How to get from beats to songs?

Yes, I like Surgeon very much. I totally trust his judgment. So I was a bit shocked when he said in an interview that he doesn’t like the Octatrack :grinning:

Yes, arrangement is the right term I was looking for. In the end there is probably no way around Live or Logic. Thanks for explaining OT and BB to me so well in a few words. Very helpful!

It sounds like a hate love relationship

I mean there are really two options in my mind–either performing things live (to some extent even if it is just launching patterns in the order you want) with a plan for structure in mind or recording stems/loops to your DAW and arranging there. I don’t see the point of recording to another piece of hardware as others have said.

If that is what you want to do you I’d recommend getting OB sorted in your DAW of choice and recording your jams then you can edit out the best parts and arrange them into something coherent.

You are very correct. Most people don’t want to listen to 10 minutes of someone free form jamming on one idea.

2 Likes

Yes, the minimalistic approach is often the most challenging – “reduce it to the max”.

Indeed.

You’ll definitely get a lot out of consciously working on arrangement/composition/structure. A DAW can help you with this by giving you tools to lay out parts of a track and move them around until they “fit just right”… but you don’t need a DAW for this. You can work it into your jamming.

Imagine you’re in a band. In a band you “play songs” and jam with several other musicians. You each know (roughly) what you’re doing AND what everyone else is (meant to be) doing, for the whole song. It takes a few (many) passes, some discussion, some mistakes, the occasional beer thrown across the room, but eventually a consensus emerges.

As a solo producer, you and your gear are all of the band. Learn with what each “instrument” (drums/bass/drones/noises/melodies) do for the track. Learn this by playing about, and by listening to other peoples’ music, or by learning more music theory. There’s standard forms (e.g.: verse-chorus-verse-chorus-mid8-chorus, intro-groove-breakdown-drop-groove), and every possible variation. Write yourself notes: write it out; compose a poem and play to it; draw diagrams and follow them; invent your own notation (any shape that helps you remember what you’re supposed to be doing and for how long is fine).

(Don’t pour beer on your Elektrons.)

2 Likes

I mean you could arrange in digitakt as well…

Say you have a 4 bar loop going. With a beat melody and the whole shebang.

Copy that pattern over to several patterns. Change them up, record a filter sweep on one pattern for a drop maybe… remove some tracks and add other parts for a b-section. maybe dissect one pattern for an intro… start from the beat and then use each pattern that you have pasted to add to the arrangement… when your done sample chain them and record it.

In reality that is basically what song mode is isn’t it, it’s just that your doing it yourself.

Be creative with what you got!

4 Likes

if you really want to make songs (and not looping that same thing forever) you need to make more things that you can change in a song.

Perhaps you need to think of song structures.
verse - chorus - verse - chorus…
or:
chorus - drop - verse - raise - drop - chorus,…

there are plenty of usefull links in se internetzt.
depends on your music style you want to make.

you could perhaps take a pattern and change on your digitakt 4 tracks into something different. voila, you have another building block - a pattern that has the spirit of your song but differs from the heritage pattern…

i recommend to record your stuff via overbridge. it’s possible to use such boxes for complete songs, the youtube is full of it, but anyway - it’s easier to fix, fatten up, rearrange, mix and master that thing.

oh: and don’t be affraid to make shitty songs. you don’t need to release them into the wild anyway. but going through the whole process helps to gain speed.

5 Likes

You don’t hate the Octatrack, the Octatrack hates you.

5 Likes

And come join us in the “bad music” thread! We’re prioritizing completion over quality so we can learn.

1 Like

we’ve probably all been in the same spot. i am most of the time. something that helped me was focusing on more structured song arrangements with at least variations, if not fully developed verse/chorus/verse parts. you can definitely do this with just the DT by copying patterns to new slots and tweaking the copied pattern to create that variation.

try to focus less on “finishing” than adding the type of variation you want to keep things interesting. maybe practice that on shorter songs, like a main part and a variation or two. shorter songs can also be harder in some ways if your instinct is to jam out. keeping the ideas bite sized is a good exercise, though.

when using the DT, i multitrack everything via overbridge to ableton, then arrange from there. it helped me break out of the “jam on one pattern with mutes” mode, even if it’s just automating the mutes in ableton. lately, i’ve been doing everything in ableton because it’s been more convenient, but there’s definitely some benefit to learning how to arrange and create variation with the DT and other hardware.

i’d focus on using what you have first, though. it’s just a matter of approaching the music in a different way.

2 Likes

I multitrack into a DAW. After I’ve got a bunch of material with variations, I can approach the arrangement process with an editorial mindset. I need the time to think.

I admire all you once-and-done people. I’d like nothing more than to ditch the DAW and live record songs into stereo directly off the board. My skills aren’t there yet.

This is the best advice for anyone wanting to improve the quality of live jams. By internalizing how arrangements and transitions work in a genre, one can improvise songs instead of loops. I need to follow this advice!

