectable line-up of four-track recorders, including the R-44, which uses SD solid state cards and has 4 balanced XLR inputs. Battery life is decent (4 hours or recording time) and the unit promises decent performance (I haven't tested it myself). If you visit your local Guitar Center, you will find a few different portable multi-track recorders, but they're designed primarily for aspiring young musicians, so they are not equipped with the robust features we need in the field.

Scenario #2: The hack

I will be honest with you. The last few times I did proper multi-track audio recording was in 2002 in a professional recording studio and during the digital taping (yes, we used ADAT tape!) of a live show that same year. A year or so later, I did a project for which I rented multi-track equipment, but it was just too physically demanding for me to use that stuff, so I gave up on the idea and haven't done any of it since. If I were to do multi-track recording again, I would probably use the hack I am going to describe in this section.

More people, more trouble

As you know, dealing with people in an interview situation requires a lot of so-called "people skills." You certainly want to be respectful, accommodating. You need to come prepared, you want to be unobtrusive. You need to give people the freedom of movement, allow them to feel comfortable. Seating four or more people around a multi-track recorder, tethered, will make it difficult for you to achieve these goals, especially if you work alone (as most of us do). Wireless technology might be a solution for some, but, in my experience, it tends to complicate things even more (see the bottom paragraph of this page).

A possible alternative

I was going to call this alternative "hack-a-track," but that's a bit corny. Suppose you need to interview a small group of people engaged in social interaction, such as a family event, or a holiday ritual. People very rarely just sit still, so you would need a full-fledged wireless system to handle this type of situation. Instead, you can try giving each person a small digital recorder, and a lavalier microphone (or even a lightweight headset). Clip the microphone on to their clothing and attach the recorder to a belt clip, arm band, put it in a pocket, or use any of the wide variety of carrying accessories available for MP3 players. In fact, for your subjects, it would be almost exactly like wearing an iPod but with a clip-on microphone instead of earphones.

Next, push the "record" button on each recorder at about the same time, set recording levels ("auto" should work just fine), and activate the "hold" function so people can't accidentally change the settings or cancel recording alltogether. Optionally, you can clap your hands once, loudly, to create a audible marker in the recording, which will help you synchronize the audio files created by each individual recorder later, in post production.

At the end of the session, collect the recorders and download the newly created audio files to your computer for analysis and/or editing. This kind of set up has many of the advantages of a wireless system, but at a lower price and with less hassle. There are many recorder and microphone options in this market segment. I would recommend a small CF card recorder with a 1/8-inch (3.5 mm) stereo microphone input (with plug-n power) and a self-powered (or plug-in power compatible) lavalier microphone (omnidirectional is better if you plan to do acoustic analysis of speech) with a 1/8-inch (3.5 mm) connector. I, personally, would use a mono microphone, but you can use a stereo one as well, as long as it matches your recorder properly. For instance, if you have a recorder with a plug-in power microphone input, you will be able to match it with a stereo lavalier microphone that requires plug-in power. Just check your equipment spec sheets, or send me an email and I can help you figure it out.

I think that "hack-a-track" is a "hack of a solution" and that it might give you surprisingly good results (Sorry, I couldn't help myself). The following recorder-microphone pair should work quite well: