Modern Approaches: Arrangement

Arrangement is a fundamental component of practically every piece of music ever written or recorded. We understand creating music, at its most basic level, to be the act of “arranging” sounds in some sort of order over a given period of time. Still, what constitutes a “good” arrangement is not always so easy to pin down. This is in part because arrangement is one of those tricky things that is actually most successful when it goes virtually unnoticed, so subtle in its execution that the listener is unaware of the effect it is having on their experience.

In earlier eras, arrangements were often determined by functional constraints – certain instruments were designed to fill different frequency ranges, and each individual musician's ability would in part help determine what notes were played and for how long. But, like it did for many aspects of music, the proliferation of gear and digital production capabilities have profoundly changed what is possible when arranging a composition. Especially for electronic artists, who often produce music without the use of traditional instrumentation, arrangements can theoretically go in any and all directions.

With so many options, electronic artists can find themselves at a crossroads: Should productions made on a computer or various pieces of hardware use these tools to drive the arrangements in unorthodox directions, or should they attempt a more natural sense of musical movement, mirroring the arrangement patterns of more traditional bands or ensembles? When it comes to the more functional duties of arrangement – making music that fits certain genre constraints, especially in terms of music made for dancefloors – how closely should producers follow certain proven arrangement formulas in order to fit into a desired musical context?

Of all the the aspects of production Modern Approaches has tackled, arrangement is probably the most slippery. Fortunately, we've assembled another quality panel of inventive producers to share some of their personal insights and help us make sense of it all. These artists share a diverse array of perspectives on the subject of arrangement, as well as a number of imaginative tips for overcoming the potential creative blocks many encounter while arranging their tracks.

I am not keen on spending too much time in front of the computer screen, and especially not so keen on spending time in front of the computer screen just moving around little boxes and doing an arrangement in that way. In its visual translation on a computer screen, an arrangement is a movement from left to right, and I think that orientation can influence the creative choices you make while you do an arrangement. The way a song is laid out on a screen leads you to make certain choices based on visual feedback because you have the impression of, “Okay, there is not much going on in this part, so I should move some things over to that empty space.” I think it is very important to shut down the computer screen from time to time and just listen. It may sound quite obvious, but you will make different arrangement decisions if you just let your ears decide.

With that in mind, I try to develop an arrangement by playing it, rather than writing it down or making a few recordings and then spending a lot of time on the computer side of the process. For me, I like to play instruments – by this, I also mean synthesizers and drum machines and such – so deciding where the kick comes in or when the hi-hat begins, all these things I’ve experienced to be more fun when I do a live recording. In general, I try to rehearse or learn a song in some way so that I can actually perform it, and then at some point, after playing it over and over again, I reach this “perfect” take when I record it all live. This perfect take has a lot of advantages. Number one is that, obviously, if you play a song again and again, and sing it again and again, you will become more skillful at what you do, because there is a lot of practice and rehearsal involved. For me, it is also a very playful way to come to the final arrangement of a song, because you can do things like say, “Well, this part feels really good, so I’m going to keep playing it a little bit longer.” Personally, I prefer to record a song 20 times until I have the perfect take, then to record two times and then have to edit and move stuff around for 20 hours.

With dance music, arranging can get quite interesting, because when it comes to the lengths of sections and songs it can be difficult to decide what is right. Especially with club tracks, they tend to seem longer when you are in the studio or at home than when you listen to them in the club – the perception of time is very different. In a club, you can maybe dance to a certain rhythm for three minutes without it changing much, but those same three minutes at home will feel a little bit long. So what I do when I’m working on more obviously “dance” tracks, is that I have a strobe light and a cheap light FX ball, and I’ll turn off all the lights in the studio and then turn the strobe and FX ball on, so that I can listen to the track in a more simulated-disco situation and dance to it. I’ll notice different things in those moments, and it can lead me to make different decisions about the arrangement, because I am feeling the need of the body – I’m dancing and I think, “This part is so cool, I want to keep on dancing to this part for a little bit longer,” and then I’ll start to feel like it’s time to move another part of the body, so I’ll bring the claps in, or something like that.

Even though I make dance music, I approach arrangement somewhat like it’s a band playing all the different parts. I grew up playing percussion, and playing with other percussionists allowed me to learn how to listen for a place in a song where I could come in, or where to add a fill, or when it was right to just keep a steady rhythm. I think it’s a good thing to have played with a band or have been a part of some kind of communal music making before going into making your own dance music, because it can give you a more natural sense of how the pieces can fit together; where a bass player or drummer would throw in a fill, those sorts of things.

A lot of the time, when I finish a track it is ten to 14 minutes long, but I’m not running my own label yet, so for most 12"s and EPs they want at least three or four tracks, and I have to go back and edit down my productions. I like the longer track lengths, though, because it allows me to keep the listener in the groove. Usually when I do something that long, there will be a bunch of changes in the music and arrangement. Basically, there will be three or four parts to each song, and it’s not that I’ve created a composition with three or four distinctly separate parts, but I’ve found ways to rearrange the instruments so that there is a different feeling to the track in each section. That’s what I like about extending a track: it gives you a chance to bring out different experiences within one song, which can be hard to do in a short amount of time – it can feel like ADD trying to do too much in a four minute song (which I’ve done, but try not to). If the track is extended, you can actually spend time hearing the changes and actually feeling the experience of the song going through its metamorphosis.

Often, just being willing to take elements in and out of a production can really help an arrangement – it helps create space and movement. If you’re going to add another part, gradually taking something out before you bring the new element in works really well, because when you take out a part, then the listener is waiting for something else to happen, so they are in some ways more ready for a new element to take over. Then, one or two bars after, you can bring back in what you had removed ahead of the transition. Alternately, if you’re working with longer loops, you can do something subtle, like change the frequency of the loop, or run it through a weird filter so it pans at the part where you bring in a new instrument. Those sorts of things can help break up the monotony without losing the momentum of the groove.

Recently I’ve been using Native Instruments’ Maschine a lot, and I like to start a song on there because it gets me away from the computer screen, but also because it introduces happy accidents, which I try to encourage. I’ll just work on a loop on the Maschine for 15 to 20 minutes, and then I’ll start to make a new loop while I’m still in that same zone. From there I’ll try to come up with four or five loops that are all part of the same session (and which use the same sounds and samples) and then they can become the different sections of the song. If you write a really good 16-bar loop with a few interesting layers, you can usually break that down, and then you already have 50-70% of the song.

With those four or five loops I’ve made on the Maschine, I can go back and forth between the scenes quickly and hear things like “Okay, that’s a potential A section, and this can be the B section.” Sometimes one of the loops doesn’t fit, so I’ll just drop it, or maybe it could be a cool random B section or something, but the idea is to get the sections out fast while I’m in the zone. Once I get into Ableton, then I can take a step back and re-evaluate the structure, but I still have these pieces, and then later on I can spend the time combining sections or working on the transitions.

Another little tip that helped me learn about strong structure was to just bring a song that I liked into a session, and then I would just copy that same strong structure. I’d go in and drop markers for “Intro,” “Chorus,” “First Verse,” etc..., and then I could just follow that general structure with my own production. At the end of the day, if the song structure worked for that song, it can probably work for your song, and it’s just a good way to learn how other people have gone about structuring their tracks and maybe find patterns you can use later on.

For me, arrangement is a way of unfolding narrative in music to produce the musical depiction of an event, an adventure or a special feeling. It can also be a way of organizing chaos, but to make expressive music you must do more than merely organize the drawer, as it were. For me, some of the most impressionable or beautiful experiences I’ve had musically are much like the dumbfounding experience of “nature” and are beyond understanding in a way; beyond the possibility of complete knowing while at once being unmistakably perfect in alignment. Arrangements must unfold in a similar expression of unknowing that teeters between what is mindfully constructed and also makes even me wonder how I made it.

It may sound semantic, but I think the arrangement is at once determined and unpredictable at the strike of the first note. The first note already contains so much arrangement-relevant information: context, purpose, history/origin of the instrument or sound and character. Is it a muted sound or a certain damaged, gentle threat? Is it percussive and lively or sporadic and full of energy? Is its decay quick or sustained?

While for me it is important that an arrangement should ultimately try to evade the grid of mundane life, I constantly consider its construction as time is “X” and event is “Y.” Arrangements are built upon the X and Y axis: X is a horizontal construction – how the music unfolds over time – whereas Y is a particular event that occurs at a discernible point in the time continuum (I personally think of the Y axis as the input of harmonic information, and the X axis being more related to space). But it’s important to move beyond whatever compositional framework you are using because that in itself suggests (if not demands) an arrangement formula. It’s easy to let your thoughts be informed by the cubical structure of the grid, lines of the score or the nebulous structure of the patch. When I arrange, I choose if I am going to make aware of, celebrate or move beyond those compositional confines.

For me, an arrangement is done when the parts I have arranged express themselves clearly and support the others in the whole. There should be some sense of balance amongst voices and objects, and each note must have meaning, purpose and function. Sometimes creative or experimental arranging is a process of adding a whole lot of stuff and then removing them one by one, but it’s a seer’s goal to understand what the piece needs before the piece itself knows it.

