How to Write Progressive Trance Melodies and Plucks (Full Workflow)

Am–F–C–G with a delay on top is not a progressive trance melody. It’s a chord loop wearing a costume. The reason an Anjunabeats or Colorize record hooks you at minute four isn’t the progression — it’s the pocket the pluck sits in, the way the lead answers instead of interrupts, and the ruthless frequency carve that lets a bell layer sparkle without stepping on the supersaw underneath.

This is the workflow — chord voicings, rhythm-first pluck design, two-synth layering, dotted delays, sidechain per element, and the automation that turns a loop into a track.

TL;DR

• Voice chords with a static top so the pluck reads as the main event, not the changes.
• Program the pluck rhythm on a single note first — rhythm is 70% of the feel.
• Layer two synths: a detuned saw body and an FM or sine bell on top, on one bus.
• Let a dotted delay play half the notes for you; keep the send quiet (-12 dB-ish).
• Sidechain pluck, pad, lead and bass with different settings — one preset for all is amateur hour.
• Carve complementary EQ dips between pluck and lead around 2–5 kHz.
• Automate filter, sends and velocity across every 16–32 bar phrase.

Voice the chords so the pluck has somewhere to live

The stock progressions — i–VI–III–VII (Am–F–C–G in A minor), i–v–VI–III, or vi–IV–I–V borrowed from pop — all work. They resolve without ever fully landing, which is exactly what you need for a track that has to travel for six or seven minutes. Argue about the numerals all you like; the audience never hears them.

What the audience does hear is voicing. Play Am–F–C–G as block triads in the same octave and every chord change lands like a door slamming, with no vertical room left for a pluck. Instead:

1. Put the root of each chord low, around C3, on beat 1.
2. Voice the top two notes tightly around C4–E4 and move them as little as possible between chords — ideally one voice stays put, one moves a step.
3. Drop a third or fifth from at least one chord. Sus2, add9 and maj7 voicings leave the door open for the pluck to walk through.

In A minor at 132 BPM, try Am(add9) – Fmaj7 – C(add9) – G6. Same emotional arc as the vanilla progression, but the top voices barely move. The pad becomes a bed instead of a co-lead.

Keep the pad patch itself boring: a detuned supersaw with a 300–500 ms attack, high-pass around 200 Hz, and a low-pass you’ll automate from roughly 2 kHz in the verse up to 8 kHz at the drop. Save the big reverb for later — a short plate around 1.5 s at ~15% wet is enough now. If the pad already sounds “finished” in isolation, it’s too much.

Rhythm first. Notes later.

This is the step almost everyone skips and it’s why their plucks sound like presets. Open a fresh MIDI clip, pick one note — A3 is fine — and program two bars of rhythm. No chord tones. No melodic movement. Just where the hits land.

Progressive plucks live on the 16th grid, but the identity of the pattern is in which 16ths you leave out. Every 16th filled and it’s a bad arp. The Anjuna signature is offbeat 8ths with a single 16th “tuck-in” pulling you back to the downbeat — a hit on the “a” of 4 that shoves the ear into bar 2. Find your version of that tuck-in and the pattern will carry the whole track.

Three 2-bar frames worth stealing (each character = one 16th; X = hit, – = rest):

• Driving: X-X-X--XX-X-X--X
• Rolling: X--XX-X-X--XX-X-
• Sparse: X-----X-X---X--X

Loop the single-note version for a full minute while you do something else — make coffee, answer an email. If it still feels good when you come back, keep it. If it’s already annoying, redesign it now. No EQ move fixes a bad rhythm.

Two synths, one bus

A single synth can’t cover the full pluck range without sounding either thin up top or muddy in the mids. You want a body and a top, from two different synthesis types so they don’t fight in the same band.

Body layer — Serum, Sylenth1 or Diva:

• Two saw oscillators, slightly detuned against each other. Serum’s per-oscillator detune around 0.20–0.25 is a fine starting point; tune by ear until the two saws just start to chorus.
• Unison 3 or 5 voices, spread narrow (around 30%). Wide unison here fights the bell layer later.
• Amp envelope: attack 1 ms, decay 180 ms, sustain 0, release around 100 ms.
• Low-pass around 4 kHz with a touch of envelope opening it briefly on each note — this is where the “pluck” articulation comes from, more than the amp shape.

Top layer — switch synthesis type. An FM patch (Serum FM tables, FM8, Operator) with a bell-like decay, or a sine plus an octave-up sine with an 80–120 ms decay and no sustain. Pan it opposite to the body’s imaging, or run it through a stereo widener working only above 3 kHz so the body stays mono-solid in the lows.

Route both layers to one bus. Reverb, delay, EQ and sidechain go on that bus, not on the individual patches. If you process the layers separately they drift out of alignment and stop reading as one instrument.

The delay is doing half the melody

Here’s the part that trips up producers coming from other genres: in progressive trance, the delay isn’t an effect on the pluck — it’s part of the pluck. You play four notes across two bars, and a dotted delay fills in the twelve rhythmic slots you didn’t touch.

Set a stereo delay on a send, not an insert. Sync it to 1/8 dotted (or 1/4 dotted for a slower, more rolling feel). Straight 1/8 delays land on your existing 16ths and thicken instead of syncopate; dotted delays land in the gaps and create motion. Feedback around 40%, wet fully up on the send. High-pass the return at 500 Hz and low-pass around 5 kHz so the repeats don’t cloud the pad or clash with the lead’s air.

Send the pluck bus in at roughly -12 dB. If the delay is obviously loud when you solo the bus, you’re at about the right level — in the full mix it will disappear into the groove exactly where you want it. Try muting every downbeat note in the pluck MIDI and letting the delay tail put it back. That gap-and-return feel is what makes the pattern sound expensive rather than busy.

Browse Progressive Trance packs at Myloops →

Write the lead as a conversation, not a solo

The single biggest mistake in progressive trance leads: playing them while the pluck is playing. The lead should live in the pluck’s negative space. Think two bars of pluck grooving, two bars of lead answering, one bar of overlap, one bar of silence. That’s a natural 8-bar phrase and it’s how nearly every memorable Anjuna-era lead is constructed.

Constraints that force the melody to be good:

1. Stay inside a 1.5–2 octave range. Wide leaps look clever in the piano roll and evaporate on a club system.
2. Use no more than five distinct pitches in your first 8-bar motif. Five. This is what gets whistled back at 2 a.m.
3. Repeat bars 1–2 in bars 3–4 with one note changed. Bars 5–8 get a longer variation that lifts.
4. Land the final note of a phrase on the root or fifth of the next chord, not the current one. That’s the pull that yanks the listener across the barline.

For the patch, a supersaw (7 unison voices, moderate detune, portamento 40–60 ms) is fine — it’s supposed to sound familiar. The melody has to be the fingerprint, not the timbre. Producers who spend a week designing a lead patch and an hour writing the melody have the ratio backwards.

Sidechain per element, not per project

Every element that overlaps the kick needs sidechain compression, but slamming the same preset across the bus is what makes a mix sound flat and pumpy at the same time. Different elements want different pump for different reasons.

The pad gets the most movement because it’s static — pumping is what gives it life. The lead gets the least because a wobbling lead melody sounds seasick. Tune release times so the compressor is fully open just before the next kick lands. At 132 BPM a quarter note is roughly 455 ms, so a release in the 100–150 ms window on the pluck keeps the tail intact without choking it.

Use an actual compressor triggered by the kick, or a shaper like LFOTool or Kickstart. A volume LFO and a real compressor sound different: the compressor’s non-linear release is a big part of the genre’s DNA and a straight LFO can’t fake it.

Carve the pluck and lead apart

Pluck and lead both live between roughly 500 Hz and 6 kHz. Without complementary EQ, one will always sound muddy while the other sounds thin, and no amount of level automation fixes it.

