Don’t build a structure by just muting/unmuting parts

It’s pretty tempting.

You’ve spent days developing your utterly brilliant eight-bar loop.It sounds full and thick. All your EQs and compressors are perfectly set. It almost makes you want to get up and dance.

But it’s only sixteen seconds long.

And you didn’t want to make a sixteen second song. You want to stretch it out over five minutes. So first you duplicate your eight bars until it fills five minutes. That’s almost twenty repetitions. And your eight bars already has a lot of repetition in it. So you start muting parts. Let the intro be pretty sparse. Then bring in some more synths. Then the kick drum. Then drop it all away for a bit. Then build up and suddenly drop everything in. Sit on that groove for a minute or so, then tear back the layers until the track ends.

That’s how it goes, doesn’t it?

Except the end result is a bit lacklustre. You can’t quite put your finger on it, but it’s not *special*. Maybe add a few whooshes, a few risers, tweak things a bit here and there… And then what?

The problem is that you’re still thinking in layers. You’re hearing the music as a stack of simultaneous components. You’re arranging your musical ideas by layering them on top of each other. Most listeners, however, hear music as a sequence of sections or landmarks. They prefer to hear musical ideas one ofter the other. In other words, you’re thinking vertically and your listeners are thinking horizontally. You think you’ve got five minutes of music, but your listeners are hearing the same sixteen seconds twenty times.

The solution is not in how you mute or unmute your parts. It’s not in where you added your whooshes and risers. It’s not even in the way you set your EQs and compressors. The solution is in changing your workflow of building a track by stacking musical ideas on top of each other.

Try to build your initial musical ideas side by side. Think about developing sections (you don’t have to worry about the order at first). Give yourself more than sixteen seconds to express your musical ideas. Develop several different ideas, and then put them in the blender. See what happens when you mix and match them. Build some transitions from one section to the next.

And then – once you’ve got some reasonably well-developed musical material – you can start to assemble the structure of the track. Pay particular attention to the contour of the track. This is the time to think about rates of change, primary and secondary themes, listener expectations, momentum, etc. The key difference is that if your starting with a lot more musical material, you have a *lot* more scope for doing interesting things with your track. Your ideas are the building blocks. You don’t have to use all of them, but you’ll be glad you gave yourself the options.


  1. absolutely! man, it can be so hard for some of us to think that way–come up with a new part before layering too much, especially when we’re on an instrument we don’t really have great chops on. i struggle with this and usualy manage but the easy road is pretty tempting. BUT- I really do feel dissatisfied when i hear something i’ve done vertically, as you say, so that alone helps alot.

    Thanks, kim! love the blog!


    • Stanton
    • August 15th, 2011

    Great post!

    • D Krueger
    • August 15th, 2011

    This is so true. I wish you had posted this last year, when I was trying to do this way. Fortunately I made an effort to change my workflow, and that made all the difference. Coming up with different unrelated sections (written in the same key for me) is great for progressing between them and creating interest.

    If you are having trouble getting away from making an 8 bar wall-of-sound loop like I did, I suggest importing a favorite similar song as a reference track into your project, and leave it muted at the top of the sequencer. That way if you get stuck knowing what should come next, you can see what your favorite artist did at that point in the song.

  2. @Mark Bodah
    The problem isn’t the layering – the problem is the assumption that layering is all you need to build up a track. Using an instrument that you’re not familiar with is a great opportunity to experiment and explore new sounds!

    Thanks – glad you enjoyed it!

    @D Krueger
    That’s good advice as well. It’s important to pay close attention to your favourite music and learn about what makes it work. It’s also important to listen to and appreciate a wide variety of music so that you can draw upon many different styles and genres. It’ll widen your musical vocabulary, help you avoid getting stuck in a rut and make your own music more interesting.


  3. Hello Kim, My name is Carl and I love your music information that you provide on your blog. I wanted to know if there is any way I can feed your information on my site so people can view when they visit my site?

  4. @Carl Bolden
    Hi Carl, glad to hear you like this blog. We can probably sort something out – send me an email with a bit more detail of what you have in mind.


    • Grover
    • August 15th, 2011

    “Try to build your initial musical ideas side by side”. are you talking about thinking more horizontal than vertical? i have been stacking 16 bar loops then spreading them out horizontally.

  5. @Grover
    I think you’re on the right track – it’s certainly about more ‘horizontal’ thinking than ‘vertical’ thinking.

    What do you mean by stacking 16 bar loops and spreading them horizontally?


