Since I started using the moniker ÄUTODIDACT, I wanted to do a song that was a bit of a shout-out, y’know the way the rappers do :-)

I also wanted to challenge myself by writing a song that utilised ALL of my equipment: the DS, the iPhone (with as many apps as I could shoehorn in), the microKORG (inc vocoder), the mini KP, and the Yamaha. The challenge was to use it in a non-gratutious way.

There were a few bits and pieces of tunes and song scraps knocking around that I wanted to try and cobble together into one song. Another challenge.


iPhone (GarageBand, ThumbJam, iKaossilator); KORG DS-10; Yamaha NP-11; Mini Kaoss Pad; MicroKORG


I knew from the off it was going to be a song of multiple parts, I was just uncertain how many. To help myself out I decreed that all the parts would be in the same key (C Major) at the same tempo (120 BPM) and in the same time signature (4/4 – though I ended up breaking this!).

After a bit of noodling, I had it conceptualized to three parts of roughly 2 mins each – though it didn’t quite turn out that way.

Part 1

My broad idea for this was to have the vocal ‘äutodidact’ (that’s ow-toe-die-dact) repeat over a drum pattern that becomes increasingly more forceful, and synths that become louder and more discordant, building into a crescendo that ushers in part 2.

I used the vocoder to record the äutodidact vocal (I’ll write up the detail of that another time), and then used the mini KP to record a distorted version. By mixing the two tracks together the vocal starts out completely distorted and gradually becomes decipherable. Originally, I was going to do that the other way around, but it sounded much better in juxtaposition.

I had a 4- part drum track on the DS I’d made a while ago, with an assymetrical rhythm I really liked, but it wasn’t anything more than that, just a pattern: no crescendo. I mulled over this for a bit, and the solution came thanks to an old friend called delay (is there anything it can’t do?). By setting some delay on the drums synchronised to the BPM, and then turning the effect depth from 100% dry to 100% wet, the complex, assymetrical rhythm becomes a single pounding beat (which I later mixed in with the drums for part 2).

The electrical noise comes courtesy of the mK, as does an arrpegiated bass synth (with extra reverb) that slowly builds throughout the part. I mussed around with the mod wheel on both as the crescendo reaches its climax. The reverb is hard panned to the left; the bass synth starts hard on the right and tracks across to join it. I modified a voice from the mK using virtual patches (I’ll write that up as well at some point), for the cyber-banshee wail, that increases in pitch and volume at it reaches climax; originally it was an octave higher, but it sounded like a kettle coming to the boil.

The last pieces to go in were two iKaossilated bass parts, the same loop with different voices, gradually getting louder and louder, and a (virtual) crash cymbal, panned hard left > right at the climax. (The crash resurfaces in part 2).

Part 2

This was largely  a ‘groove’, as they’re known, which I’d composed on the iKaossilator last year, but never knew what to do with. Recording it was tricky. The iK is great, but there are two major drawbacks – you can’t mix within the app, and you can’t pan. So each part had to be recorded separately (drums, Lead 1, Lead 2 and Bass) and synced manually in the DAW, the levels set, then panned. Lots of tinkering with automated faders! The main lead also has a complex pan.

The nice baleric synth came courtesy of the mK, one of the in-built voices (B68) with some slight modification to the attack and cutoff. The chords are F C G, played via MIDI with the Yamaha.

I re-recorded the äutodidact vocal (after trying and failing to manipulate the original recording – sometimes the simplist thing to do is just re-record. I must remember this!) and distorted it with the mini KP for when part 2 fades out. The different bass sounds for the fade out come from cycling through the bass options on the iK as the part was recorded. The final additions to the fade out/fade into part 3 were a reprise of the electrical noise (modulated) from part 1.

Part 3

I’d always had in mind some kind of piano-led outro, fading in as part 2 fades out. Tying the two parts together was the F C G microKORG synth; I differentiated them by keeping the tempo but changing the time signature from 4/4 to 3/4. This gives the part its melancholic flavour.

The piano melody is played deliberately haltingly – which is just as well as haltingly is my natural style! I’d like to say I played the whole thing flawlessly in one take but that would be a lie. It’s a composite MIDI take with some touch-ups here and there. I used the same reverb as on the arpeggiated bass synth in part 1, and delay at 0.12ms, to give it a haunting quality.

I wrote a second piece for piano, rising in volume and intensity just as the song begins to fade. Basically, it’s the chords, but also the fundamental of each chord going up through the octaves. Separate reverb, same delay as the first piano part but panned to the other side.

Two different guitar sounds via MIDI on Garageband, exported to the DAW and panned alongside each other created the ‘guitar’ part: B A G. I gradually increased the volume of the ‘rockier’ guitar sound as the song fades, pace the second piano.

I really wanted a vocal part, something high-pitched, Beatles-like, along the lines of the ‘oohing’ and ‘ahhing’ of You Never Give Me Your Money. However, there was one problem with this idea: I can neither sing like a Beatle, nor harmonise like one. After a couple of evenings of frustration at the mike, no matter how many overdubs I tried, and even experimenting with the dark arts of auto-tune, it just wasn’t good enough, so I shelved that idea. You have to know your limits.

Instead came up with a melody high up on the keyboard that complimented the other two pieces of piano and recorded it over the second piano part. I played the same melody via MIDI on ThumbJam, with a glockenspiel sample, introducing it earlier. This took longer than I thought, primarily because of my ignorance of buffering MIDI; the samples kept distorting. It took a while to get a clean take.

Then it was a case of getting the different levels, panning, and frequency slotting for each instrument right. This took a succession of mixes: I spent more time mixing part 3 that the other two parts put together.

What I like

  • The ambition. This is only the 4th song I’ve recorded, and I’ve tried to put three different distinct parts into it with multiple time signatures, dozens of instruments/sounds on dozens of tracks. Do I pull it off? I don’t know. But I’ve learned a hell of a lot for the next one.
  • The Balearic synth. The part of creating the song I enjoyed the most of all was dropping that synth into the mix.
  • The transition from part 2 to part 3: you have to be in the know to hear where the synth switches from 4/4 time to 3/4 time, which is what I wanted.
  • Faders. I went crazy for faders on this song, and really felt I got a handle on how to use them effectively.
  • Panning. Building on previous knowledge. There is some reasonably complex panning, but my favourite piece is as part 2 pans out to right as the piano to part 3 pans in from the left.
  • Mixing. I spent a long time on the mixing, particularly on part 3. At times it was frustrating, but it was rewarding. I’ve learned a lot.
  • Bringing it in under 6 minutes. 6 minutes was what I had in mind, but I was sure I was going to spill over. I’m very pleased with 5:42.

What I dislike

  • The ambition. This is only the 4th song I’ve recorded! Probably, I bit more off than I can chew. Do I pull it off? I don’t know. But I’ve learned a hell of a lot for the next one.
  • Not being able to put a vocal on part 3. The song doesn’t finish how I imagined; it will always not quite be the song I wanted.
  • Part 1. The intro is my least favourite part. Couldn’t properly nail it the way that I wanted. The second and third parts outweigh the first.
  • DS Drums. DS Drums aren’t really cutting the mustard, sound very lightweight indeed. Think I need to look into a better drum sequencer/samples.
  • The Mini KP elements. The more I use the Mini KP, the more I hanker for a Kaoss Pad 3.
  • My playing. Actually, I think this is OK, but it takes me sooooo long to get there. I could have recorded the song in a fraction of the time if I took fewer takes, ie was a better musician.


I really threw the kitchen sink at this song, and there is a sink load of influences.

Part 1

Pink Floyd, particularly Speak to Me and On the Run, from Dark Side of the Moon. The vocodered voice is so Daft Punk they’ve practically trademarked it, not to mention Kraftwerk. The Beatles’ Day in the Life, natch.

Part 2

This is a subconscious mish-mash of every 90s Ibiza tune I’ve ever heard! Nailed on influences are Chicago by Groove Armada and Childhood by Dusted. In a less literal way, also 8 Ball by Underworld. And I was listening to Kid Velo by Rival Consoles when I was recording this part, so a bit of that got in there too (especially on the outro).

Part 3

Owes a huge debt to Shine a Light by Spiritualized as it shares the same chords and time signature! The glockenspiel must have come from the part of my brain that used to listen to Tubular Bells by Mike Oldfield when I was a teenager. I had the piano intro of Perfect Sense (Part 1) by Roger Waters in mind for the halting piano melody.


4 responses to “ÄUTODIDACT

  1. Pingback: iKaossilator by KORG «·

  2. Pingback: V for Vocoding «·

  3. Pingback: Virtual Patching «·

  4. Pingback: So this is Christmas… And what have you done? | projectÄUTODIDACT·

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s