I can’t dance. Well, I can technically dance, but I’m way too self-conscious (and old) to do it in public. This is a shame, because dancing is a proven way to improve your mood and feel better. But I can write dance music and create dance tracks. Like this one.
(It’s also available on Spotify, Apple Music and all the many other streaming services.)
This post is about how the track was written, arranged and recorded. And anything else I felt like adding. If you like it, consider subscribing so you’ll get more like it as I release songs. Or if you don’t like it, that’s also good.
The Words
When you’re young, the idea of a night out can be full of promise, of losing yourself (and your worries) in moving to music, wrapped in light and sound and dancing. The night is something to be anticipated, to lose yourself in. That’s the first image I had in mind when I started jotting down ideas for words, and it’s what the first part of the first verse is about.
But this has been something of a dark year, both globally and personally. Here in the northern hemisphere it’s winter, and Vancouver is grey and rainy (and this song is being released the day after the solstice, when the nights are longest). Also, that election result happened, and there’s climate change, and war. And a few weeks ago my cousin Danny died, too young and too fast. So I wanted to add a second sense to the words, so they say two things at the same time; the night is the pressure of the world we’re in, and the darkness is how it weighs us down, but the light is the connections between people, and the dance is how we keep going. We just keep moving through the night.
Of course, you can ignore all that metaphorical stuff and then it’s just a groovy song about dancing, as many of the best disco songs are.
City lights down far below
We’re on the edge, and letting go
All the night is breathing
If the daylight got you down
We got the answer, got the sound
Better start believing
We don’t ever let the darkness in
Don’t ever wanna let it win
Not gonna disappear
And this connection that I feel
Holding onto what is real
Is what we need to be here in...
The night, surrounds us
Suspended here in the atmosphere
Tonight, around us
All the people who we are, all the people who we are
And in the moment, in the heat
Close your eyes and feel the beat
Leave the silence out there
Let it in then let it go
It was always ever so
There’s an answer somewhere
We don’t wanna let the darkness in
Don’t give up hope and let it win
Won’t take it anymore
This connection that I see
When you look right back at me
I just wanna be right here…
Tonight, beside ya
There ain’t no shadow to hide ya
In the light, around ya
Close your eyes and feel the night, close your eyes and you feel the night
I don’t wanna break the spell
Got to save it and I can tell
This could end too soon
All I said and all I’ve done
All I lost and all I won
All I gave to fortune
Give it up and let the groove begin
We’re gonna let the music in
We’re gonna make it clear
Hold the feeling ‘til it’s gone
Let the time keep running on
For a moment stay here in...
The night, around us
Let the rhythm come right back down
Tonight, to hold us
Lift us up and control us
In the light, I found you
All the music around you
In the night, that calls us
Just keep on moving through the night, keep on moving through the night
The Artwork
The image is a detail from a photograph on Unsplash by Solen Feyissa, of oil drops and bubbles floating on water. I liked the colours, and I wanted a break from the more subdued, darker artwork on the last few releases. The colours don’t feel like night, but they might feel like dancing.
The Backstory
About six months ago, I rolled a die twice to make two more-or-less random choices. The first choice was a musical genre, and fate chose disco. The second choice was a musical scale, and the die commanded me to use a Lydian scale. My mission, should I choose to accept it, was to write a disco song, using the Lydian scale.
The Lydian scale is… interesting (see the music theory bit below), and as far as I can tell, there are no disco songs that use it. Then again, it wasn’t supposed to be an easy challenge.
I started by listening to a lot of disco, especially songs that reminded me of school disco playlists when I was a teenager; Chic, the Bee Gees, Earth, Wind And Fire. It turned out to be a lot of fun. Disco is generally relentlessly optimistic, upbeat and downright groovy, and listening to these old songs from a musical, analytical point of view showed me how well-written, arranged and performed they were.
