Lately I’ve noticed more and more intro tracks on EDM albums. It’s an interesting trend, and it makes sense when you think about the evolution of EDM and the live performance of EDM. Except between artists, there is typically no breaks in the music during a live EDM performance. The DJ plays one track after the other without a break. Sometimes they are mixed together. Other times, there is a sharp break that immediately transitions into the next track.
The point is that 90% of the time, live EDM tracks do not begin with silence. Live EDM evolves continuously from one track to the next. Often, the lines between tracks are blurry or non-existent. It’s not like a rock show, where the band might talk to the audience between tracks, tell a joke, or relay an anecdote about a song. At a rock show, there is a distinct break between tracks, but this doesn’t happen in live EDM.
So it makes sense that EDM artists don’t want their music to start from silence. The intro track, or prelude, is a way of setting up the first track on the album.
I compiled a playlist of some of my favorite intro tracks. These mostly fall under the broad umbrella of “EDM” but there are a couple hiphop artists in there, and some of the tracks don’t us much electronics, they are simply by an electronic musician.
The playlist shows how intro tracks have a distinct identity. There is frequently talking in intro tracks. This often seems to relay the concept or feeling that underlies the album. Musically, these tracks are often unrelated to the rest of the album. Sometimes, as in the case of Steve Aoki, they are similar to the rest of the album, but other times they are quite different.
For any new readers, the goal of the imitation experiment is to see what I can learn about a piece of music by trying to recreate it.
Coffee Pot (The Percolator) is a classic Chicago House track by Curtis Jones aka Cajmere aka Green Velvet.
In The Underground is Massive, Michaelangelo Matos says that The Percolator was a massive hit in the nascent Chicago House scene.
“When ‘The Percolator’ came out, it was like a tidal wave,” remembers Justin Long. “Around Chicago during that time, there was definitely a Percolator dance. When a dance is made after your song, you know it’s something special – like the Electric Slide.” Another Chicago-bred house jock, DJ Sneak, adds: “It was a track that everybody could play. Not just the ghetto South Side kids, not just Bad Boy Bill on the radio – everybody came for that record.”
I think The Percolator has aged well. It still has the power to move your feet, and it shows how a little clever knob-twiddling could go a long way, even with the gear available in 1992.
There are not many elements in The Percolator. After several listens I hear the vocals, the drum machine, and the lead synthesizer. At first I thought there was a quiet sub-bass in there, but now I think that’s just a low tom.
In many modern EDM tracks, there is a lot of dynamics processing, EQ, and other effects that bring out aspects of the mix. In The Percolator, it sounds like there is very little dynamics processing or equalization. This is not necessarily a product of when the track was made. From interviews, we know that Jones took the track into a recording studio when he had figured out the sounds he wanted. According to the interview in The Quietus, the engineer was an expert. So the lack of extreme dynamic processing may simply indicate that Jones spent time finding sounds that naturally worked well together.
The vocals are looped and sliced, but seem otherwise dry.
I was unable to find a quote where Jones specifically identified the drum machine used in The Percolator. The obvious candidates were the 808 or the 909, but in a Reddit thread on the drum machine in The Percolator, commenters agreed that it sounded like a Boss Dr. Rhythm 660. The drum pattern varies every few bars throughout the piece, usually by adding a new element to the previous pattern.
The lead synthesizer is the most interesting aspect of the piece. The vocal line might be the hook that people remember, but the synthesizer is what moves their feet. To me, it sounds like a low pass filter with very high resonance where the performer turns the frequency from the maximum to the minimum to generate the pattern. This could be any number of available synthesizers, and none of the sources available to me list a specific synth favored by Jones.
There seems to be very little use of the stereo field.
As far as EQ, the track is clearly EQed for a thumping bass 90s club speaker system more than it is EQed for listening on earbuds from your iphone.
Outside of the vocals, there’s only one sample in the piece. Around 3:10 every other part cuts out and a roll on a hand drum is heard for a single measure. It’s a moment to give dancers a measure to take a breath and cheer before the piece wraps up.
To remake this track, I started with the drums. I tried to remake the sounds using Drumatic, my favorite drum synthesizer. The 808 preset on that synthesizer comes pretty close to the sound in The Percolator. Then I tried the 808 Tuned Kit preset in Battery 4, and I thought that sounded closer to the recording with less work. Also Battery is a more portable format, so it’s easier to share. Here’s the list of samples I ended up using. It’s mostly 808 samples with a few other samples mixed in.
