Every other guide in this set teaches a part. This one is about the order you touch them in — because the same eight elements, assembled in a different sequence, produce a track that sounds intentional or a pile of loops that sounds like nothing. Some of what follows is settled technique. Some of it is one producer's opinion. Both are labelled, so you can tell which is which.
Eight elements, one bar each, entering in whatever order you set. Load a preset or reorder the list yourself with the arrows, then press play — each stage adds its element to whatever is already sounding. The same eight parts, three different orders, three different experiences of the same loop.
Same set, different order. All three presets use exactly the same eight elements — nothing added, nothing removed. Low end first gets the groove locked before anything melodic enters, which is the order most house and techno records are actually built in. Atmosphere first starts from a mood and builds the groove underneath it — slower to get moving, but the vibe never has to be reverse-engineered later. Melody first puts the tune in before there is anywhere for it to sit, which is why it will sound directionless even though every element is identical to the other two.
This is opinion, usefully. Opinion Neither order is correct. What matters is picking one on purpose instead of drifting into whichever element happened to be easiest to start with that day.
What you already have decides the first move. Find your situation.
| You have | Do this first | Then open |
|---|
If none of these match, you have nothing yet — which is the easiest state to be in. Pick a kick sample and a bass patch, get four bars looping that you would nod along to with no melody at all, and build from there.
Three ways to make a loop feel alive without adding a new instrument. All three are established technique, not opinion.
Quiet, often shortened or filtered kicks placed on the 16ths between the main hits. They do not compete with the main kick for attention — you feel them more than hear them. The number below scores the ghost pattern on its own, the same way every syncopation figure in this set scores one part at a time — measuring it against the main kick would be a different, mostly meaningless number, since the two are locked together on purpose.
Rolling off the highs on a chord or pad is the fastest route to the "warm" sound deep house is named for. This is measured on the actual rendered audio, not asserted — the numbers below come from analysing the two buffers.
Two ways to make a 16-bar loop feel like it's building. Version A introduces a new instrument every four bars — the "1, 2, 3, here's a new part" pattern almost everyone starts with. Version B keeps the same instruments present from bar one and only automates a filter and a volume. Count what each one actually adds.
Split your automation by register. If you only automate one thing, automate the top of a chord or arp part while the bass note underneath stays still — the movement reads clearly instead of the whole part just getting quieter or brighter as a block.
Most house tracks live on one melodic idea. Established When a section needs more energy, adding percussion usually works better than adding a second melodic part — because percussion adds nothing for the ear to reconcile with what is already there.
Both raise the energy. The difference is what they cost: Version A adds zero new pitch classes to the harmony — nothing to clash. Version B adds a real melodic line the ear now has to reconcile with the first one, and if the two do not agree they will fight, which is exactly what the Melody & Harmony guide measures directly.
One producer's rule of thumb is to spend roughly four times longer listening than producing. Opinion — no study backs a specific ratio, and it will vary by how experienced you are. What is not in dispute is that reference tracks work, and that dissecting one properly takes a structure, not just a replay button.
Fill this in while listening, not after. The point is not the document — it is that answering these questions forces the kind of attention that "listening critically" actually means. Do it for one track before every session for a month and the pattern-matching becomes automatic.
Opinion One producer's habit: do not sit down to write unless there is a block of time to finish a draft, and treat three to four hours as that block. The reasoning is real even if the exact number is personal — a track worked on for three weeks loses you your own objectivity about whether it is good.
The phase split is a starting point, not a rule. Sound selection light, low end and groove heaviest, then arrangement, then a short automation pass, then stop and bounce — whether or not it feels finished. A worse-but-finished draft teaches you more than a better-but-abandoned one, and you can always come back to it in the next session with fresh ears.
Opinion, but close to universal among working producers Finding something unique to your production is the hardest thing on this page and the least teachable — it is not necessary for a functional track, but it is what separates your tracks from everyone else's once the groove is already solid. Nobody can hand you this. What follows is a way to force the question instead of waiting for the answer.
The idea is not the constraint itself — it is that a forced limitation produces a decision that is yours specifically, made under pressure, rather than the default setting on a plugin. Sing something badly into a phone mic and pitch-correct it wrong on purpose. That mistake, kept, is a voice. A preset, however good, never is.
This guide is different from the other twelve. They teach mechanisms you can verify — a semitone count, a syncopation score, a peak level dropping under phase cancellation. Half of this one is the same kind of claim. The other half is one experienced producer's opinion about workflow, clearly marked as such, included because it is useful even though it cannot be proven the way the rest of this set can.
All audio synthesised live. The warmth lab's frequency figures are measured on the actual rendered buffers, not asserted.