This track is not really disco; it’s a bit too raw and unpolished. It’s disco only in the same way as Blondie’s Heart Of Glass. But it definitely has disco somewhere in its ancestry. The most obvious influences include:
The guitar style borrows heavily Nile Rogers (the genius behind Chic, and the rhythm guitarist on Daft Punk’s Get Lucky and Lose Yourself To Dance). I took inspiration from Nile’s playing on I Want Your Love and Good Times. I have nothing like his ability or sense of groove, but this pushed me to play guitar in a whole new way.
The rhythm arrangement is based on Stayin’ Alive by the Bee Gees, which was the first record to use a drum loop. The drummer for the album sessions (Dennis Bryan) wasn’t available when the track was recorded, so they took two bars of the drum part from Night Fever and created an actual physical tape loop to repeat those bars over and over. I used almost the same tempo (Stayin’ Alive is 103bpm, The Night is 100bpm), but I did play it all because I have a thing about not using loops.
The idea of basing the whole arrangement around a single repeated groove and bassline came from listening to the Talking Heads (who were heavily influenced by disco and funk). Tracks like Slippery People or the ever-wonderful Once In A Lifetime are little masterclasses in how to build dynamics over a fundamentally unchanging groove, and the Talking Heads developed that into a fine art while working with Brian Eno. Listen to Born Under Punches or Crosseyed And Painless for sheer funk!
The Arrangement
The Rhythm
Let’s talk about groove. It’s a difficult thing to define… so I did some reading and this seems like a pretty good way to explain it:
Groove is what happens when rhythmic elements interact in a way that creates forward momentum and makes you want to move. It's not just about playing in time - it's about how different parts weave together, creating tiny pushes and pulls against the basic pulse. Think of it like a conversation between the instruments, where everyone's speaking together but with their own subtle inflections.
A track has groove when all the elements - drums, bass, guitar, etc. - work together to create a pocket, a sweet spot where everything feels like it's in exactly the right place, even (or especially) when it's not mathematically perfect. It's that quality that makes you nod your head without thinking about it.
I think this track grooves a bit, so let’s see how that’s done, starting with whatever funky rhythm the drums are playing:
That is… very clearly not funky. And that’s deliberate; the drums are an anchor, playing a strict, simple beat that sets up a rhythmic grid for everything else to bounce off (For a fantastic example of this kind of drumming, see this video of Jonathan “Sugarfoot” Moffett playing Billie Jean). Oh, and in the full track there’s also a cowbell in there to emphasise the strict beat a little more.
Playing that drumbeat was difficult because it has to be so precisely in time. Playing the bongos was difficult in a whole other way, because they are the first element that’s grooving. Here are the bongos on their own, with the cowbell so that you can still hear the strict beat:
The bongos play around that rigid drumbeat, adding accents on offbeats. The best approach I’ve found for this sort of percussion is to play it while moving; it needs to match up with how your body responds to the music and the beat. All these drum and percussion parts were recorded twice; first as simple guides and then again once most of the rest of the track was complete, so that I could feel the groove while I was playing.
There’s a whole art of using the different tones of the bongos to create dynamics, and with the help of a lot of good Youtube videos and experimenting, I came up with something that I think works well, which is pretty much the same pattern all the way through the track. Of course, I didn’t use real bongos - these are sampled, played with a drum pad.
From the point where the bass comes in, I added a set of shakers. A rhythm track needs to balance low, mid and high frequencies. The kick drum is the low end, the snare and bongos come in the middle, and the hi-hat and shakers add some energy up top. The shakers aren’t obvious in the mix, but they add a sense of movement. Here’s the full rhythm:
Both the bongos and the shakers move from left to right in stereo. There are producers who say that this is unrealistic, because you’d never really hear two separate bongos on opposite sides of a stage, and you’d need arms ten feet long to have shakers move as much as these appear to. I, on the other hand, don’t care because I’m trying to get people to move.
Later on, especially in the choruses, I added handclaps, because disco.
The Instrumentation
Guitar, bass, Rhodes electric piano, and a 70s-style funky clavinet (which is a plain old clavinet with an auto-wah effect on). I’ve mentioned before, in other articles, that one of the General Rules of mothershout projects is that a track should sound like a band playing. That rule exists to keep me from adding too many parts and over-cluttering the whole thing up, and it also means that I have to think carefully about how many imaginary band members I have on the imaginary stage, and what each of them is doing.
I’ll start with the guitar, because that’s the first instrument to come in. The pattern it plays is eight bars long, made up of two four-bar variations. Here’s the full thing, on its own:
Technically, that’s two guitars, one on the left and one on the right, playing exactly the same pattern. The tiny variations between the two performances add some movement and humanity (you could also call it imperfection), but a band could do this perfectly well with a single guitar and a little bit of audio trickery.
As soon as you hear that guitar, you’re hearing the “core groove” of the whole song, where the guitar plays chords. Two beats, a gap, three beats, two delayed beats. That’s the foundation of it all - the rest is variation. The bass follows that same pattern with its own variations:
I’ll add the rhythm track back in next, because the guitar, bass and drums interact with each other in a carefully chosen way.
The first two gap in the guitar-and-bass core groove makes space for the snare hits. The second note of the delayed double-beats (half way through) lines up exactly with the next snare. Each of those snare hits also has a kick drum hit to emphasize it. The parts all fit together, but also leave space for each other.
I came up with this by creating an eight-bar loop of the drums and percussion, adding guitar, then bass, then redoing the guitar, then the drums, then the bass… and so on. Each time I played one part along with the others, I tried out variations until I found something that I liked better, bouncing off and responding to my own playing. I’m pretty pleased with how it came out.
A clavinet is a mechanical keyboard that works like a guitar - it has strings that are struck by hammers. On its own, it sounds kinda twangy and uninteresting, like this:
But Stevie Wonder used one on Superstition and it became a signature disco-era sound. Because it’s got a sharp-percussive sound with lots of frequency content, it sounds great when you play it with short snappy notes through a whole load of effects:
There’s an auto-wah, which makes it sound almost like a human voice, as well as a compressor to level out the sound, a touch of distortion for some edge, a flanger so that the repeated notes vary in sound and a fancy stereo tremolo to bounce it from left to right.
I Hear Voices
If you’ve listened to other mothershout tracks, you are an excellent person. And you might also notice that I used a different lead vocalist. This is Nyl, the latest voice from Eclipsed Sounds (who produced all the voices on this track). The voice provider for Nyl (whose pronouns are officially they) is singer and musician Zacariah “Zay” Driver. I haven’t heard Zay’s music, but Nyl is a really versatile and flexible voice, with great range and rhythm.
From the first chorus onwards, there’s a second voice singing an upper harmony above Nyl’s line. This is Saros, voice singing in falsetto mode, because falsetto vocals are very disco. Like Nyl, Saros is described as nonbinary, but I was able to dial in a slightly more male vocal tone, which combines very well with Nyl. Later on, two more voices join in, singing harmonies and then call-and-response with the lead vocals. These are Solaria on the left and Asterian on the right. Asterian’s a bass voice, but pushing him up in his range gives a seriously good dance-track vocal performance.
These are the same four voices that I used on In Silence, but singing in a very different style.
The Build
It’s common to have a track build as it plays, starting from a quiet introduction, building in energy and level to a peak, then maybe dropping in energy before a final crescendo, or maybe dying away quietly. Here’s a “map” I put together of The Night and how it builds, to help me keep focused as I was arranging and playing.
As the track progresses, more vocals come in; first the harmony vocal (singing above the lead) and then the backing vocals (on the left and right). Notice that the backing vocals get more dense (darker colour) towards the end.
The electric piano part gets higher on each chorus (not easy to hear, but it adds to the energy). The funky clavinet comes in from the second chorus.
The “breakdown” is where everything drops out except the guitars, kick drum and percussion; it’s a tried-and-tested way to add a breathing space to an arrangement, which can then be followed by everybody coming back in for a Big Ending. And talking of endings…
That Ending
I did something for the first time on this track; I ended with a fade. That’s a technical violation of my own rule about sounding like a band, but disco records had fades for several good reasons.