Arrangement is a comprehensive thing – it’s central to the basic impact of a track, so each recording’s arrangement has to be customized somewhat. (However, I still use a lot of repetition and themes across my productions.) Depending on the project and record, I’ll settle on certain unifying ideas about arrangement, sounds, structure, etc..., at the outset. I like to commit to a lot up front; it helps the entire process to flow and I find it especially useful to have a “blueprint” to refer to when I get stuck. Of course, there is always some adding, subtracting, rearranging and general spontaneity that comes later (sometimes really late) in the process, but I try to minimize all that stuff while not restricting it completely. For example, my Egyptrixx and Ceramic TL projects both exist along a basic spectrum of both impactful and tranquil ideas, and these feelings are reflected in the arrangements, so the tracks tend to toggle between really dense, smashed arrangements and sparse, spacious ones.

I like to work with non-musical inspiration for a lot of my arrangements. Mining for non-musical ideas to incorporate into my material has in some ways always been an important device for me and it serves a few purposes: It can function as a type of conceptual North Star and bring a unique unity to a project, and it can prevent obvious references to other material and help avoid being derivative. These non-musical reference points can also operate on a symbolic level as well as a more literal sound level. For example, the Anamai material (a collaboration between myself and Anna Mayberry of noise-punk outfit HSY) uses ideas of depth, location and 3D-circularity in its sounds and arrangements. On the other hand, the Egyptrixx material is usually guided by ideas like microplastics in waves washing up on the shore, or activated circuits, as is the case on the track “Lake Of Contemplation Pool of Fundamental Bond.” On another Egyptrixx track, “Transfer of Energy [Feelings of Power],” I was trying to channel the activation of a power grid – the precise moment of the introduction of electricity. With the most recent Ceramic TL record, I was trying to channel the euphoric/dismal dichotomy of environmental decay, like how an oil spill can have little swirling pools of hypercolor.

Generally speaking, I have a pretty live approach to making tracks. I usually start by making a percussive skeleton and use this as a basic 32-bar loop to jam over and record stuff live. Sometimes, I’ll send MIDI or CV to different synths, but a lot of times I’ll just jam by playing live for 30 minutes or so, and then cut up the best bits. After that, I’ll start to form a rough version of the track and begin arranging it all into something that’s a bit more concrete.

I do my best to give my tracks somewhat of a “natural” feel. What really helps is incorporating a real instrument like a guitar, live drums, or something along those lines. It doesn’t even have to be something that’s really audible – something subtle can go a long way. What can also help is to divide the arrangement up into instrument groups, like an orchestra where an oboe will ask the question and the violin section will answer. I’ve found that it can be very useful to apply the question-answer method to my music, as it really helps tell a musical story.

Arrangement incorporates so much: it’s a matter of style, of function, drama, adapting to formats. At the same time, it somehow feels less important than ever (talking about recorded music in general here), since every kid out there is doing edits and DJs can change arrangements on the fly easier than ever before. Still, I would say super good arrangement work is one of the aspects that really separates the best producers of dance music today from the rest. Ideally, I know an arrangement is working when there’s a clear idea all the way through but also something unexpected, so it doesn’t ever get boring.

I work in Ableton and try to move my songs to the Arrangement View as soon as possible, but then I really enjoy playing a lot with the clips and loops in the Session Mix, going at it hard and raw and recording takes and keeping mistakes. Then, at the final stages of working on a tune, I like to go nuts with tape delay, recording long takes of different elements being sent to it with long feedback tails. This recorded audio is very useful when I need to tie together sections of a song. In addition, some of these tape takes might even become useful in other tracks, or even set the foundation for entirely new tracks. I think getting a more “natural” feeling in an arrangement comes out of doing what feels natural to you, not necessarily from emulating live band tropes. But, of course, if you have the same short, monotonous pattern going through all of your track (like a repetitive kick drum, for example), some slight randomization can feel good. I usually get this from using analog equipment or randomization effects in soft synths or on my DAW.

I guess it’s due to my musical socialization with ’80s pop music that arrangement plays a rather large role in my productions. Most of my productions show a clear song-like structure. My friends call me an “arrangement Nazi” because I can’t just let a loop play for a while without making any changes. I love it when tracks have a real beginning and an end.

In the techno world, it’s become quite a rare commodity to find tracks that are completely thought through. In these times of digital DJing, the loop button seems to have made such a thorough arrangement obsolete, but I could never fade out a track at the end or finish off with two minutes of a looped beat. Hey, that’s cheating! It’s pure laziness! I want you to enjoy my tracks even when they’re not played in the mix. Working for an hour on the final bars of a track gives me a strange pleasure – it feels like I’m adding a little secret to my track that nobody will ever hear.

