I’m a member of the band Gert. Until we played together in person this summer, our year-long collaboration was entirely virtual. 6 song writers, a continent apart, connected by musical tastes and the Internet. We’re still a band in the general sense, but in place of schedule conflicts, angry neighbours, and ego clashes, we deal with time zones, bit rates, and ego clashes.

The 6 of us have played in bands for years, but most of our past experience isn’t relevant to creating music virtually. Collaborating with 3000 miles between band mates has little in common with traditional jamming. Though that’s not always a bad thing, as we’ve discovered in the last year. On one hand, music – especially for performers – is interactive. With some limited exceptions, the Internet does not facilitate real-time musical interaction, so musicians who rely on visual or audible cues are at a disadvantage. But on the other hand, creativity is usually a personal and private experience. Many people feel liberated with no band mates or groupies in the room, which in turn opens up creative avenues not possible with the traditional approach.

In Gert, we’ve learned to work around the obstacles to our creativity, and in fact we’ve capitalized on the benefits the Internet affords us to create some fantastic music. Below, I provide insight into what it takes musicians in different cities to write and record a song together. I also describe in detail how we wrote Sweet Lovin’ Woman (I Hate You), with samples of the song as it progressed, and the discussion around the choices we made while writing.

If you’ve considered online collaboration yourself, or are simply curious about how distributed creativity works, read on.

In our year working together, we’ve refined our process for remote collaboration, both technically and creatively.

The technical details are straightforward:

  • We use a phpBB board for discussion (visit Gert Mansion.) Though we could just as easily use email, we prefer the bulletin board because it keeps things centralized.
  • We email MP3 versions of rough ideas. Since these ideas are only for discussion, quality isn’t important.
  • We work to a click-track to keep things synched. To ensure that the raw tracks line up as intended, we agree on the song’s tempo beforehand, and include a bar or two of click-track at the start of each WAV file, in case a manual synch is needed.
  • We collect raw tracks on a shared FTP server. When quality does matter, specifically for the individual tracks that make up the final mix, we use a shared FTP account and transfer 24 bit/44.1 KHz WAV files. Since none of us use the same DAW (collectively, we use Sonar, Cubase, and Digital Performer, on PCs and Macs,) we avoid program-specific formats and use raw WAV files for everything.

I say this is straightforward because most musicians collaborating online will, from necessity, learn these details quickly. (With the possible exception of the click-track, though it may not be important to everyone. In Gert, we realized early that we work chaotically, throwing parts at the wall to see what sticks. Without the luxury of first recording drums and bass as bed tracks, which is the traditional “band in the flesh” approach, we couldn’t keep things coherent without a click-track.)

The creative side of our process is more complicated. Like all creative activities, it varies somewhat each time we do it. However, after collaborating on 11 songs, we’ve learned what works and what doesn’t:

  • Every songs start with an idea from one of us, usually a guitar riff with simple accompaniment (or in the case of Sweet Lovin Woman’, a bass loop.) Generally, we propose ideas via email with a quickly-recorded scratch track. However, this doesn’t guarantee the band will immediately work on the new song, as we have 3 or 4 ideas on the table at any time. We don’t make real progress on a song until one of us arranges a structure. (More on this below.)
  • Each member of Gert has presented ideas. We write by collaboration. While this may not work for all bands (for example, many depend on a primary song writer for ideas,) Gert has 6 musicians with solo catalogs and egos.
  • We don’t present “finished” ideas. Since we aim to collaborate, we’ve found it best to build the song collectively rather than rely on a single songwriter as architect.
  • We agree on a structure. The ball doesn’t get rolling on our songs until someone is inspired to arrange a proper structure. Mostly, this involves deciding how the song should flow: Will there be a chorus? 3 or 4 verses? 16 bars of guitar wankery, or should we let Mike sing a little longer? You can see this clearly below, where I take Paul’s idea, propose a chorus for it, and map out the song’s sections.
  • One person volunteers to mix. There’s a technical reason for this (i.e. it’s not possible for multiple people in different cities to simultaneously mix the song,) but it’s also a creative requirement. We need a designated mixer because, while everyone in Gert maintains their creative freedom, we recognize that someone has to make final decisions. Sometimes a riff or harmony just doesn’t fit, so we have the mixer to serve as director (and occasionally, benevolent dictator … remember: Egos!) In the example below, I played this role.
  • We write … This is the meat of Gert’s creative process. For a few days, we enjoy a creative free-for-all. Anything goes: guitar riffs, lyrics, melodies, rhythms, harmonies. All 6 members of the band sing and play guitar, and 3 of us are drummers, so there are no designated instruments (though for consistency between songs, Tom always records the final drum track.) Ideally, as the song builds, we play off each other’s ideas with the bigger picture in mind, so each new part adds something to the song. When we’re done with our “audio brainstorming,” the mixer decides which pieces work best together, and starts on the final mix.
  • We all approve the final mix. A second (or in our case, fifth) set of ears will hear nuances that the mixer’s ears glossed over. And since we all monitor through different equipment, feedback from the band is the virtual equivalent of checking how our mix translates to 5 different systems.