Kick 808 6 – Tuned up 5.88 semitones
MidTom 808 2 – Tuned up 4 semitones
HiTom 808 3 – Tuned up 4.45 semitones
Snare 808 12 – Tuned down 6 semitones
Clap 808 2 – Tuned down 9 semitones with a 230ms decay
Kick 909 11 – Tuned down 2 semitones with a Low Pass Filter at 300Hz
ClosedHH House 3 – Tuned up 2 semitones
OpenHH Dubway 1
Snare 808 12 – Tuned down 6 semitones with a 209ms decay
Snare 808 12 – Tuned down 6 semitones with a 66ms decay
Remaking that synth sound was the trickiest part. It sounds like a fast filter sweeping across a sine tone. It sounds like he performed the rhythm by twirling the knob. Unfortunately, I don’t have whatever synth he used. So I tried to recreate it in Massive. I was able to get pretty close using two low pass filters hooked up to a performer, which is a curve sequencer that acts a bit like an LFO.
My sound is clearly not as funky as the original. I attribute this to several things. First, the human performance aspect of the original probably helped. Second, the filter he was using was a little edgier than most digital filters.
An interesting side-effect of the filter sweeps is that they add little percussive clicks whenever the filter reaches 0Hz. These clicks sound a bit like synthesized kick drums with tight low pass filters. In other words, a side effect of this performance technique may have been to add extra thump to the drums. That may be part of what made it stand out when DJs played it back in 1992.
Remaking the vocals wasn’t tricky in terms of editing, but it was essentially impossible in terms of timbre and liveliness. I tried to record the line after an eleven hour day at work. I made five attempts with three different microphones, but none of them had the liveliness of the Curtis Jones line. It sounds like he’s speaking monotonously, but really, he speaks with authority and musical rhythm that is difficult to recapture for someone like me who almost never performs with his voice.
The sound of Curtis Jones speaking “it’s time for the perculayda” is a bit like The Amen Break. It might seem simple on the surface, but the magic in the original can’t be duplicated.
Here’s my remake on SoundCloud.
The Percolator is a great track because it combines unique rhythmic sound design with a quirky, adventurous approach to composition. Although the piece was written when EDM was in its infancy, it reflects a deep understanding of EDM at the time and where EDM was headed.
There aren’t that many elements in the percolator. Just the vocals, synthesizer, and drums. Interest is generated through the internal structure of each part, and how that structure interacts with the other parts.
The lead synthesizer is just awesome in The Percolator. It’s an electronic screech performed in a very rhythmic, danceable way. It’s simultaneously robotic and human at the same time. Also, it reinforces the beat. The filter sweeps create thumping rhythmic patterns that accentuate and contrast with the drums. It’s easy to hear why dancers found it so mesmerizing at the time.
Until I did the remake, I didn’t hear the quirky form that underpins The Percolator. The form is a fascinating hybrid of a layered form with a pop-style verse-choruse form. The drums follow a layered form; they build up element by element, then break down, then build up again. The synthesizer, on the other hand, alternates between three contrasting loops in a generally sequential pattern of intro, verse, chorus. This is easiest to see in an annotated screenshot of my remake.
A final interesting note about this piece is that it may have been created with only a drum machine (and the vocal loop, obviously). The lead synthesizer sounds very much like a filter opening and closing on a classic synthesized kick drum.
The goal is not to make an exact duplicate of the piece. The goal of this experiment is to learn about a piece of electronic music by remaking it. The assumption is that whatever is most difficult to imitate probably represents something unique and interesting.
Clarity was released by Interscope records on October 5, 2012. Stache is track seven, and I never heard it on the radio. It’s written in the key of G minor at 128bpm. It lasts four minutes and two seconds.
Here’s a waveform of the piece.
The form could be called theme-and-variations if you want to attach a traditional label to it. It employs two riffs that are repeated and varied. It employs two chord progressions that are repeated and varied. In both cases, variation is primarily timbral. The piece never modulates to another key, but the synthesizer patches are constantly evolving.
The piece is definitely not in a verse-chorus form that is typical of pop music. There are no contrasting A and B sections, and certainly no repeated chorus.
It also fits with a form that I know colloquially as DJ-mix-form. Basically this label applies to repetitive, layered music where layers are built up several times. In three places it breaks down then builds back up.
This annotated waveform shows how I broke down the form (click for larger).
The first riff is an eighth-note arch pattern that lasts for a single measure. The first appearance of the first riff occurs over a pedal bass G. Then it is varied over the initial chord progression of Gm | Bb | Eb | F.
A second riff is introduced around 2:45. It is a syncopated pattern that lasts four measures. It is played over the second chord progression of Gm | Bb | Edim7 | Eb F.
The piece opens with a yell. Outside of drum samples, that seems to be the only sampled audio.
Then a hard beat kicks in with a simple bassline on G, and the first riff. Throughout most of the piece there is not much use of the stereo field. Everything seems to be panned center unless a section is being emphasized. At one point the bass begins bouncing back and forth between the left and right speakers, and from then on the stereo field is used as an effect.
