Unlock your creativity: just do it, then iterate

The first version is supposed to be bad. Its only job is to exist, so the second version has something to react to.

Almost every creative block is the same mistake: trying to make something good immediately, in one pass, from nothing. That is not how anything gets made.

Start badly, on purpose

Give yourself permission to make something obviously rough. A crude loop. A melody played badly. A drum pattern that is nearly right.

The moment it exists, the task changes completely — from “invent something out of nothing”, which is genuinely hard, to “improve this thing”, which is comparatively easy. Reacting is a much lower-energy activity than originating, and you can do it for hours.

Practical version: set a timer for ten minutes and make something inside it, deliberately not good. The clock removes the option of being careful.

The power of iterations

One pass produces a normal result. Five passes produce something nobody else would arrive at.

The mechanism is compounding. Each pass applies your judgement to material that already carries your previous judgement, and the distance from the starting point grows faster than you expect.

In practice:

1. Make the rough version.
2. Render it.
3. Chop it, stretch it, reverse it, filter it.
4. Render that.
5. Do it again to the result.

By the third generation you are working with material that has no obvious origin, which is exactly where original-sounding work comes from.

Render, chop, stretch, destroy

This is the most useful creative technique in modern production, and it is mechanical rather than mystical.

Render. Commit the sound to audio. Now it is material rather than a settings page, and you can treat it the way you would treat a sample. It also frees the CPU — see printing effects to audio.

Chop. Cut it at transients. Rearrange the pieces. Play them out of order. Use one fragment and discard the rest — often the best two hundred milliseconds of a bar is the whole idea.

Stretch. Slow it to a tenth speed. Time-stretch artefacts are not always damage; they are frequently the most interesting part of the result.

Destroy. Bitcrush, saturate, over-compress, resample at the wrong rate, run it through a reverb and keep only the tail. Then render that and use it as a new layer.

Granular playback takes this furthest — see what is granular synthesis and SPC Granular X. Spectral processing is the other route: what is spectral processing.

Constraints make decisions

Infinite options produce paralysis. Every professional workflow is full of self-imposed limits.

Try one of these per session:

• Four sounds only.
• One synth, no presets.
• Finish in two hours.
• No plugin you have not used before.
• Everything derived from one source recording.
• One key, one tempo, no changes.

The limitation is not a handicap. It removes the decisions that were consuming your attention without improving the music.

A template is the same idea applied to setup: fewer decisions before the work starts.

Separate the modes

Writing and judging are different activities, and doing them simultaneously stops both.

Writing mode: everything is allowed, nothing gets deleted, no plugins are adjusted. Speed matters more than quality.

Judging mode: listen back, decide what stays, be ruthless. This happens later, ideally on a different day.

Most people judge while writing, which is why the first eight bars of their projects are immaculate and there is nothing after them.

Finish more things

Finishing is a skill and it only improves by finishing.

An unfinished project teaches you the first twenty percent of the process over and over. A finished one, even a bad one, teaches you all of it — arrangement, mixing, mastering, the discipline of stopping.

Set an arbitrary deadline. Accept the result. Move on. See making your first song for the full path.

Your ideas are not as fragile as they feel

The fear of ruining something is what stops most experimentation. Two things dissolve it:

Save a version first. One keystroke. Now nothing you do can be lost, and “destroy it and see” becomes free.

Duplicate the track. Mangle the copy. Compare.

The safety net is trivial to build and it changes what you are willing to try.

The crazy idea is the point

The idea you dismiss because it is too strange is usually the one worth doing, precisely because nobody else is doing it.

Nobody remembers the correct version of anything. What sounds like a mistake in isolation is frequently the thing that makes a track identifiable — and it connects directly to finding your own sound, which is mostly a matter of keeping the accidents other people would have deleted.

The summary

• Make the bad first version quickly.
• Render it, then destroy the render.
• Repeat until the material has no obvious origin.
• Use constraints to remove decisions.
• Write and judge at different times.
• Finish things, including the ones you have lost faith in.
• Save a version before every experiment.

Creativity is not a state you wait for. It is what happens after the first rough version exists.