Creating great songs starts with recording great tracks—and here’s how.
By Joe Albano
Before you spend long hours in front of your DAW clicking on plug-ins and editing tools, you have to capture the sounds that make up your arrangement. Tracking is mainly done “out of the box;” while there are modern considerations in computer rigs, many aspects of tracking haven’t changed in decades. Let’s take a look at what comes into play during that all-important first stage.
The big questions
Every recordist will have to deal with certain issues regarding tracking, no matter how large or small the studio. The two biggest ones, at least to my mind, are:
~ To record all-at-once, or track-by-track (overdub)?
~ To use a click track, or not?
Both can have a big impact on the musicality of the finished recording. There are no right or wrong answers, but these issues will rear their heads at some point, and it’s worth spending time considering them. Later we’ll go over the usual technical considerations, like setup and monitoring, but let’s start with these issues of overall approach first.
When is a band not a band?
Ever since the introduction of multitrack recording, the option to use overdubbing—recording a song piecemeal, one track at a time, as opposed to capturing all the performers playing together—has presented artists and producers with some thorny decisions. There are pros and cons to both methods, of course, so let’s pick up there.
The traditional way to record an ensemble of musicians was always to set them up in a room, mike everybody up, and record the group’s performance(s) as they played, preserving all the nuances of their musical interactions. There’s still no better way to capture a group performance, but it’s not always a viable option for a variety of musical and technical reasons.
There may not be a suitable space available or affordable; there may not be enough musicians available for the session(s); the arrangements may not be finalized at the time the recording process begins; the equipment (mics, console, headphones) may not be sufficient; artists talented enough to record themselves as a one-man band may prefer to perform all the parts; and the many artists and songwriters on a tight budget may simply need to do as much as possible by themselves. Despite all these good reasons to use part-by-part overdubbing in place of the live ensemble approach, very often the goal is still for the finished recording to sound like a live performance. And with a little care, it can, as we’ll discuss below.
Overdubbing track-by-track offers its own advantages. Zeroing in on just one instrument at a time (or two, as in the case of simultaneously tracking drums and bass) allows the artist and producer to focus without distraction, with attention to detail, both musical and technical, that can add that subtle something extra to the finished product.
Since there’ll be little or no leakage on overdubbed tracks, it’ll be easier to edit or comp together the recordings later. And, of course, that’s another advantage—other than using closed-ear headphones, there’s little need to deal with leakage, allowing for better control when processing the tracks in the mix.
Overdubbing allows artist, performer, and producer to work together more closely, which can enhance communication, and sometimes offer a more friendly (read: less intimidating) environment for some performers. Everybody may feel more free to offer opinions and suggestions than they would in front of a larger group of players. On the other side of the coin, if a part isn’t happening, sometimes musical interaction with other performers may pull a better idea or performance out of a struggling player.
Mixing it up
In practice, it’s logical and common to combine the two approaches. An initial ensemble recording is done, with anything from just the core rhythm section—drums, bass, rhythm guitar/keyboard—to the full ensemble, sometimes including a vocal (or at least a scratch vocal, to give the other players a better musical context). But then many of the parts are replaced by overdubs, one by one, to fine-tune the arrangement and fuss over details in a way that they can’t be in a live, all-in take. It’s not uncommon for the focus of the initial live session to be getting just the drums (and maybe bass) down as “keepers”, with the assumption that everything else will be replayed piecemeal. This is a kind of “best of both worlds” approach—get a good ensemble feel for that all-important rhythm section, and then add in the additional arrangement elements around it.
This comes back to my earlier comment that there are ways to make even a recording of track-by-track overdubs still sound like a live ensemble. Getting an actual live performance from the key rhythmic elements, and then building the rest of the arrangement around that performance, is a good way to have your cake and eat it too. But this can be applied to songs even where there is no initial live performance, like the one-man-band scenario.
If I need to record a tune where I have to perform all the musical parts, I’ll initially get down a skeleton reference—sometimes drums, but more often a guitar or keyboard part that can define the various sections of the song more clearly. If there’s a vocal, I like to have a scratch track of that at this point as well, if possible. Then I’ll overdub drums and bass, and begin building up the rest of the arrangement. As the tracks get fleshed out, I can revisit some or all of the previous recordings while listening to the more fully-formed arrangement and re-track some or all of them. I can react musically to the subtle little licks and twists in the existing tracks, just as I would if I were part of a live ensemble of musicians listening and reacting to each other as they play. Obviously, this is time- and labor-intensive, but the results are well worth it—the resulting track can have much more of that elusive “ensemble feel,” even if it’s only one or two players doing all the work.
Of course, overdubbing parts can introduce timing issues that must be dealt with. Playing along with a pre-recorded part—no matter how well that part was played, and how good the musical “feel”—is not the same as playing along with another musician live. The recorded part can’t react to what you do; it plows along and you must follow.