1 Like

I can relate to this. I write more in the intro/verse/chorus/break/chorus traditional sort of song structure. and I just don’t write well with a sampler, when using it for melodies and such. if I’m just using it for beats or “texture” kinda stuff out of my modular (or whatever) it’s fine, because then I’m using other synths over top to write the actual melodies and song flow. but if I get a melody going out of a sample/loop, then the most I can do is create one pattern out of it. I can’t work out how to vary that up and change it enough to develop into a song flow. if that makes sense…

the only solution I currently have to this problem is going lower level with my sampling for the melodic/song flow parts. chopping the sample up small enough that I can play/sequence it chromatically in the sampler (OT for me). and then I could vary that melody for another pattern. but I am usually not doing this, because my outlook is; “that’s why I have synths!”

but for song development and variations, what others have suggested is my typical approach: write a pattern you like, then copy it to a new pattern and change something drastically. for me, that would be like completely changing the main melody. usually the bass line would follow suit, and then maybe the drums need to work with the new bass part, etc… before you know it you could have three or four variations on a theme, and your track is mostly done.

1 Like

Hi.

The only way to learn something is doing it.

Start make songs instead of beats. It’s okay if you will not be able to achieve ideal results in the beginning.
Focus on smaller targets like “make song with defined structure” or “make song in defined genre”. Don’t try to get a perfect result overall. You’ll spread you focus, spend a lot of time and an overall result anyway will be worse than if you focus on a certain part of production. Listen your results critically, don’t afraid to publish it on some feedback-threads, take notes and improve your next songs based on these notes.


It’s a quite a longread.
I’ll describe “project” based approach which used in many art and business departments. And I’ll show how I used it in my recent track.

This approach is based on pipelines with strictly described and separated steps. Actual steps may be different depending on your needs. A simple idea: eat an elephant one bite at a time.

The base idea is about dividing the whole project to smaller steps. The steps should be focused on certain types of jobs.
I’ll show my way what works well for me.
If we talk about music production steps can be like that:

  • definition
  • scoring
  • mixing
  • mastering
  • publishing / promotion

Now, in more detail about each of the steps.

Definition.

Why do you make this track? Is it commercial job or your hobby? Do you have deadlines? Is it profit oriented (mainstream genres or sound) or pure art? And so on.
Answers on these questions define limits (like deadlines) and some choices of next steps (song structure, sound design solutions, mixing solutions, etc).
These questions should set some estimations in your head as well. Like “For this song I need two main parts, a bridge between them and an intro. The track should be four minutes length.”. It simplifies your next work.

Scoring.

I’m used to dividing this step by two parts: arrangement and sound design. I consider these parts as a single step because they are very related to each other.

When you’re working on an arrangement you shouldn’t care about sound at all. Focus on the composition. If you build your song based on an existing beat, ok, use the sounds of this beat. But I prefer to get a piano and some generic drum samples. If your arrangement sounds good with the only instrument like piano, you have MUCH less problems on the mixing step due your instruments are spread well from the very beginning and have less frequency conflicts.
On the arrangement step you define your parts like intro, A, B, bridge, etc. If you’ve decided to focus on a specific genre on the definition step, you likely already have a structure guides. Basic knowledge of composition will be helpful on this step. I use rough melodies and beats during arrangement. Like looped drums without fills. I detail them on the sound design step because I have a choice to add a few notes or change a synth patch (or something like that).

When the arrangement step is finished, start to work on sound design. Again, on the definition step you should define an overall sound of the song. Should it sounds aggressive or softer? Do you want to make it electronic or more real? Does the chosen genre have some defined sound or not? And so on. These decisions as well as composition by itself help you to choose instruments and sounds.
On this step don’t care about mixing. Your definition of done should be like “record all tracks of my arrangement using appropriate instruments”. In the end of this step your track should have the right feel due to arrangement and instrumentation, but it definitely shouldn’t sound finished.
Try to use only volume and panorama faders to mix track to each other. Use compression only as an effect (sidechain, crushing) not a conflicts solution. The same about eq, delays, reverbs and so on. If delay or reverb is the core part of a sound, use it. If you want to add some space and breath to your sound, leave it until the mixing step. Actually good sound design choices (instruments, timbres, patches) eliminate most of mixing problems.

In the end of this step you have a finished song. Next you can mix it by yourself or send it to a mixing engineer.
If you mix it by yourself, a good idea is to bounce all track to wav files and start a new project. It prevents you from endless fixing loop and optimizes computational resources if you use a lot of plugins on the sound design steps.

Important note. If you finished a step it means you FINISHED the step. Don’t come back to it. Small critical fixes are acceptable, but consider it like material you can’t change. For example, on mixing stage I bounce all tracks to wav files and start a new project on my DAW, so I can’t change arrangement or sound design anymore.

Mixing, mastering, publishing.

