Improvise on a film scene, a video game video or on what you can see outside the window for a few minutes. Now use this as your guide track. If there is not much happening visually, chances are that your improvisation may only be an A part. If you observe a lot changes in the visuals, you may even end up with several parts in the end.
3 Likes
Go to a club, take an E, and listen. Call it “field study”.
3 Likes
This reads like an excellent psilocybin advertisement.
1 Like
One thing that definitely annoys me about pattern based composing is how hard it is to start a melody before the 1. A really nice and common way to introduce a phrase but a fiddly mess to implement with pattern chaining , you’d need a whole blank pattern just to have the first notes at the end of a bar, then record the rest of the melody on a different pattern?
I used to be able to ‘compose’ freely on old midi sequencers like the Roland mc50 which was essential like a midi tape recorder. I really hope some company decides to recreate something like that but smaller and with usb c. (I’ve looked into the retro kits one but it’s still pattern based from what I can tell?)
4 Likes
Thanks, everyone, this has been very interesting and I’ve garnered some interesting suggestions.
A few people have mentioned probability. I get that, but what I haven’t yet nailed is how to ensure this actually sounds decent! If your trigs are probabilistic, I find it hard to ensure that the result actually sounds good. In theory, different tracks could all play at the same time, or none at all, if they’re probabilistic, but this speaks more to my lack of experience here. Probably it’s a combination of probability and hard-coded control.
@wing good to know I’m not the only one coming at it from this angle! I agree with most of your points. I have much less of an issue with harmony when dealing with electronic music. You allude to the more nebulous concept of harmony in a lot of electronic music, the lack of a traditional root-dominant duality. I find that’s largely down to the timbres; I’m always amazed at how (theoretically) discordant electronic music can be - conflicting scales, bass notes that are actually a semi-tone “out” - but it works because the timbres are smudgey, not accute, as it would be with an orchestra or really any organic instruments.
I definitely need to live-play/overdub more. All I ever do is program things in meticulously. And yet I’m perfectly comfortable improvising at the piano so perhaps all is not lost.
Composing vs. arrangement - ah, that’s interesting. That’s definitely a distinction I lack; I marvel at how producers can turn up to a show and remix their track live, make it something other than what the studio version was. If I did a show I’d just press play on my track and go sit in the corner (for this and many other reasons, I don’t do shows - no bookings, please.)
3 Likes
That’s exactly how I use the last trig conditions and why I keeps on asking it for tonverk, usually my last trig conditions are a start of the next melody or some kind of risers at the end if the finishing pattern.
You can also combine no last and last to bring drums cuts or drums variations on last pattern play
2 Likes
@mitya33 Regarding probability, one thing you can try is to use probability or other semi-random methods and then record the generated MIDI or audio so you can chop out the bits you like.
It obviously depends on your tools as to how easy this is, but it’s certainly easy enough in most DAWs.
You can also do Elektron-style note conditions in Bitwig (and probably elsewhere with the right plugins or Max4Live devices) where you can create a pattern and then make different notes have conditions like “only trigger on the 3rd and 5th out of every 7 times played” which can introduce a lot of interesting variation and a sort of semi-random polymetric behaviour but in a controlled way.
I don’t have a musical background like yourself and I literally have almost no music theory at all so obviously it’s a lot of experimentation and trial and error, but I’m starting to realise that this isn’t so unusual and that a lot of electronic music producers lean heavily into a pattern of experimenting and then editing/curating the output to keep what they like. I suppose you could think of it a bit like making a collage but you create the raw material first yourself.
I think a lot of it comes down to whether you want (or need) to have a clear idea of what you want in your head or whether your’e happy to see it as a journey of experimentation and seeing what pours out at the time you sit to make music.
1 Like
I was tempted by that, OPN being one of my favorite artists… did you feel it was worth it? It’s not cheap, but it’s less expensive than whatever new gear I’d buy instead!
I enjoyed it. there were thousands on the zoom call, so it didn’t feel personally that interactive, so if you were to do the course, based on the recorded sessions, it wouldn’t feel that dissimilar. If you’re a fan, you’ll dig it. I’m not actually that clued up on his music, but still got plenty from it.
Some of the best music I made, related to this thread was just stitching audio jams together and collaging stuff in ableton off the grid. made my music way more fluid and escape the grid/pattern structure stuff and just lead with your ears, transposing audio clips till things ‘fit’.
2 Likes
Yes, i mute and unmute tracks while recording a live mix. But it’s not a random version. While producing the track i get a feeling for how it could evolve. Or when which tracks could come in. But there’s room for experimentation when i record the mix.
I made a lot of music where everything was arranged but my music changed and i enjoy tweaking stuff during the mixdown.
1 Like
I use Probability as a « B part » or as a complementary tool. The technic is to compose a melody and then P-lock at 100% the Notes/Samples that you want for the first part. Now, with the Probability set to 0%, you will only hear the Notes/Samples P-locked to 100% and when you’re ready to introduce the full melody, you put the Probability to 100% or lower if you want some random variations.
5 Likes
I agree that it is a chore. I also build my tracks like that, and even if it is just one pattern, I want to create a dramatic form.