  6. taking a progression using synfire pro.

    building pads, textures, bassline, keys, synths. basically building a chorus type section with all instruments in the mix. then thin the elements and build the song from the parts. initially you might have a 16 bar “loop” then use these different loops and elements to make your song.

  7. @Grover
    That sounds like a hybrid of the two approaches. Also, that Synfire software looks interesting!


  8. Really great post, in a way this is so obvious yet Iv not adopted it in 5 years! I generally opt for the vertical approach at present but have a loop of between 30-60 tracks which have to be mixed with solo buttons to avoid it sounding awful – however this way will avoid clutter and help me to consider the arrangement in a more structured way. Thanks

    I try and arrange the elements as quickly as possible as soon as I feel the loop has enough substance. The loop with the most parts playing has generally been viewed as the peak of the track im producing, but as you say by subtractive arrangement can leave you a bit disappointed. Will be trying this later

  9. @bump
    30-60 tracks sounds like a lot to have before you even have a structure laid out!

    Try this: Attempt to lay out the structure as *soon* as possible, instead of waiting until you can’t add any more tracks. See if you can get the vibe of the song happening with only half a dozen tracks, then work out how the let the music ‘speak’ over the course of a few minutes. After you’ve got that sorted out you can add background parts, special effects, etc. You might find that this approach results in a more organic finished product.


    • Kronsteen
    • August 19th, 2011

    I tend to come up with the song structure before the song. Something like:

    Intro (4 bars):
    – Pad or drone.
    – Basic beat coming in halfway through
    – Movie sample

    Verse 1 (8 bars)
    – Full beat
    – Bassline
    – Busy but unobtrusive treble synth bleeps
    – Building chords in the final 2 bars

    Chorus 1 (2 bars, repeated twice):
    – Full beat (dropping final bar of 2nd loop)
    – Bassline 2
    – Piano chords

    Drum Break: (1 bar)
    – Some kind of roll
    – Maybe another movie sample

    Verse 2 (6 bars)
    – Same as Verse 1 but adding gated chords.

    Chorus 2 (2 bars, repeated 4 times):
    – As 1, but denser piano chords
    – All fading out except piano on loops 3&4.

    I can (and probably will) change it later, but for now I have a skeleton, on to which I can put a chord structure, while thinking about what kind of bass sound, what drums etc to try out.

    The lyrics come last – making this the polar opposite method of the songwriter who uses music as a background to the important thing for them, the words.

  10. @Kronsteen
    Coming up with the structure first sounds like a good approach. That way you know what you’re building towards. Does it also help you to stay on track – so you avoid getting bogged down?


    • Kronsteen
    • August 19th, 2011

    Well, I think ‘getting bogged down’ is something music makers – especially audio engineers – are naturally prone to. But taking a top-down approach I think directs our obsession with detail away from details which don’t matter so much, onto the important ones.

    The important ones might be the EQ, editing the vocal to correct timing errors, or indeed writing the deep and meaningful lyrics – depending on the genre and your own preferences. But I’d suggest that, provided the overall song structure isn’t absurd – say, three verses followed by a fake ending, half a chorus, a two minute drum solo and the full chorus a capella repeating to fade – it isn’t something which benefits much from tweaking.

    And I’m saying that as someone guilty of minute-long ambient intros to three minute futurepop bleepfests.

    What a top-down approach does prevent is two things.

    First, a working pattern than looks like this: Make the intro, fiddle with it till you’re happy with it, fiddle with it some more because you’ve have another idea, make the first verse, fiddle with it, make the chorus, fiddle, go back to the verse, fiddle some more…and maybe never get to the end of the song.

    And second, a tendency to make the composition too cluttered with ideas which occur to you as you’re working. Or to make it unevenly cluttered – where you wind up with 10 drum layers in the intro, leading into a bare verse with just acoustic guitar and voice, then enormous soaring strings in the chorus building excitement until…you lose it all with a bare second verse.

    • Anton
    • August 20th, 2011

    @D Krueger
    I totally agree with you. I think that remaking songs is the good practice.
    Not so long ago I read post about loopitis and time to time make remakes. So now I know that good song aren’t just various combination of layers, it more complex object because one layer can have some variations and with right arrangement it sounds cool!

  11. @Kronsteen
    Excellent points – I couldn’t have written it better myself.

    That’s right – it’s important to explore the expressive range of each loop and each sound. Variations are one of the first steps towards more expressive music.


  1. No trackbacks yet.

Comments are closed.
%d bloggers like this: