The Trouble with Overdubs

I suppose that if you don’t record music, this post might not mean very much to you.  It’s about a technique called “overdubbing”.  Overdubbing is the process of recording something once, and then making another recording alongside of the first, by playing along with the first recording and at the same time, recording what is being played now.  The second recording is called an overdub.  You can do that for as long as you have vacant tracks on a multi-track tape machine, or to the limit of the performance of your computer and software, if you have a Digital Audio Workstation (DAW).

Overdubbing is an exceedingly common technique, in music production.  Very rarely are even live performances published without further overdubs, these days.  Overdubs are used to correct mistakes (“drop-ins”) or to fatten up tracks, by doubling or even tripling the parts.  Lead vocals are very frequently “double tracked” through overdubbing the lead singer, to give the vocal more prominence.  Backing vocals are almost always overdubbed, to even out the choral effect.

In home and project studios, some tracks are constructed almost entirely from overdubs.  A single part is played by the lone musician, then the next, then the next and so on.  It produces a particular sonic aesthetic, in my view.

There are two very effective experiments that can show the effect on performance of overdubbing, versus live interaction between musicians that are playing ensemble.

Experiment one:  Take a band that has three people in it that are capable of playing guitar and begin playing twelve bar blues as an improvisational jam session; with one of the guitarists playing the bass instead (you can have drums).  The rules are that the bass sound cannot be changed.  No tweaking of the amp settings, the effects, or the volume and tone settings on the bass guitar is permitted.  Record the result, as played live.  Now, have one of the guitarists swap his guitar for the bass, with the bass player.  Now play the same improvisational piece again and record it.  Finally, let the remaining guitar player be the bass player, while the other two play guitar and play the same improvisational piece again, recording a third version.  Now listen back, paying attention to the sound and prominence of the bass guitar part.  I guarantee that the three pieces will sound extraordinarily different to one another and that if you listen to the bass parts on the three recordings, all recorded on the same guitar with the same amplifier and effects settings, the actual part will sound very different in each of the three cuts.  The bass will have been played entirely differently by each of the three guitarists.  Some will have played hard, some soft, some with lots of attack and some with a softer tone.  All of this sonic difference comes from their hands and minds alone.  They will have been playing in reaction to what the rest of the band is presently playing and attempting to shape the sound of the ensemble in their own way.

Experiment two:  Choose a song and rehearse it, in a rehearsal studio, recording the result.  Now play it in the same rehearsal space, but this time with girlfriends or people you want to impress in the same room.  Record that version as number two.  Now record the same song played on stage, live in from of an audience.  The fourth recording will be one recorded in a recording studio, as a band.  Finally, recording five should be made in a recording studio, but with each part recorded individually, while no other musicians are playing, as a series of overdubs.  In other words, record the drums, then the bass, etc, with each part played in isolation.  Listen back to the five versions of the same song, played by the same musicians (usually on the same instruments).  You will notice very quickly that each of the five versions has an entirely different feel to it.  Some will be exciting and raw, some clean and clinical, some less engaged and bored and others full of mistakes but sizzling with life.  Notice that the fifth version, while perhaps note perfect, probably has the least life, dynamics, feel and interplay present in it.  It’s the flattest and most sterile.

This is the trouble with overdubs:  It might be the cleanest way to capture an instrumental part with full audio isolation, or the only way available to make a track, if you do not have a band, but something important is lost, in the process.  It might also be the best way to correct bum notes and mistakes made while playing the part.  You might be able to construct one entire longitudinal part by dropping in, phrase-by-phrase, given a lack of ability to play the song in one go, end to end.  But it doesn’t sound the same as if the band had been playing, to people they want to impress, in an exciting atmosphere.  It just never does.

What makes least sense of all is to have a band, but to never rehearse the song or play it live, before taking it into the studio and recording it, part-by-part, never playing it at all as a band.  Writing songs in that way, in the studio, part-by-part, is an even more impoverished experience, in prospect.