The songs on my next album will be “one-pattern-songs”, and because of that, I have planned to use Overbridge, stream all tracks into Reaper for the length of a song and make the form there (mute and unmute), so I can SEE and be fast. And then I can re-record tracks to manipulate parameters according to the then established form. And so forth.
For some future stuff I planned similar ways, recording stuff into the computer, cutting it up there … It’s an unsolved problem, thus a process, which I have come to accept for the time being 
1 Like
I like applying probability to grace notes (melodic or rhythmic) as a way to “naturalise” a programmed part
1 Like
Don’t forget a lot of house and techno producers start as DJs. I did. Still love DJing.
A lot of us produce the way we DJ.
Let it ride for as long as it sounds good 
5 Likes
That’s really well said! To my point about tonic–dominant functional harmony, I would say most modern music doesn’t really use it the way we’re taught in classical composition anyway. With electronic music, if I could summarize my longer idea about harmony, it’s that it feels less phrase–progression–cadence based. More layered with textures coming and going, harmony less strict. Totally agree, harmony is often more discordant and textural/timbre-based.
I personally don’t struggle with this harmonic aspect so much as I love dissonance anyway, but trying to “unthink” everything I’ve been taught about melodic and motific development. Approach music as patterns and sequences. Also I love rubato which is sadly hard to translate to sequenced grid-based music.
So this is funny but out of curiosity I figured out how to play both my Syntakt and Digitone from my notation software (Dorico). Not that I would use them in this way with serious intent, but it was a curious way to explore the hybrid worlds, and get some counterpoint in 4 voices going on the Digitone haha.
I know what you mean about programming things meticulously – it’s not terribly different from writing a score in that way. But sounds like you have that improvisational side waiting for you. Maybe it’s just about thinking of whichever tools you’re using as actual musical instruments and less like sequencers/devices - playing it like a piano basically.
I agree with what others said about arrangement. I recently played my first show and perhaps I’d be one of the producers you marvel at, but my approach was really nothing special: what I did was just create patterns with every track doing something (in my case it was on a Syntakt, so 12 tracks), and then I would copy these patterns and create slight variations so I could switch back and forth. And then live, all I did was save my project with mute states so they would start minimal and I would add and subtract live. Someone actually came up to me and asked “are you writing all this live?” and I told him I wrote everything in advance and was just moving shit around – so arrangement, essentially. I know shows aren’t your goal, but that was a recent lesson I learned myself in how I can possibly make 64 steps interesting with “motives” and “development” lol.
Good luck!
3 Likes
I went to music school for a year when I was young and stubborn and stupid i thought I’d cracked the code: I didn’t like modern music because it didn’t utilize clever voice leading, “airy” room sounds, varied through composition or general imperfection — it was ruined by gridded lack of feel, sterile recordings, extreme repetition, and a million people who just wanted to be “heard” for their own aggrandizement rather than contributing to the form.
I was just looking to be right and prove everyone wrong and try to figure out where things had gotten messed up in music history that led to modern music malaise (I didn’t understand much then) — something that I could make a music career out of. I never made anything of that conviction because none of it actually mattered and it’s all fairly incidental to whether people enjoy anything or whether a composition-recording is valuable to someone.
The answer was always a mix of mute-unmute arrangement, overdubbing transitions, cutting and pasting sections and smoothing them out. There’s nothing magic about the classical/jazz/tinpanalley past techniques, and few listeners are as rigid as I was
5 Likes
Probably should’ve mentioned that my music is rooted in modernist music [20th century, Arnold Schönberg and the endless branches of ideas that followed him] and not classical [18th century] or romantic [19th century] music.
And there, one has the freedom to explore things, do things differently … for example, concentrate on colour, or how it would be called in electronic music - sound:
With score (time stamped to the correct movement):
This piece was written in 1909.
3 Likes
This is what I’m learning has been the workflow behind a lot of my favourite music. I’ve been stuck on this idea that a song needs to be composed linearly with intention throughout. Still struggle to get my head around the idea of editing and postproduction because music to me always sounds as if it’s being played ‘live’ even when it’s clearly a composite if ideas/improvs that happened at completely different times
2 Likes
I also studied classical composition and I’ve been trying to write both drone and techno.
I’ve found that my training has given me a tendency want to make harmony, melody & development more complex, but in techno the main development is the real-time manipulation of timbre over time via effects/synth parameters/filters, plus introducing and removing pre-written interlocking rhythmic patterns.
I.e. it’s an improvisational art-form where the main aim is to create a groove and then fiddle around with the elements of the groove and whatever synths are playing over the top.
Usually harmony / melody are static and very repetitive but they don’t get boring because the timbral changes and the adding/removing of elements creates the interest, rather than the harmony & melody creating the interest.
5 Likes
To the OP, if this really is your goal, just RUN. A DAW will me so much less hassle, and a lot more powerful, and with a visual representation of music that looks pretty much like a score.
The closest thing to a score in the hardware world would be the Hapax.
That said, pattern based machines are great for what they are made for : untraditional sequencing, improvisation, surprise-driven composition. Also great if you like losing work.
7 Likes