While some musicians play along with overdubbed parts effortlessly, others struggle against the unyielding performances already recorded, despite being able to perform very well with live musicians. And the most unyielding of all, cold and musically unfeeling to boot, is the dreaded click track!
To click or not to click?
In any recording scenario, live or overdubbed, you’ll have to confront the issue of whether or not to make use of a click track. Most DAWs (with the possible exception of Pro Tools) assume you’ll be playing to a metronome click, and most of the cool digital editing capabilities we take for granted nowadays—rearrangement, quick comping, time compression/expansion, MIDI quantizing—will require the music to be referenced to the Session Tempo via the click, or at least perform better when it is. But most musicians can be very colorful when describing how they hate playing to a click, as it can suck all the vibe out of a performance.
Still, many recordings make use of a click, sometimes despite the musicians’ distaste for it, simply because of all the creative options it facilitates down the line. If you’re producing a session, you’ll have to decide when to use one and when not to (and deal with the consequences later).
I learned to play piano years ago with a wooden metronome clicking away, but I found that recording to a click was a new skill that had to be specifically practiced. No amount of ensemble playing will substitute for it. I’ve seen many sessions where a performer or a whole band assures me that, “yeah, sure, no problem, I/we can play to a click, no sweat”… only to discover in short order that they couldn’t.
From that experience, I learned that as a producer or recordist, you just can’t force the issue. If a performer can’t do it, they can’t do it, and no amount of “bearing down” or “focusing” will keep them in good time. Sometimes, you just have to let them play in free time (rubato), and then match their performance to the Session Tempo after the fact. This is done either with beat-mapping (creating a changing tempo map to match their timing), or by forcing their timing to match the (fixed) Session Tempo through quantizing or time-stretching tools. These capabilities are fantastic, but they’re not as quick and easy to implement as we’d like to believe. If you elect to use them, you’ll need to factor in extra time in the session to do so.
That said, there are a few tricks to try when it comes to the click. In a live ensemble, try feeding the click track to only the drummer and letting everybody else follow his timing. This can often work—many drummers can stay with a click reasonably well, and the other players are used to following the drums as a timing reference. However, if it’s one of those bands where some other player—bandleader, bass player, rhythm guitarist—is really the one driving the ensemble’s timing, then you’ll have to adapt.
If a player is having trouble playing to a click while putting down the first track, and it’s not a drum part (maybe a rhythm guitar or keyboard), it can be very helpful to substitute a simple quantized drum pattern. An actual backbeat seems to be easier for many musicians to follow, even ones who have inordinate trouble with the steady clicking of a simple metronome sound. Just make sure that the pattern is not at odds with whatever the real drum part will be doing when it’s eventually added, otherwise you may end up with some musically clashing rhythms.
When laying down overdubs, as soon as a suitable rhythm part has been recorded, get the click out of there! Let everyone play against a real performance—it’s still not a lock, but they’ll do much better, guaranteed. Just make sure that if there are pauses in the track that’s serving as the timing reference, you ensure that something comes in to provide time in those sections.
Now we can turn to the procedural aspects of tracking. With any approach, there are a variety of issues that must be dealt with when tracking parts.
Setting up for tracking a live ensemble, assuming there’s an adequate space, can range from fairly straight-forward to dauntingly complicated. One key consideration is avoiding leakage—unwanted sound from one or more instruments leaking into other instruments’ mics. The solutions for this will depend on the configuration of the available space. For the most effective isolation, if the space offers extra small booths, the players and their instruments can be physically isolated from each other. Just make sure they can see one another, or you’ll lose half the benefit of recording everyone together!
Sometimes, either the loudest or the softest instrument of the group is the only one that needs isolation. Drums, for example, will leak into everybody else’s mics, and piano mics will pick up everything else in the room. Isolating either, when possible, will help to avoid the worst of the leakage. Since loud guitar amp cabinets can be physically separated from the performers, they can be placed in iso booths, purpose-built amp isolation enclosures, or even just in closets, to control their sound. Since scratch vocals are often replaced with different performances later, the singer should track in an iso booth if feasible.
Short of complete isolation, instruments in a room can be partially isolated with baffles or gobos (go-betweens), which are placed around each instrument to contain its sound as much as possible. This, combined with the usual close-miking techniques employed in any modern recording, should offer reasonably good isolation for most instruments in most situations.
If it’s expected that some of the parts going down in the initial live session will be replaced with overdubs, then it’s especially important to achieve reasonably good isolation—listen to some of those “phantom” guitar solos on some of the really old Beatles tracks for a good example of what may happen otherwise!
Of course, if the goal is to avoid overdubbing and use all the parts played live, then you can be a lot more forgiving about leakage. While it does take away some editing control over the individual tracks, it’s by no means a bad thing. As long as you’re careful to avoid unpleasant phasing (observe the 3-to-1 miking rule, and check all the mics in combinations for any weird phase issues before recording), then a little leakage can be a good thing. It can add air and depth to the overall recording, requiring less artificial ambience at the mixing stage, and provide a nice sense of three-dimensionality.
