Modulate This! Show Report: Depeche Mode Live at Red Rocks 2009
I'll Be Doing a Presentation on Vocoding in Ableton Live 8 at the Denver Ableton Live Users Group Tonight, September 27th at 6:30

The Making of The Electronica Single "They Walk Among Us" Using Ableton Live + VSTs

image

In this epic post I’m going to share some behind the scenes notes on what it took to compose the song “They Walk Among Us”. If you have any friends who think that all us electronic musicians do is just drop canned loops and hit buttons, have them read this – of course they’ll have no idea what I’m talking about ;^).

About The Song
Here is a fun blurb I put together when I first posted the song that will give a feel for what I was going for:

This song has an interesting story. You see -  I traveled back in time with my music production laptop and a pile of virtual synthesizers to score a 50's Sci-Fi B-Movie. “They Walk Among Us” is the result, Will any humans survive? You’ll have to listen to find out. Extra points if you can tell me the Morse code message in the end of the song. So If you like theremins, 50’s SCI-FI B-Movies, flying saucer sounds, big beats and ear candy I think you’ll really dig this track.

DAW, Synths, Tools, and CPU Usage 
imageThe song was composed, produced, mixed, and mastered entirely with Ableton Live 8 on an HP DV6500T Laptop using Windows Vista 32-bit Premium.

I composed all elements of the song in Session view using a Novation Remote SL controller,Korg NANOKeys, TriggerFinger, and the Qwerty keyboard to play notes. I also used a Wacom Tablet and my Laptop touchpad to enter and edit notes. I experimented with arrangements and created scenes in session view. I then practiced with these scenes to the point where I could perform the song structure live. I did a real-time recording of my work in session view to create the Arrangement view of the song. Then I stayed in Arrangement view and added a few more parts, mixed, and automating.

The song is all digital and no external hardware was used. There are 31 tracks. Most are “MIDI” and most are third party VSTs. As you can see below, I’ve used A LOT of Sylenth1 for this song. While I did use some factory presets, there are many heavily tweaked or effected presets. I also did some custom preset programming from scratch which I’ll describe in more detail below.

- 2 Drum Racks
- 2 Impulse
- 3 Audio tracks used for re-sampling
- 10 Sylenth1 tracks
- 1 Absynth 4
- 2 Vanguard

- 1 Gladiator 2
- 2 Camel Audio Alchemy
- 4 Dimension Pro
- 1 Morphine
- 1 Ableton Operator
- 2 Ableton Collision

On the effects side I used built-in effects within the VSTs. I then added more effects using only built-in Ableton Live effects. I used many, many instances of these FX so instead of giving you a count I’ll list the ones I used. I also did both envelope and capture of real-time automation within clips.

- Beat Repeat
- Reverb
- Compressor
- Redux
- Arpeggiator (MIDI)
- Pitch (MIDI)
- PingPong Delay

- EQ3
- AutoFilter
- Frequency Shifter
- Utility (Bump up some volume)
- Chorus
- Grain Delay

While most of the tracks are MIDI, I used audio tracks for re-sampling and slicing within Live. I also did some re-sampling through Image-Line’s Gross beat.

image

I tend to mix with the tracks as I go along with mastering in mind. By the time I’m done things are nice and balanced without needing a heavy mastering pass. I do use Image-Line’s Multiband Compressor effect from their Juice Pack on the master channel so that my songs match in volume when they are played in album form.

On my new HP DV6 Intel Core 2 Duo (2.53 GHZ P8700), the CPU usage with no frozen tracks is between 20-40% on average spiking to 45% when there is a lot going on.

Running Commentary
Hit play on the player below and read along. If you don’t see the Flash Player, you can grab a Free 128-bit MP3.