That’s all very well, but there are lots of examples of great music that was made in a DAW, or multi-track recording studio, by overdubbing extensively.  If you think about those rich, beloved Queen harmonies, or Brian May’s guitars, these sounds couldn’t have existed at all, without overdubbing.  Mike Oldfield’s “Tubular Bells” couldn’t have been made, without them.  Boston’s “More Than a Feeling” was heavily overdubbed and remains much loved.  Gotye’s work is constructed, track-by-track, but is very popular.  Let’s not also forget that the inventor of multi-track overdubbing, Les Paul, couldn’t have produced the lively, space age hits of Les Paul and Mary Ford, without the ability to overdub.  Overdubbing is ubiquitous.

There are many aspects to a musical performance, or ensemble of individual musical performances, that contribute to the dynamics, contour, emotional impact and passion of a song.  As a musician, you tend to play with more intensity and sensitivity, when others are playing at the same time.  You also play to balance the overall sound and to match the mood of the other players’ musical outpourings.  There’s something worth preserving, in all of that.  How do you capture or even simulate those things, when overdubbing?

A band is an interactive thing.  Band members interact.  There is a lot of unspoken and unwritten give and take going on.  A DAW, on the other hand, doesn’t interact with you much at all.  The backing track you are recording against is fixed and doesn’t vary or respond to what you are playing at all.  You give, it takes.  I guess that’s why they call individual overdubs “takes”, right?

When you are using overdubs to remove or blur harsh edges, you can get them to “flatten” the music, in some sense.  It’s like polishing a stone.  You can make the music devoid of passion, smooth, slick and perfect, if that’s the effect you are trying to achieve.  And many people like that effect.  On the other hand, to other ears, it can sound artificial, overworked, safe, bland, blunt and lacking any form of risk.  In short, it can be very unexciting.

So how do you overdub, but keep the music lively, exciting, edgy and interesting?  Is it even possible?  There are examples of overdubbed music where, arguably, the musicians and producer have achieved a happy compromise between overdubbing and producing great music.  What are some of the things that you can do to get that outcome?

Here are some suggested solutions:

1)       Record as a band first, then overdub only what you absolutely have to.  The Beatles did it this way.  They recorded a number of takes, playing the song as a band, and then “sweetened” the best take with overdubs, to create the released version.  This, if course, assumes you have a band, a room big enough to record it in, with acoustics good enough to permit isolation of each track you record, even though the band is playing at once.  In this approach, overdubbing is used as a condiment, rather than an ingredient.  It produces a pretty good compromise.

2)      If you don’t have a band, you could record a solo guide track, as a performance, to give the song its form and shape.  If you treat it like a performance, rather than a solo recording, you can retain some of the passion, especially if you wrote the song.  You can even record with an audience, if this is practical and it helps.  This allows other musicians and/or instruments to be added as overdubs and for the guide track to be entirely replaced, in the end, but the song retains some of the character of the first performance captured.  What makes this difficult is that if you use a DAW and don’t record to the metronome (because the metronome can be the very thing that squashes the life out of the performance), then today’s DAWs can make it exceedingly difficult to add MIDI parts and tempo changes, because the first recording will not be synced in any way to the DAW’s own grid, tempo or downbeat positions.  DAWs are not that great at working out where your downbeats are, where your accents and stresses were, how your beats were actually spaced in the bar, or what your tempo and tempo changes were, across the length of the recorded piece.

3)      If you have no choice but to construct your music track-by-track, part-by-part, instrument-by-instrument, then record the parts, but when you are done, re-record the first parts again, only last.  That way, the song may have developed a form and shape, during its making and you can add more spice to the second performance of those earliest tracks you laid down.  In other words, you can improve the feel of the earliest tracks you recorded, by re-recording them, relative to the more finished work.  In some senses, you can “re-interact” with your own song.

4)      If you’re into painstaking detail, you can programme a MIDI part first, including all the song dynamics, tempo changes and the contour of the song’s development, in minute granularity.  Having constructed this MIDI part, you can then play in the overdubs, conformant with the MIDI part’s feel.  This is a slow and artificial way to work, but has the advantage of allowing easy additions of MIDI parts and audio edits, because the tempo, downbeats and MIDI timing grid are all well defined.  It has the disadvantage of not being very spontaneous, however.