First, a fade is useful to DJs, who get a longer time to match the tempo and create a transition to the next record. Second, a fade solves the problem of writing an ending to a track built around a solid, repeating groove. And finally, a fader lets the listener feel as though the dance is still going on after the music ends. So it felt right for this track.
Getting a fade right is surprisingly difficult. If you do a “linear fade”, where the level drops smoothly and steadily, it can make the ending feel weaker (and I’m not sure why). I used an S-curve fade, which drops more between the lead vocal phrases and keeps the energy up. Then you have to choose the last thing that the listener hears (if they keep listening to the very end); if you end on a downbeat, the song feels like it’s ended. I chose to end just as the lead vocalist is starting a final line, to add to the sense of the song carrying on.
The Music Theory Bit
But wait! It’s not all super-music-geeky, and I have real audio examples to show off what I was trying to do.
The Lydian mode has a specific sound, sometimes described as “ethereal”, or “dreamy”. It’s the scale used in The Simpsons theme. But let me try and explain it using simpler examples.
Here’s a plain old E Major scale. It’s sung for us by Nyl, the lead voice on the track, using the good old Do Re Mi solfege syllables. (If you don’t know solfege, think The Sound Of Music and that Doe, a deer, a female deer song).
And here’s an E Lydian scale. It’s identical to a major scale, except that the fourth note is altered. It’s one half-step sharp (and sung as Fi instead of Fa in solfege):
Subtle, isn’t it? It gets more interesting when we start thinking about the chords that fit the scales.
Building chords from a scale is easy. Here are the notes of a major scale, in solfege again.
Do Re Mi Fa So La Te Do
To create a chord, choose a starting note, then pick every other note from the scale (working upwards) until you have enough notes for a chord; usually three. (If you want to create a fancy chord, you can keep skipping and adding, but I’ll stick to three note chords, or triads).
For example, to build a chord on the first note of the scale, start with Do, skip Re, pick Mi, skip Fa, add So and now you have a three-note chord: Do, Mi, So.
If we start on Re and add Fa and La. And so on. Here are all the chords for a major scale. I played the chords as arpeggios (playing the notes one after the other) so that you can hear all the notes:
If we do the same chord-building trick, but using the Lydian scale, we get different chords. Listen to the second chord in particular. In the E major scale, that’s an F# minor. In E Lydian, it’s an F# major:
Remember that altered fourth note of the Lydian scale (Fi instead of Fa)? That shows up in the second chord and changes it from minor to major. That altered fourth note is the “signature sound” of the Lydian scale, so I chose to have the song move between the first chord (E major) and the second chord (F# major), to make it sound clearly Lydian. The piano part of the song plays these chords:
That second chord is the key to the sound. You might recognise it if you’ve listened to Pink Floyd’s track Learning To Fly - the first two chords use this trick (though it’s not a Lydian-scale song).
An older example is Fleetwood Mac’s Dreams, which is subtly Lydian. Like The Night, it alternates between the first and second chords of the scale, both major (F and G). But the melody mostly stays away from that sharpened fourth note, landing on it briefly (on the last notes of it’s only right and it’s only me, and a few of the final woh-oh-oh notes).
Another example that’s a bit more up-to-date is Katy Perry’s Teenage Dream, where the two main chords (Eb and F) could be interpreted the first and second chords of Eb Lydian.
Teenage Dream (which is a brilliant little master-class in song arrangement and production all by itself) exploits another thing about these two chords. You can hear them as the first and second chords of a Lydian scale, or the fourth and fifth chords of a major scale. Teenage Dream never resolves this ambiguity, which I think gives it a kind of unsettled, floating quality. I did not want The Night to feel unsettled, so the bass keeps hitting an E note, to ground the listener in E Lydian.
There was another reason for picking these two chords; the mothershout song Rise also uses two major chords one step apart, but in Rise they’re two different chords from a different scale (the I and bVII of B Mixolydian). I liked the symmetry of using the same "harmonic movement” between two major chords, but in a very different way.