To show how this works in practice, I’ve assembled an audio chronology of our process for the song Sweet Lovin’ Woman. If your browser supports Flash you can hear the tracks below using the “play” button beside each sample, or you can stream/download an MP3 via the “mp3″ link.

The “discussion” box that follows each sound file expands to show some of the emails we exchanged while working with the idea at that stage. Click on this box to see how we interact while collaborating.

If you’re not familiar with the song, you might enjoy following its creation before listening to the final version.


Paul started us off with a simple loop idea. His original track sounds nothing like the final song, though it does establish the bass line which became the song’s main instrumental hook.


Discussion (click to display)

Paul’s idea didn’t immediately inspire anyone else, so it sat neglected for a few days. This is actually the norm with our songwriting efforts, and it happens not because the initial ideas are bad, but because of the fickle nature of creativity. Writers rely on a creative spark, and until an idea manages to strike that spark, there’s not much we can do with it. In this case, Paul laid an ad-libbed vocal (though much of it remains in the final version) over top of his original idea, lighting a fire under our creativity and the debate that accompanies it.

Lively debate on how to proceed is a sure sign that an idea resonates with the band. The main topic of debate here was whether we should add a chorus section or leave this track as a “jam.”

Discussion (click to display)

The mixer is charged with discarding pieces that don’t fit the final mix, but we also drop concepts as we write. When it’s obvious something doesn’t fit, or would take us in the wrong direction, we put the idea aside. In this example, Rich proposed a Radiohead-esque vocal line that wasn’t so well-received.

Discussion (click to display)

We hadn’t agreed yet on whether or not the song needed a change – be it a chorus or bridge – but as I listened to Paul’s idea I heard what seemed like a natural change. So I recorded my thoughts, and stitched them together with Paul’s idea. (Note: The song sounds significantly different here because I added a drum machine to tie our parts together.)

Along with my recording, I mapped out how I thought the song should flow, start to finish. You can see this in the discussion below. I went with a standard rock song arrangement, with verses, choruses, and a solo break. The mp3 of this idea that I shared with the band ran the full 96 bars so others could use it as a backing track for their subsequent ideas.

Discussion (click to display)

The next few samples show what I described as the meat of our writing process. The song evolves iteratively, with each of us trying ideas and adding parts to previous mixes. In this sample, Rob added some keyboard noodling and a percussion riff between the first two verses.

At this point, I also volunteered to mix the final track, as Tom (who originally wanted a crack at it) was too busy. The mixer plays a creative role, and each of us is qualified for the job, so we take turns at mixing. Tom did the next song.

Discussion (click to display)

Here, Rich added a “porno” guitar riff (or “Shaft” riff if you prefer.) More importantly, he presented a vocal for the chorus section that we agreed unanimously was perfect. Given the importance of the chorus as hook in most rock songs, it might seem odd to write the chorus in the middle of the process, almost as an afterthought. However, that’s one of our favorite aspects of this approach. If something doesn’t work, we drop it, and when something works as well as Rich’s chorus, it stands out!

(This sample also illustrates our lack of concern for audio quality while we share ideas. Rich didn’t waste time balancing his mix, because his simple “faders-up” mix conveyed all we needed to hear.)

Discussion (click to display)

Next, Rob added some ebow’d guitar. This became my favorite moment in the song, because of the way it carries the listener into the second verse.

Discussion (click to display)

This sample features Paul’s re-tracked vocals (with no ad-libs this time,) and Tom’s first contribution to the track (with lots of ad-libs.) His drum part is clearly a jam, but hearing his thoughts let us give him the feedback he needed to nail the part.

Discussion (click to display)

Here, Mike added his distinctive voice to the song’s intro.

Discussion (click to display)

As we near the end of our writing process, Tom needs a drum-free mix to record his final drum take with. The drum-free mix makes it easier for him to follow changes and accents, and the bass guitar.

The sample here features the final versions of all parts (except the guitar solos,) including some harmonies that Rich and I added.

Discussion (click to display)

And here is the final mix, with Tom’s “keeper” drum performance. I made a few changes as I mixed, adding a second guitar solo in the break section, and an extra beat at the end of the intro to help the first verse’s impact.

Discussion (click to display)

We were thrilled with how this track turned out. And 9 successful collaborations later, Sweet Lovin’ Woman is still my favorite Gert song, because we clicked so well creatively while writing it. (You can hear our other work on Gert’s web site.) Even though geography separates us, we’ve found, in our virtual creative process, a way to connect and create some great music.


For more articles on collaboration and home recording in general,
Subscribe to the Hometracked feed.

Tags: ,