7.12.12. STRUDEL-12 — Synthesis

daStrudel ships five built-in oscillators, two noise generators, and a two-operator FM engine — enough to produce most classic subtractive and FM timbres without loading a single sample. Every source is selected the same way: s("name") picks the voice, and the usual ADSR / filter chain shapes the result.

7.12.12.1. Part A: Oscillators

s("sine" | "triangle" | "square" | "sawtooth" | "supersaw") picks the oscillator. They share one API — the same attack / decay / sustain / release and filter words apply to all of them:

require strudel/strudel

let pat <- s("<sine triangle square sawtooth>") |> note(48.0) |> attack(0.02) |> decay(0.1) |> sustain(0.7) |> release(0.3) |> lpf(2000.0)

Tonal character:

sine — pure, no harmonics
triangle — soft, odd harmonics, gentle roll-off
square — hollow and woody, odd harmonics only
sawtooth — bright and buzzy, all harmonics

7.12.12.2. Part B: Supersaw

supersaw is seven detuned saws stacked into one voice — the classic “trance” pad sound. It is much wider and fatter than a single saw; a low-pass filter tames the top end:

let pat <- note("c3,eb3,g3,c4", "supersaw") |> attack(0.1) |> decay(0.1) |> sustain(0.8) |> release(0.6) |> lpf(3500.0)

Commas inside note() build chords — all four notes sound simultaneously per step.

7.12.12.3. Part C: Noise

"white" and "pink" are pitch-less noise generators. Pink noise has more energy in the low end — perceptually “warmer” than white. Run them through filters to shape specific colours (ocean, wind, static, hi-hat):

let pat <- atom("pink") |> sustain(1.0) |> lpf(400.0) |> gain(0.4)

atom() wraps a plain string as a one-step pattern — useful when you want a continuous source with no rhythm.

7.12.12.4. Part D: FM Synthesis

FM modulates a sine carrier’s frequency with a second sine. Two knobs control the timbre:

fm(index) — modulation depth: higher = more harmonics, brighter sound
fmh(harmonicity) — ratio of modulator to carrier frequency

Integer harmonicity values (1, 2, 3) sound musical; non-integer values produce inharmonic, bell- or clang-like timbres:

// Metallic bell — high inharmonic ratio
let bells <- note("c5", "sine") |> fm(1.0) |> fmh(7.0) |> decay(0.3) |> release(0.8)

// Gentle brass-like tone — integer ratio
let brass <- note("<c3 e3 g3 c4>", "sine") |> fm(2.0) |> fmh(2.0) |> attack(0.02) |> decay(0.2) |> sustain(0.6) |> release(0.3)

The note("c5", "sine") form is shorthand for s("sine") |> note("c5") — the second argument picks the oscillator.

See Lambdas and Closures for how |> threads a pattern through a chain of modifiers — each |> returns a new Pattern.

7.12.12.5. Part E: Shaping FM timbre with `fm` and `fmh`

The two FM knobs interact. fm(index) is modulation depth — more index means a brighter, busier spectrum. fmh(harmonicity) is the modulator-to-carrier ratio, which sets where the partials land: integer ratios stay harmonic (brass, organ), while non-integer ratios smear the partials into inharmonic, metallic, or industrial tones. Compare a harsh non-integer ratio against a clean integer one:

// Harsh, inharmonic — non-integer ratio
let harsh <- note("c3", "sine") |> fm(5.0) |> fmh(1.4) |> attack(0.01) |> decay(0.2) |> release(0.3)

// Bright but musical — integer ratio
let bright <- note("c3", "sine") |> fm(3.0) |> fmh(3.0) |> attack(0.01) |> decay(0.1) |> sustain(0.8) |> release(0.3)

7.12.12.6. Part F: Direct-Hz synthesis with `freq`

note() derives the oscillator frequency from a MIDI number; freq() sets the frequency directly in Hz, bypassing the note-name → MIDI → Hz path. Use it for raw frequency content or non-equal-tempered tunings. Here a sawtooth steps through 220 / 277 / 330 / 440 Hz (an A-major-ish chord):

let pat <- (
    s("sawtooth*4")
    |> freq(note("220 277 330 440"))
    |> lpf(2500.0) |> release(0.2) |> gain(0.5)
)