Live or die by the monitor mix
One aspect of tracking that can be either simple or complicated is monitoring. I’ve often seen engineers be really haphazard with this, quickly throwing up a few phones at the end of the setup, but the headphone monitor mix provided to the musicians as they perform is a really important aspect of any recording session, from live ensembles to individual overdubs. To avoid sound leaking from the headphones into the mics, closed-ear headphones are always used. They can get a little hot after a while, but the isolation they provide is absolutely necessary to avoid that high-frequency tst-tst-tst from getting to the mics (you know the sound I mean!). Any model that seals well and plays loud enough to be heard over nearby brass players will work.
At a gig, players hear each other in the room and don’t need phones. But with the use of iso booths and baffles, not to mention that some instruments—bass, synthesizers, guitar processors—may be recorded direct into the console and not be making any sound in the room at all, headphones are a necessary part of a live session. And for overdubs, naturally, the players need to hear a feed of the previously recorded tracks, as well as their own instrument/voice. Since everything they play is based on what they hear in the phones, the musicians really need a monitor mix that provides them with a reference that allows them to achieve their best performance. If they can’t hear a key part, or themselves, well enough in the phones, their performance will suffer—and so will the finished track.
Sometimes musicians won’t call attention to problems with their monitor mixes. Maybe they don’t want to be seen as “complaining”, or they’re intimidated by the exposure of their instrument or voice in relative isolation. Problems like level imbalances, or a dodgy cable, or excessive latency (digital delay from the DAW) may go unmentioned and unfixed, impacting the performance. That’s why it’s important for the engineer to check each headphone for two things: technical issues, like those I just mentioned, and balance issues.
For the second, you’ll need to ask the performers what they want in their phones—but you’ll also have to use your judgment as to what they need in their phones. I’ve had performers who wanted so much reverb on their voices that it threw off their pitch, or who wanted themselves so loud in the phones that they couldn’t hear that they were drifting out of time with the tracks they were playing to. If these situations come up, the producer should be the one to deal with them, gently coaxing the artist to come around to what you know they’ll need to do their best work.
Once the recording begins, you need to keep an eye (and ear) on the monitor mixes. Many vocalists like to remove one earcup as they sing—it helps them stay on pitch, but sound leaking out of the unused earcup will get into the mic and may ruin the recording. If you see them do it, you can quickly kill the feed to that side, possibly saving the take.
Setting up the cans
And that brings me to the last monitoring issue: setting up the monitor mix itself. There are many variations of this. In the simplest setup, you’d use the same Aux Send (e.g. Send 1) from each channel of the board or DAW mixer to create an independent musicians’ monitor mix, send them (pre-fader, so level tweaks made to the Channel Faders during a take won’t distract the players) to an Aux channel, and send that mix out an extra stereo output to a headphone amp. If it’s an overdub session with one player at a time, you’re good to go. If it’s a session with several musicians, then there are further considerations.
You could take that one monitor mix and mult it to several players. Ideally, this would be done via a multi-headphone amp with separate connections and volume levels for each phone, so each player can have a comfortable level in his/her ears. If the phones must be multed from the same amp, with only a single volume control, then at least try to use matching headphones—otherwise, some will likely be much louder than others, and everyone will be unhappy.
In a more complex setup, different players may need different monitor mixes in their phones. To accomplish this, you’ll need to use several different Sends and Aux Channels, and if you’re doing it in a DAW, the interface will need to have multiple outputs. For example, if you have eight outs, two can be used for a stereo control-room mix, and the other six for individual mono mixes (or three stereo mixes). Stereo monitor mixes are nice for separating the player’s own instrument from the other/backing tracks, but mono is fine, as long as everything can be heard clearly.
I often create monitor mixes on a small analog mixer, independent of the DAW. I feed a mix of any pre-recorded DAW tracks to that mixer and run or split all the live mics to it, using the 4-6 Aux Sends there to create independent monitor mixes. This has two benefits—it provides true zero-latency monitoring for the live instruments/mics (one less detail to worry about), and, since that mixer can be off to the side, it lets me delegate the constant back-and-forth talkback chatter about monitor levels and mixes to the assistant engineer, so I can focus on the main recording.
Bigger studios often use multi-monitor systems that put mix controls for the monitor mixes in the hands of the individual players, via small boxes clipped to their mic stands. These systems, while extremely convenient, can be costly, and they can give artists too much control over what they want vs. what they need.
Once this setup has been taken care of, with mics and baffles well positioned, headphones and monitor mixes at the ready, and an appropriate tracking strategy is in place—live ensemble or overdubs, or both—then you’re ready to start recording. There are other articles in our Resource Library that touch on some of the other obvious technical aspects of recording, like setting levels and whether to use compression on the way in, etc. This article was more about dealing with issues and approaches to tracking; I hope it’s been of some help to those who want to be sure they’re really ready to count off the song and hit that big red button.