5)      If your DAW has groove matching features (e.g. PreSonus Studio One Pro v2 – http://www.youtube.com/v/VYDJ_UpPFVY), play your music in, as a performance, then let the DAW create a groove that conforms to where you placed your beats and emphases, in each bar.  That also gives you the ability to match MIDI parts to the feel of your live performance and preserves the subtle variations in timing and beat placement that go into creating a good live feel, but it does little for the song’s contour and emotional development.  It’s better than nothing, though.

6)      Comping (i.e. composite take editing) and loop recording is the process of letting the DAW run, repeating a particular section of the track, and automatically and continuously recording your playing, each pass.  That way, you rapidly build up multiple takes, without stopping.  This gives you a chance to warm up and get into the groove.  After a few loops, you may be playing “right in the pocket”, just because you are now more rehearsed.  DAWs that have this feature let you edit a single take out of all of your takes, so you can choose some of the spontaneous bits in the earlier takes, if you want, while keeping the better executed sections from later takes.  It’s all done through editing.  A DAW that does this is Cubase 6.5 (http://www.youtube.com/watch?v=R1wJCvjPsNA&feature=related)

7)      When you play your overdubbed part, imagine you’re in the moment, playing the part of a live musician, as a character.  Using your acting abilities can help you forget that you’re in a quiet studio, trying to capture the perfect, mistake-free take and actually convince yourself that you’re on stage, in front of screaming fans, putting out your best playing.  This little Jedi mind trick can make a huge difference to the part that gets recorded.

8)      Fix it in the mix.  If you are playing self-consciously, to avoid any little mistake in timing or intonation, or if you are throwing away take after take, because of tiny fluffs and bum notes, consider keeping those overdubs anyway and using the DAW’s Celemony-alike pitch and timing editing features to tweak the occasional problem.  Otherwise, you might be increasing the frustration, with each take and getting no better a performance, as people concentrate on not screwing up, instead of on playing the most openly they can play.  (Here are three DAWs and their approach to this form of editing: http://www.youtube.com/watch?v=I0N_bNp4cv4,   http://www.youtube.com/watch?v=jbhQ7Ap5-TMhttp://www.youtube.com/watch?annotation_id=annotation_783634&feature=iv&src_vid=Q_VElRfTZ4k&v=MkpaM3fmYF0)

9)      Actively map out and plan the song’s contour, its high and low points, its breakdowns and tempo changes, on paper, ahead of recording the song.  Having this reference, made in pre-production, can be invaluable, in the studio.  At least it communicates to any collaborators what the song writer’s or producer’s intention is.  If you mark up the points of drama, resolution, tension, light and shade, at least musicians can be mindful of these points, during their overdubs and adjust their playing accordingly.

10)   Don’t get distracted by the mechanics of playing to sync and within the recording headroom available, when overdubbing your part.  Let the engineer take care of the levels.  If you are in a project studio, set up your recording channel to accommodate the likely peaks in your playing, ahead of time, or use a little compression and limiting on the channel.  The important thing is not to play in lock step sync and tune, with safe, constrained, characterless articulations; it’s to give a good performance that is near enough in timing and pitch.  Tiny mistakes or slight deviations from time or tune will actually add to the character of the overall song.

11)   Have a jam session, using the song you are about to record, before you record the release version.  Permit and encourage improvisation and record this jam session.  This gives you a reference recording that may have some novel performance elements in it, which you might want to recreate in the clean version.  You should pre-rehearse your song, even if you are a one man band, because this free form recording may capture some character elements you want to preserve, when it comes to recording the track for real.  If there are other musical collaborators to jam with, but who won’t appear on the finished track, you can borrow their musical ideas (with permission and acknowledgement!).  Time spent in pre-production always pays dividends in the studio, even if it is your own project studio and studio time costs you nothing.   If you record the pre-production jam session, you could even consider using it (or parts of it) as an erasable guide track, which can form the scaffolding of the song, as you record your overdubs, part-by-part.

12)   Most importantly, when you overdub, let go of it, when it’s good enough.  Take 37 is not going to be significantly better than take 36, typically, and usually it’s much more boring, lifeless and overworked.  By the time you get to take 37, you’re also bored, sick to death of the song, disinterested and thinking about being elsewhere.  This applies especially if you are a one man band, singer, songwriter, and self-producer.  Don’t let fatigue destroy the feel of the music, especially when takes 1 or 2 were just fine.