The mixing, mastering and publishing steps work in the same way. The definition of done of the mixing stage is the best sound you’re able to achieve, but loudness. Mastering is about compliance with the requirements of chosen distribution method (loudness for streaming, compression for vinyl and so on).


And, as I promised, the real world example.

I was needed to make a soundtrack for my video.
What were my steps.

  1. Definition

When your work on a soundtrack, the music should complement and emphasize a picture. Actually, I didn’t have a video at the moment. I had more than 100Gb of footage from my drone and gopro. But I defined a structure and mood of the project. Footage is a mix of scenery views and action scenes. Also, I filmed an incident which I wanted to add to the movie.

I defined my structure as two main parts: exposition (scenery) and climax (action) with bridge (the incident, ) between them. And I was needed an intro for titles.
Because I didn’t have a video yet, I was needed one more step in my pipeline. I decided to make a draft track using the described structure, make a video using this draft and then finish my music using the video (exact scenes, mood and so on).
And I wanted a video length about 4 minutes.
With these definitions I started to work. As it was my hobby project, I didn’t have any deadlines.

  1. Scoring

As I described earlier I was needed three main sections: part A for scenery scenes, part B for action scenes and some epic bridge between them for the incident.

For the part A I wanted something light and fast. I’ve chosen drum’n’bass. For part B I wanted something more tight and heavy. I’ve chosen rock drums and more heavy mix (I mean arrangement, not sound yet). And for bridge I wanted something epic and cinematic.

I have hundreds of tunes in my phone. I just record every decent thing from my head. So I scrolled through tracks in my phone and selected two ideas.
Part A: White Sea. Part A Origin by diswest | Discrete West | Free Listening on SoundCloud
Part B: White Sea. Part B Origin by diswest | Discrete West | Free Listening on SoundCloud

They were not related to each other, but I decided they would work well together in this track.

Then I made a whole track based on these two ideas.
I was using Maschine (because of ideas view, nice arranger, unlimited tracks and hardware feel), some drums samples and a piano.
Draft: White Sea. Draft by diswest | Discrete West | Free Listening on SoundCloud
You can hear it sounds well even with drums and piano only.

Using this draft I made a video and started sound design.
I moved all my tracks to cubase as it’s my main DAW.
Actually, I wanted to make sound design with hardware, but in the middle of the process decided to move completely ITB.
To be honest, it doesn’t matter what you use to make your music if your tool can do everything you need. I just recreated my A4 and Digitone patches using plugins.
Early hardware version looked like that: https://www.youtube.com/watch?v=fh9x2SI3pLs

I’ll not describe each part of sound design, but show a few examples to demonstrate the idea of definition and pipeline.

As it’s mostly drone shooting I wanted airy pads and softer sound.
As it’s nature shooting I wanted to keep some realistic sounds like pluck arpeggios. I was using a harp and a kalimba mixed with synth patches.
As I was needed heavy sound in the end, so I’ve chosen rock drums sound for the part B, my drum’n’bass part is using acoustic drums as well.
As I wanted some epic and cinematic stuff there are orchestral strings and brass in the bridge and part B.
And so on.

I hadn’t problems with a question “what to do to finish the track?”. I just had some kind of checklist in my head like “add pads, record drums, add this epic stuff”. When I’ve completed all the items of my list, track was done.
And I didn’t touch any mixing tools but volume and panorama.
When I’ve finished with scoring, I bounced all my tracks to separate wav files and created a new Cubase project for mixing.

My rough mix at the moment: White Sea. Rough Mix by diswest | Discrete West | Free Listening on SoundCloud
Again, it sounds nice even without mixing.

  1. Mixing

On the mixing step I did all I can to get the sound I want. Just step by step: mix drums, mix bass with drums, add main parts to the mix, add other parts to the mix, add space and punch to the mix.
In the end of the mixing step the track was like that: White Sea. Final Mix by diswest | Discrete West | Free Listening on SoundCloud

  1. Mastering

And then I sent this final mix to a mastering engineer.


You can see, I had certain steps with certain targets on each of this step. I didn’t care about finished track. I had definitions of done on each step and focused on the result of each of steps.
This approach makes big projects much easier as you don’t need to keep a whole project in your head.

14 Likes

Well, having a “Song mode” would certainly help in building more complex arrangements…

2 Likes

I’ve cueued song chains a lot, its no problem…

1 Like

Ha! We must have played in the same band!

2 Likes

Yeah I think he has also said in interviews that even in his DAW arrangements he often lays down a full jams of mono synth before he goes in and fills it out with drums and pads and such. I feel like my workflow fits pretty well with advice he has given out over the years. Really love those raw trax.

Personally I am a big fan of “jam” style set ups for creating full songs, it just takes time to really learn how to make everything flow together. recording multiple full takes and fading between them can be useful at times if you are having trouble getting everything down in one go.

If you take a look at how surgeon uses an Octatrack with a fader fox he basically uses it as a Digitakt + FX processor or looper + basic mixer. I would be surprised if he ever uses the arranger on it.