<a href="http://markmosher.bandcamp.com/track/they-walk-among-us">They Walk Among Us by Mark Mosher</a>

0:00 The song opens with just drums and a rhythmic synth. I composed this section first and it was the genesis of the song. It’s the bed on which the “verses” are laid upon. I add the bass line on the second pass through the motif. You’ll notice there are no high-hats so as to leave room for all the other layers down the road and to make you really notice when I do add hat and hat-like elements later on the song.

When I was working out the note structure for the rhythmic synth, I was striving to create a rhythm guitar feel but also a “scratch” like rhythm.
52:00 I was noodling around with ideas on what to lay on top of the bed and stumbled on a killer Theremin-like sound. The whole song changed for me at this point – and it became Theremin/Sci-Fi Mania. I improvised a solo around the bed. After some practice with my performance trying to sound like I’m playing a real Theremin I recorded a clip. You’ll notice that the notes are not quantized. This helps give some tension and anticipation complimenting the the rigid bed tracks.
1:11 We go around a second time. This time I added another Theremin track. Slightly different waveform  a lower octave – more menacing ;^). The second track is not simply a copy of the first. It has slightly different timing and tends to be earlier in the pocket. I also added high-hats. I used Ableton’s Redux to downsample the hats to make them more dirty. I added subtle panning in the session view clip. I really like the hats in this section. Then things are full on, there are really three counter melodies working together here.
image
1:30 A little bridge getting you to a B part. Again no hats. This leaves room for the the delay coming back from the bass synth to act as a hat. There is a volume dynamic at the end of this section with the drum track. Lots of work on the kick and snare to make it sound like a drummer is pounding this out.
1:40 Now we’re in the B part. Again, this is a bed of tracks that will support the next section. Plenty of room for a meandering Theremin part.
1:59 Repeating the last section but with some dirty rhythmic bell melody.
2:09 A dynamic build-up in both volume and complexity.
2:18 Ok, I want to get back into a FULL on A part but need a way out of this section. I like the bridge from 1:30 – but don’t want to simply repeat it so I resample that through Gross Beat and use an Audio clip version for the stutter effect.
2:23 Back to A. There is a new bell track in here, again using Redux to downsample. With that there are are 4 countermelodies working together.
2:44 Since we’ve heard the A part a few times, I don’t sustain and instead switch to another break. This break has an oscillating metallic sound. I spent about 45 minutes creating this sound from scratch using re-synthesis and additive synthesis in an instrument called Morphine. This is a precursor to the saucer sound coming up in part C.
 image
2:52 Time for the flying saucers to warm up. I really wanted to mix things up in this C part so the volume dynamics drop completely with only a deep acoustic piano and symbol being played. This anchors the section and makes a lot of room for all upcoming sound effects. The oscillating saucer sound is a preset I created from scratch using Ableton Operator. This patch probably took me about an hour and half to make. I placed this preset into a rack and assigned macro knobs to allow me to “perform” with the saucer in real-time. This is a case where modulating an FM synthesis parameter has a HUGE impact on the sound.
image
3:11 I bring the bell back in from 2:23. It was a supporting line back in 2:23, now it’s the primary line. I continue to build up parts mixing elements from A & C. I love the saucer sound when it goes sub-harmonic in this section.
3:33 Back in with the kitchen sink. This is the climax of the song.
3:49 A transition to the ending.
3:51 Ending. There actually 8 synths instances used to create this with some custom programming. With the exception of the piano all tracks so far are synthetic. So I threw in some menacing tremolo strings at the very end just for good measure. Finally, we finish with some Morse code – “CQ”.

How Long
I’m not 100% sure how long it took to produce this song from start to finish. Elapsed time it was a three weeks. My guess is it was about 15-20 hours of work – and I LOVED every minute.

Share Your Notes
If you ever do a post like this on your blog and want to share, leave a comment with a link so we can all see what your up to.

Mark Mosher
Electronic Musician, Composer, Sound Designer
Louisville/Denver/Boulder
http://www.modulatethis.com
http://www.markmoshermusic.com
http://www.twitter.com/markmosher

Comments