On the pluck bus:

• High-pass at 200 Hz. Low end is the bass’s job.
• 2–3 dB dip around 300–400 Hz to clear boxy build-up.
• 1–2 dB shelf around 2–3 kHz for presence.
• Optional 1–2 dB air shelf above 10 kHz if the top layer feels dull.

On the lead, do the mirror move: gentle dip around 2–3 kHz where the pluck is now sitting, and a small boost around 4–5 kHz where the lead should own the presence range. The two elements still occupy the same broad range, but they own different narrow bands within it. This one carve — the deliberate mirror-image EQ between pluck and lead — is the mixing move that separates released-quality progressive from demo-quality progressive. Nothing else comes close.

Two reverbs, doing two different jobs

One big reverb on the pluck erases the rhythm you spent an hour designing. Split it into two sends:

1. Short room: 0.8–1.2 s decay, no pre-delay, high-passed at 400 Hz. This is the glue reverb — it plants the pluck in a space without smearing the transients.
2. Long hall or plate: 4–6 s decay, 30–50 ms pre-delay, high-passed at 800 Hz and low-passed at 6 kHz. Send only the dry pluck in, and quietly (around -18 dB).

Now the crucial move: sidechain the long reverb return to the kick, same aggressive settings as the pad (fast attack, ~200 ms release, 6–8 dB reduction). Without this, the tail builds up bar after bar and the mix drowns by the second drop. With it, the reverb inhales on every kick and exhales between them, and the pluck sounds huge without any of the sludge.

Automate, or accept that people will skip

A static pluck patch playing the same MIDI through the same filter for 32 bars loses listeners before the drop. Automation is what turns a two-bar loop into a track that carries an arrangement.

At minimum, ride these:

• Low-pass filter on the pluck bus: opening from around 2 kHz at the top of a section to 8 kHz just before the drop.
• Dotted delay send: -18 dB at the start, climbing to -9 dB over 16 bars, then pulled back to reset the ear.
• Long reverb send: only opening in the last 4 bars of a 16-bar phrase, as a transition cue.
• MIDI velocity: no two notes identical. Randomise ±10, then draw an accent on every downbeat by hand. This is the “humanization” that presets can’t fake.

The pluck notes stay the same. Everything around them keeps moving. That’s how you get six minutes of momentum from a two-bar idea.

Browse Progressive Trance packs at Myloops →

The second drop needs a counter-melody

By the second drop your pluck and lead have been in the listener’s head for minutes. Fatigue is real, and just making things louder is a beginner’s answer. The professional answer is a counter-melody layer sitting above the lead.

Grab a bell-ish patch — an FM cell, or a sine plus octave sine with a fast attack and roughly 300 ms decay. Take your main pluck pattern, transpose it up an octave, then delete about two-thirds of the notes so only four hits remain across a 4-bar phrase, landing on strong beats. Pan it opposite the main pluck, drench it in the short room reverb (not the long one), and set it around -12 dB below the main pluck.

You now have three melodic layers stacked vertically: pluck in the mids, lead in the mid-highs, counter in the highs. Each owns its own frequency window and its own rhythmic slot. That’s the density progressive trance is famous for — and it comes from restraint, not from adding more parts.

The bottom line

Progressive trance plucks sound expensive because of restraint. The rhythm gets designed before the notes. The pluck is really two layers plus a dotted delay doing melodic work you didn’t have to write. The lead answers the pluck instead of talking over it. Each element gets its own sidechain shape and its own narrow band in the carve. The reverb tail pumps in time with the kick instead of washing over it.

None of the individual moves are exotic. A saw pluck, a supersaw lead, a ping-pong delay, a sidechain compressor — the ingredients are boring. What separates a released track from a demo is how carefully these boring ingredients are stacked, carved and automated. The chord progression is almost never the reason a track works. The 180 ms decay on the pluck and the -12 dB dotted delay send probably are.

Build one 8-bar loop where pluck, lead, pad and kick lock together and breathe with each other. Don’t move on until it feels inevitable. Once that loop exists, the arrangement — breakdown, second drop, outro — writes itself around it.