The lead synth sounds almost like a distorted electric guitar. Using filters and other effects, the sound evolves over the course of the piece so that it is never exactly the same.
Differences and Challenges
Remaking this track was difficult. I only wanted to spend ten hours on it, but I almost certainly went over that. Why did it take me so long? Zedd clearly knows every trick in the book. The synths are programmed meticulously. The dynamic processing is perfect, and doesn’t reduce the range or impact of the sounds at all. This took me a long time because I am not as good with a DAW as Zedd.
One of the major differences in my imitation, which you will hear immediately, is the frequency content of the result. I was unable to get such a clear tone from my sounds while squashing the dynamics the way he is. After fiddling with the dynamics processing for hours, I came to the conclusion that he is using a compressor, limiter, and exciter on the main mix, as well as on some submixes. But he has different plugins than I have, and he knows how to use them better than I do.
Another major difference between my imitation and the original is the complexity of the synths and sounds. His hits are all multilayered. It sounds like he is layering cymbal crashes with white noise. Many of the synths are also multilayered. I tried to create single patches to roughly match his patches, but I couldn’t get quite there. You can hear the synth layers in the attacks on some synths, such as the bass that comes in at 1:30. I didn’t have the time to fully recreate the layered sounds.
This piece is built around a pretty conventional formula. A very small amount of melodic and harmonic material is layered and transformed using synthesizers, samplers, and sequencers. Then, at three spots, he drops all the loops out except one or two, and layers up again. The variation is continuous. There are no sections that are repeated exactly as before.
A little bit after the middle of the piece, he significantly varies the melodic and harmonic material. He adds a diminished chord to the chord progression, which sounds odd in EDM, and he starts leading with a new riff.
Into this conventional form he adds some very unique moments. At key points, a unique synthesizer, loop, or sampled sound jumps out of the mix. For instance, the piece opens with a sampled yell that doesn’t occur again. This propels the listener into the first riff and chord progression. After the listener hears several variations on the initial material, a single unique bass loop takes over around 1:30. Then a new and different synth briefly takes over again at 2:30. These moments are tentpoles that support the craftsmanship of the rest of the piece. They break up the varied repeats of familiar material to give the listener a sense that the piece is moving forward.
Conventional layered EDM form
Gm | Bb | Eb | F
Unique sounds/loops at key moments
Cymbal hits layered with white noise
Extremely tight dynamics processing and drum programming
Here are the audio files and presets I used to make this track. This zip file includes several Massive presets, a Battery preset, and a couple audio files. It doesn’t include the Drumatic preset I layered into the drums.
Recently, I was talking about The White Rabbit Project with a friend at work, and he mentioned how much he liked the opening credits. He said the music was really good. I hadn’t noticed the music, so I went back and watched the titles again.
The opening credits really rock, and I wanted to find out why. So this weekend I recorded the intro, dropped it into Reaper, and started breaking it down. Using Kontakt, Massive, and Battery, I was able to create a pretty good imitation of the musical portion of the theme.
But when I listened to the music alone it felt lifeless. It starts with a single note on the piano with a low pass filter removing the fundamental frequency, then it moves on to a simple beat made with a bass drum and snaps/claps. It’s basic, even cliche stuff. Even when the synths take over, my music-only imitation seemed to have none of the vivacity in the original.
So I watched the intro again, and this time I listened to all the stuff that isn’t made with synthesizers or drum machines. The intro actually starts with some technobabble type foley, then uses explosions, thunder claps, and glass breaking to punctuate various moments. The foley is really interesting because it is neither strictly “foley” nor is is “musique concrete”; it is neither strictly synced to the video, nor is it strictly synced to the music. The sound effects are used as foley at times, then as musical punctuation at other times.
The sound effects are the glue that hold the piece together. They being together the rocking theme and the badass visuals. Notice that there are more sound effects at the beginning of the track. They act as an introduction to the sound world of the piece and allow the kicking lead synth to land with full impact when it finally comes in.
I prefer to call this album semi-algorithmic because some of the music is purely software-generated, while other pieces are a collaboration between the software and myself. Tracks four and six are purely algorithmic, while the other tracks are a mix of software-generated material and more traditionally composed material.
The software used in the sound collage pieces (1, 3, 4, 6) was inspired by Melissa Schilling’s Small World Network Model of Cognitive Insight. Her theory essentially says that moments of cognitive insight, or creativity, occur whenever a connection is made between previously distantly related ideas. In graph theory, these types of connections are called bridges, and they have the effect of bringing entire neighborhoods of ideas closer together.
Finally, these sound graphs must be activated to generate sound collages. I used a modified boids algorithm to allow a swarm to move over the sound graph. Sounds were triggered whenever the population on a vertex surpassed a threshold.