The melody is, of course, also Lydian - the second note is that altered fourth. I’ll let Nyl demonstrate. The words and syllables in bold use that altered fourth note:
City lights down far below
We’re on the edge, and letting go
All the night is breathing
If the daylight got you down
We got the answer, got the sound
Better start believin’
There you go. Lydian chords and a Lydian melody. Challenge accepted and (more or less) met!
A Bit About Engineering
This last section is definitely for the music equipment and production geeks.
The rhythm guitar parts were played on an Epiphone Dot electric guitar, central switch position so both humbuckers are on, full tone and volume. Flatwound strings. Unusually for me, the guitars (one left, one right) don’t use any sort of amp or cab simulation. They’re quite heavily EQ’d and use a J37 tape emulation for some compression, but otherwise it’s the plain guitar sound.
The bass is a Squier Mini-P bass, with flatwound strings, though an SVT-4 Pro bass amp and 8x10 cab simulator.
The electric piano is Pianoteq’s “Rhody Basic” preset, modified. It’s in mono, but put through Waves ADT (Automatic Double Tracking) so that it appears symmetrically at two positions across the stereo image.
The clavinet is the Logic Pro stock clavinet. I started with the Cop Show Clav preset and modified the effects section a lot. The stereo movement is done by putting the mono clavinet through Logic’s stereo tremolo.
From the start of the last chorus to the end, I added a high E6 note using a simple basic string from Logic’s ES2 synth, low in the mix but quite high in the reverb. It’s barely audible, but it helps lift the last part of the song. I normally don’t use synthesizer sounds, but there were plenty used on disco records in the late 1970s, so it fits.
The drums are Native Instruments’ Abbey Road Drummer 60s kit. I found the 60s kit was closer to the sound I wanted than the more historically accurate 70s kit. The percussion is a mix of Logic’s stock percussion library and Native Instruments’ Kontakt factory instruments.
There are two reverbs; Waves’ Abbey Road Chambers for a bit of room ambience and Waves’ Abbey Road Plates for the vocals, hihat and electric piano. The vocals have a really big 75mS predelay on them, so in the mix it sounds a little like a slapback echo, but it keeps the reverb well away from swamping the vocals.
The plate reverb has a stereo widening plugin on, which increases the width gradually as the song progresses, to help the overall build.
All the vocals go through Waves’ Vocal Rider, which acts like a slow compressor and keeps the levels consistent. The lead vocals also go through Waves’ ADT to thicken them up a touch.
All the tracks have a standard set of “1970s-style” plugins (all from Waves):
A J37 tape emulation, which is also sometimes set up to include a delay. This adds a touch of tape compression, and also gives a little mid-frequency harmonic boost. It’s on every track that would be played from tape in a typical 1970s studio (i.e. generally on the tracks and not on the buses).
A 1073-style EQ (the Scheps EQ). I use this as a way to force myself to be more creative and thoughful with my EQ, since it only has a high-pass filter, a low shelf, a single mid-frequency bell and a high shelf, and the frequencies of all of these are either fixed (for example, the high shelf is 12k) or switchable amongst a fixed set of frequencies (for example, the HPF can only be at 50, 80, 160 or 300Hz).
Compression is either an LA-2A or 1176 emulation (from Waves’ CLA series) for tracks. There’s an SSL G-bus compressor on the main mix bus.
I also used Waves’ Abbey Road Saturator on a couple of the tracks (especially the lead vocals). This is a difficult plugin to figure out, since the manual is very light on detail (I spent a couple of days investigating it with plugin analysis tools and posted a writeup to the ever-useful VI-Control forum). But it’s very good at bringing a sound up in the mix without boosting the level, by adding a touch of focused harmonic distortion.
Mastering was done in two phases. I used Logic’s Match EQ to build a reference profile from a commercial track that had the right kind of sound, and applied that (carefully, at around 50%) to my track. I let Logic’s Mastering Assistant handle the dynamics, using the Punch mode.
There are 11 tracks of instrumentation and vocals, so the whole thing would have been possible on a 16-track tape setup in the 1970s.