So, overdubbing is an essential musical production tool, but it has some nasty side effects that you need to mitigate.  When overdubbing, you can attempt to make the music too perfect and lose all of its character and charm, in the process.  Some of the rough edges give the music its texture.  It gives it an organic vulnerability and sounds authentic.  Making your overdubs in such a way that you preserve those aspects of your music is a delicate art, but one well worth developing.

As Mick Jagger is reported to have replied, when his band was criticised, by a recording engineer, for their inability to play in strict tempo, “The Rolling Stones is a band, not a clock”.

Advertisements

About tropicaltheartist

You can find out more about me here: https://michaeltopic.wordpress.com/. There aren’t many people that exist in that conjunction of art, design, science and engineering, but this is where I live. I am an artist, a musician, a designer, a creator, a scientist, a technologist, an innovator and an engineer and I have a genuine, deep passion for each field. Most importantly, I am able to see the connections and similarities between each field of intellectual endeavour and apply the lessons I learn in one discipline to my other disciplines. To me, they are all part of the same continuum of creativity. I write about what I know, through my blogs, in the hope that something I write will resonate with a reader and help them enjoy their own creative life more fully. I am, in summary, a highly creative individual, but with the ability to get things done efficiently. Not all of these skills are valued by the world at large, but I am who I am and this is me. The opinions stated here are my own and not necessarily the opinion or position of my employer.
This entry was posted in Uncategorized and tagged , , , , , , , , , , , , , , , , , , , , , , , , . Bookmark the permalink.

20 Responses to The Trouble with Overdubs

  1. There are some good ideas in here. Thanks.

    But honestly, I think your focus is too performance-centric. What I mean by that is that it seems you believe that the point of recording is to emulate what is/can be produced live, by a band, in a single take, at least in theory – assuming an entire band can all simultaneously give a “perfect performance” (with “perfect” being defined however you want – slight imperfections, string noise, voice cracking, flubbed snare hits etc. can all be included in that definition).

    My feeling is then – why not just record bands playing things? Why mess around with making studio music at all? Why aren’t all albums produced to sound like live albums?

    I think it’s because, for most people, the recorded version of a song IS the definitive version of that song. I think in the early days of recording people thought of the SONG as being the thing itself, and the recording, or any given performance of it, was just an emulation, a simulacrum of the thing that exists more as an idea, perfect entity. But I think that was a long time ago. Now the recorded song IS the song, and everything else, including live recorded versions, non-recorded performances, sheet music etc are just approximations of the real thing. Yes, people still qualify comments on songs sometimes by sayint “the album version,” but that’s usuallyjust for clarification, or to contrast it with some other incarnation. In mos people’s minds, the first recorded “album version” is the definitive version of a song.

    To me a studio is another instrument. Some people think of it as a way to pursue sonic perfection that just isn’t possible with live playing. Maybe. But even then you’re limniting yourself to what is possible (if only theoretically) to play live. And that’s a huge limitation, in my book. I tend to think of the recording/composition/studio process more as “how good/cool/awesome/interesting can I get this to sound?” Sometimes that happens to correlate pretty highly with a “live” feel, and other times it’s absolutely nothing like anything anyone could create by playing an instrument. To me that’s the ultimate goal, not makigj it sound “real.”

    I guess my point is – I don’t care if it doesn’t sound believable at all as four guys in a room rocking out. Seriously, who cares? I want it to sound GOOD. Whatever that means in terms of how REAL it sounds is irrelevant to me.

    Sorry if my comment here is tangential.

    • Actually, I use overdubs all the time. I have to. I play all the instruments on my own work. The point of the article was to acknowledge what is lost in the process of overdubbing, with the aim of putting as much of it back as you can, consciously, while overdubbing. If you don’t, then you definitely get a less affective musical result. My works is practically ALL overdubs 🙂

      Thank you for commenting. You put a lot of thought and passion into your comments and I like that. Do you have a blog of your own? If not, maybe you should. You have great perspectives to share.