For me, the best music I make is where it’s really spontaneous and free-flowing and I don’t have to overthink it – it’s primarily about what the sample I’m anchoring the track to says to me, and that’s really how I decide what direction to take the arrangement in. I’ll get my basic eight or 16 bars, and then loop that for awhile, probably go outside and have a cigarette and just listen to the loop from outside the window on my balcony and think about, “What else does this need? Where could it go?” Then I’ll start adding the kick drums, the hi-hats and the claps, and I just let it evolve naturally and try to think about what I want to hear. For instance, on my last record, there is a song “Michelle”; it was a fast one and the sample held on its own, but underneath it, I programmed a different kick drum sequence basically every eight bars. That’s not something I necessarily try do on every song, but something about the sample seemed to ask for it.

There’s that fine line of not over-arranging a track, so that it still feels natural, but then also not under-arranging it to the point that it’s boring. The most important thing for me in an arrangement is just that there is movement. There always has to be something moving or evolving – opening and closing a filter, building and pulling apart percussion, things coming in and others being taken out. That’s how a track talks to me, that’s what gives it life. When I’m working towards that on a track, I’ll try to play it out in a club or bar because that’s where I can easily realize that, “Yeah, that sequence is about eight bars too long,” or “Those strings definitely need to come in earlier.” If you’re making club music, you have to play it out in that environment because certain things about the arrangement – and the mix, as well – just become clearer when you hear your songs in that context.

So many of the elements in my productions actually come from a live band that I sometimes find myself editing the arrangement and the audio to exaggerate the electronic edge of the music: throwing in a stutter, pitch bend or delays. Also, silence can be a very effective tool for transitioning to different parts of an arrangement: half a bar or even a quarter of a bar of silence before a new section adds impact; sometimes I’ll execute this with a cut that uses a snare (with added reverb) hitting on the last beat.

I consider arranging a track probably the most fun part of a production, as I just let the feeling of the music ride and let go of any rules that limit the music. If I have to agonize on anything, then it’s time for a new track. However, sometimes after playing a track in a club environment, I’ll identify something I want to change and I might go back and forth for months whilst testing the arrangement at different shows. The amount of time you can revisit an arrangement is infinite and crafting the perfect arrangement can be a never-ending quest, but it’s great when something just feels right instantly. I draw the line once I have bounced everything to audio.

I should maybe explain first that I do different projects; I run multiple labels and artist names, all with a very different approach to things. As Frits Wentink, I find it really hard to work from scratch, so I usually re-do old material that never got to a point I was happy with, or even re-do finished tracks, like I’m remixing myself. I’ll open a project, and start muting stuff until I come to a base of what I think is the essence of the track, and then I’ll start to replace instruments and sounds. For some of the other projects, I deliberately start with an empty screen and set out to find interesting sounds and build a track around those (it’s more about sound design here). Lately, I’ve been really into recording hardware jams, and since it’s recorded onto a stereo track, the arrangement and mixing can’t be altered drastically. This limitation is sometimes necessary, and obviously determines how and what I can do in arrangement.

On all my projects, something I’ve tried to always be aware of is modulation on key sound aspects, like the frequency of a low-pass filter. I have been using chaotic fluctuations to alter the modulation of my sounds, which makes the track sound more natural, basically because it’s building a slight error into the instruments. Another thing I have found really helpful is using non-synced loops of modulation curves. There is a Eurorack module called Knob Recorder, and what it basically does is record a movement and, once recorded, start looping that movement, but without a synced timeframe. This way the movement doesn’t appear at the same points in a track that you would expect it to, after eight bars for instance, but it still keeps a live feel.

I generally always start my tracks in a live/loop format; usually, a 32 or 64-bar loop that can be arranged and structured afterwards when I feel I have enough ingredients. I’ve recently been trying to change this, however, because I’ve found I waste far too much time in that dreaded endless loop, rather than beginning the arrangement process. (I took a month off from producing recently and have decided to work only in the arrangement window [of Ableton] from now on, because I’ve witnessed other friends work that way, and it’s far more efficient. It’s actually a really hard bad habit to kick, though.)

The arrangement is probably the most difficult part of working on a track for me. I feel when you have that perfect loop, you seem to lose something once you lay it down. For me, it sometimes takes a couple of hours or a couple of weeks to get right, but I’ve found it can be useful to study arrangements of similar tracks by occasionally dropping them into an audio channel and using that song as a guide. When moving from one section to another, sometimes just a subtle soft crash can make a transition smoother, or some trailing or rising tape echo towards the end of a section can help, too. Sometimes your brain just needs to hear another frequency for a transition to make more sense. If I want to give the arrangement more of a natural feel, Ableton actually has a great groove tool function that allows the user to randomize hit points slightly. I’ve found that setting the “random” percentage on there somewhere around 5-15%, using acoustic percussion and manually penning in samples a little off the grid, can really help “humanize” drum tracks. You can also use this to achieve a sort of “vintage” effect, with your channels drifting ever-so-slightly as if they were recorded live on many different machines; this can also be used to create an unquantized MPC/Dilla-esque effect on drums. (It’s equally effective on other instrument channels, too.)