    • And I’ve just re-found your blog. I’ve been on it before. It’s mostly on nutrition, right? I didn’t immediately connect you with musical blogging, so I apologise. Going to read more of your stuff 🙂

      • Read away! But I haven’t touched that blog in years, as you can tell. I said what I thought needed saying at the time and moved on.

        Like I said, I think you have a lot of really good advice and insight in here. The music I’m making actually does not use a lot of live performed parts at this point (but I’m moving more in that direction as my gear arsenal expands), but I think some of the same ideas still apply. Like learning when a take / part is already good enough, when continued reworking will have a low ROI. I’ve also been playing along to early versions of songs (sometimes just drum loops), and I find that I do tend to come up with things that I would not think to program in a sequencer or build in sample editing. I also plan to use multiple recordings / iterations of the same riff/part in an attempt to somewhat create the “live” feel of a part that is recorded “live” (as opposed to a 4-bar snippet that gets looped, for example). And I actually add in imperfections to my rhythms (percussion and melodic stuff alike) by design so that they don’t sound quantized.

        My previous comments were really more about getting at a much more expansive question. Should we simply be using computers / DAWs / overdubs as a way to record on a budget, as a way to emulate what people used to only be able to do in a studio on at least a modest budget? To record solo in lieu of having a band? Should we just be trying to achieve a sound that sounds “real” to untrained ears (it never ceases to amaze me how little the average music fan knows about how “fake” music production really is)? Or should we be trying to use computers / DAWs / overdubs to make music that is fundamentally different from what can actually be played by humans? I tend more to the latter. And by “can’t actually be played by humans” I don’t mean it should sound like space age alien techno, or some such futuristic unknown genre. I just mean that I tend to think it’s okay to step past that line where most people go “wait, what is that, how is this happening?” To basically snap people out of that willful suspension of disbelief where they can believe that everything they’re listening to is “real,” when in fact almost all modern music production is complete artifice. That is an effect that humans playing actual instruments are not good at, but it can be achieved by computers / DAWs / overdubs. I’ll leave this with a quote from Jonny Greenwood (when asked about recording Kid A), which sums up my feelings on the topic pretty well:

        “I don’t remember much time playing keyboards. It was more an obsession with sound, speakers, the whole artifice of recording. I see it like this: a voice into a microphone onto a tape, onto your CD, through your speakers is all as illusory and fake as any synthesizer—it doesn’t put Thom in your front room – but one is perceived as ‘real’ the other, somehow ‘unreal’ … It was just freeing to discard the notion of acoustic sounds being truer.”

      • Oh, I completely agree. A studio, and these days a DAW, is a valid instrument, to create with and to do what has never been done before. I also think that the creation of the song can be a separate process and that there are benefits from working this way, sometimes. My favourite way of working is to go so far, outside of the DAW, then go completely to town with the song, inside of it. I recognise that other people like to work in other ways, however. Thank you, once again, for commenting.

  2. That’s an interesting idea (working as far as you can outside of the DAW, then going in). I think that’s totally normal for a band to have that approach, but probably less common for someone like yourself that relies primarily on overdubs. I’ll have to try that once it actually becomes a feasible option.

    And I think I have a partial solution to the metronome problem you raise in point 2 above. I’ve had good luck with this method: create a simple beat in a DAW, then record a few loops worth of other percussion things to add in. I think things like shakers, tambourines and maracas are especially nice for this purpose, because they create a “wide beat.” That way, the accents and downbeats are not occurring with pinpoint accuracy as they do with a quantized beat in a DAW, so the “target beat” becomes wider. It’s like the difference between marking the target with a fine tip pen vs. a magic marker. The center of the line is quantized to the exact metronomic beat, but the percussion track makes the target beat (the magic marker line) wider. You have more freedom in where you place any single note, but you don’t slowly drift off grid as would be inevitable if not using some form or metronome or DAW. Most people feel WAY more comfortable playing to something like that than a metronome, and you preserve the ability to match everything up to the DAW time. It might be a little extra work, but it’s worth it if it can elicit a more natural performance.

    • That’s a good idea. I sometimes swipe a finished beat loop, from my library of loops. When those are played by excellent percussionists, it serves a similar purpose to your idea. I have literally thousands of loops that I have collected, over the years.

      • Yup, also a good idea. It could definitely have the same affect.

        I just have a general aversion to using other people’s sounds in my own work (excluding things like single-hit samples). But I have respect for people that can do it artfully. It’s just not what I want to do. So I want to preserve the option that the percussion track I create/record can potentially end up in the finished track.

  3. GH says:

    Listening to the Carpenters second album, “Close To You,” will give you a great example of just how amazing quality overdubbing can sound. In the song, “I’ll Never Fall In Love Again,” they lay down a 13-part, 39-voice chords that sound like a full choir singing. Not to mention that this was accomplished in 1969 & 1970, 46 years ago. In fact, all of the tracks on that album use overdubbing to perfection.

    • I love that album. I think the Carpenters were masters of the overdub. Richard Carpenter was highly influenced by the Les Paul and Mary Ford records of his youth, which are another superb example of the art of the overdub. Thank you for commenting.

    • Mike says:

      Richard Carpenter has never stopped tinkering with the mixes of Carpenters’ songs. Unless you know the source, you can’t be sure you’re listening to a mix from 1969 or 70. Most of their material was remixed and with some instruments re-recorded from the mid-1970s through the 1990s. Richard Carpenter doesn’t like the original versions to remain in print once he remixes an album.

  4. Michael, very useful reflections on challenges facing a solo artist desiring to deliver a live band feel via DAW and overdub. Some may question why would any one want to do that? Well, it’s a choice, that’s all 🙂 … my approach is to midi generate place-holder tracks for bass / acoustic / electric guitar etc., parts, then progressively replace them with live performances (timing/pitch correcting only where noticeably sloppy). The biggest challenge for me is playing guitar with personality to the click track.

  5. Rick Herron says:

    This is what i do to good effect as a musician and songwriter currently without a band. For those like me with a DAW, good amps, guitars, synths and load boxes: Download a great drum program. I have BFD Eco which is all I need for writing classic rock. As stated above in the blog I put down the rhythm guitar once I have a simple BFD drum track laid down with the proper tempo, and without fills.

    The amps and effects are running through a tube amp, through a load box, then into the interface and DAW. With the guitar already set up I go back and ad lead guitar while my fingers are hot and I have ideas pouring out from my fingers. I capture the energy of the lead and the vocal parts. If I have an idea for a back harmony I put it down. When I have the song roughly mapped out I only replace the parts that were rough.

    Only later do I ad drum fills or bass, or if I get a band together then I put the other parts down and overdub. I find that first takes can be dynamite and even playing aginst pre recorded material retain the energy. The better one becomes as a vocalist or musician the more they can control their instrument and get a feel out of it regardless of whether it is the original or overdub part they are performing.

  6. db says:

    Overdubs can more than often turn a tune to mush (in my opinion) and can turn a mediocre tune into a so called Classic.
    That is to say if it was played straight it would only sound as it is “mediocre”.
    And sometimes a great tune can be transformed into a mediocre one by the system of overdubbing.
    So I agree this overdub business is overbearing and I would rather a band take 100 takes of a song in real-time and then pick out the best one (that is the best way).
    Tedious but meaningful.
    But that means actually playing the instruments not samples flying off like a calculator.
    The band might even surprise themselves..
    Someone said the carpenters were good at the overdubs yes but there is one song she sang just with the piano the henry mancini one and I preferred it. Its just taste I guess.
    So lets be nearly over with overdubs as the technology progresses that is great but what is missing is the simple approach of there and then.
    Overdubs tend to make a band lazy.
    A good article thanks for reminding me of this I would be very interested in a list of songs that were performed without overdubs probably jazz ones rate the highest?

    • I recently learnt that most of Elvis’ studio recordings are a single take, sung live with the band. In fact, in trying to rework the master tapes, there is so much instrument spill into the vocal mic that it causes the restorers some problems.

      I think there is an artistic place for overdubs, but there is also something special about musicians who can just play and who do so on record.

      Thanks for your comments and interesting perspective.

  7. Pingback: Overdubbing Live On-Stage – The Business of Music II

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s