Synthesizers

Generate audio from oscillators, FM synthesis, physical modeling, and more.

SubtractiveSynth DX7FMSynth SuperStackedSynth DrumSynth PluckSynth PhysicalModeling DialupSynth ChipTone

SubtractiveSynth

SubtractiveSynth

Core

Classic subtractive synthesis with oscillators, filters, ADSR envelope, and LFO modulation.

Constructor Parameters

Parameter Type Default Description
sample_rate int 44100 Audio sample rate in Hz

Oscillator Properties

Property Type Default Description
osc_wave str "sine" Waveform: "sine", "square", "saw", "triangle", "pwm", "wavetable"
pwm_depth float 0.5 Pulse width for PWM waveform (0-1)
wavetable np.array None Custom wavetable array for "wavetable" mode

ADSR Envelope Properties

Property Type Default Description
attack float 0.01 Attack time in seconds
decay float 0.1 Decay time in seconds
sustain float 0.7 Sustain level (0-1)
release float 0.2 Release time in seconds

Filter Properties

Property Type Default Description
filter_type str "lowpass" "lowpass", "highpass", or "bandpass"
filter_cutoff float 1000 Cutoff frequency in Hz (50-5000)
filter_resonance float 0.7 Resonance amount (0-5)
filter_q float 1.0 Q factor (0.1-10)

LFO Properties

Property Type Default Description
lfo_freq float 5.0 LFO frequency in Hz
lfo_depth float 0.2 LFO modulation depth (0-1)
lfo_target str "filter" "filter", "pitch", or "amplitude"

Example Usage

from audio_dsp.synth import SubtractiveSynth
import soundfile as sf

# Create synth with sawtooth oscillator
synth = SubtractiveSynth(sample_rate=44100)
synth.osc_wave = "saw"

# Configure filter
synth.filter_type = "lowpass"
synth.filter_cutoff = 800
synth.filter_resonance = 2.0

# Configure envelope
synth.attack = 0.05
synth.decay = 0.2
synth.sustain = 0.6
synth.release = 0.5

# Add LFO modulation to filter
synth.lfo_target = "filter"
synth.lfo_freq = 3.0
synth.lfo_depth = 0.4

# Synthesize a note
audio = synth.synthesize(freq=220, duration=2.0)
sf.write("subtractive_synth.wav", audio, 44100)

DX7FMSynth

DX7FMSynth

Core

Yamaha DX7-inspired FM synthesis with 4 operators, 5 algorithms, and per-operator envelopes.

Constructor / Properties

Property Type Default Description
algorithm int 1 FM algorithm (1-5)
freq_ratios list[float] [1.0, 2.0, 1.0, 0.5] Frequency ratio for each of 4 operators
mod_indices list[float] [1.0, 1.0, 1.0, 1.0] Modulation index for each operator
feedback float 0.0 Op4 feedback amount (0-5)
adsr list[dict] see below ADSR envelope per operator

FM Algorithms

Algorithm Structure Description
1 Op4 → Op3 → Op2 → Op1 Linear carrier chain
2 (Op4 → Op3) + (Op2 → Op1) Two parallel chains mixed
3 (Op4 + Op3) → Op2 → Op1 Dual modulators into chain
4 Op4 → (Op3 + Op2) → Op1 Split modulation
5 Op4 + Op3 + Op2 + Op1 All operators independent (additive)

Example Usage

from audio_dsp.synth import DX7FMSynth
import soundfile as sf

# Create FM synth
synth = DX7FMSynth()

# Configure operator ratios (harmonic series)
synth.freq_ratios = [1.0, 2.0, 3.0, 4.0]

# Set modulation indices
synth.mod_indices = [1.0, 2.0, 1.5, 0.5]

# Add feedback for richness
synth.feedback = 1.0

# Synthesize with algorithm 3
audio = synth.synthesize(freq=110, duration=2.0, algorithm=3)
sf.write("fm_synth.wav", audio, 44100)

# Electric piano style
synth.freq_ratios = [1.0, 1.0, 7.0, 14.0]
synth.mod_indices = [1.0, 0.5, 2.0, 1.0]
audio = synth.synthesize(freq=440, duration=1.5, algorithm=2)

SuperStackedSynth

SuperStackedSynth

Core

Massive unison synth with up to 10,000 detuned oscillators for huge, lush sounds.

Synthesize Parameters

Parameter Type Default Description
base_freq float required Central frequency in Hz
duration float required Length in seconds
num_oscillators int 100 Number of oscillators (1-10000)
detune_spread float 0.02 Max detune amount (0.02 = ±2%)
wave_mix dict equal mix Waveform weights {"sine": 1, "saw": 1, ...}
fade_in_time float 0.1 Staggered fade-in per oscillator
output_file str "super_synth.wav" Output file path

Example Usage

from audio_dsp.synth import SuperStackedSynth

synth = SuperStackedSynth()

# Massive pad sound
synth.synthesize(
    base_freq=220,
    duration=5.0,
    num_oscillators=200,
    detune_spread=0.03,
    wave_mix={"saw": 1.0, "square": 0.3},
    output_file="massive_pad.wav"
)

# Subtle chorus effect
synth.synthesize(
    base_freq=440,
    duration=3.0,
    num_oscillators=20,
    detune_spread=0.005,
    wave_mix={"sine": 1.0},
    output_file="subtle_chorus.wav"
)

DrumSynth

DrumSynth

Core

Synthesize kicks, snares, cymbals, claps, rims, and toms with pitch sweeps and layering.

kick()
Parameter Type Default Description
length float 0.5 Duration in seconds
max_pitch float 1000 Starting pitch in Hz
min_pitch float 50 Ending pitch in Hz
decay_factor float 10 Amplitude decay speed
signal_mix str "sine:1.0" Waveform mix "sine:1.0,fm:0.3"
snare()
Parameter Type Default Description
length float 0.3 Duration in seconds
high_pitch float 300 High tone frequency
low_pitch float 150 Low tone frequency
mix float 0.5 Noise vs tone balance (0-1)
signal_mix str "sine:1.0" Waveform mix for tone
cymbal()
Parameter Type Default Description
length float 1.0 Duration in seconds
op_a_freq float 587 FM operator A frequency
op_b_freq float 845 FM operator B frequency
noise_env float 5 Noise decay envelope
tone_env float 10 Tone decay envelope
cutoff float 8000 Filter cutoff Hz
mix float 0.5 Noise/tone mix (0-1)
clap()
Parameter Type Default Description
length float 0.3 Duration in seconds
pitch float 1000 Tone frequency
burst_count int 3 Number of burst hits
noise_mix float 0.8 Noise/tone mix (0-1)
tom() / rim()

Similar parameters to kick/snare with pitch sweep and noise mix controls.

Example Usage

from audio_dsp.synth import DrumSynth
import soundfile as sf
import numpy as np

drums = DrumSynth()

# Create a punchy kick
kick = drums.kick(
    length=0.5,
    max_pitch=800,
    min_pitch=40,
    decay_factor=15,
    signal_mix="sine:1.0,fm:0.2"
)

# Crispy snare
snare = drums.snare(
    length=0.25,
    high_pitch=350,
    low_pitch=180,
    mix=0.6
)

# Hi-hat cymbal
hihat = drums.cymbal(
    length=0.15,
    cutoff=10000,
    mix=0.3
)

# Combine into a pattern
pattern = np.concatenate([kick, snare, hihat, hihat])
sf.write("drum_pattern.wav", pattern, 44100)

Pluck Synth (Karplus-Strong)

generate_string_pluck

synth.pluck

Physical modeling string synthesis using Karplus-Strong algorithm with extended parameters.

Parameters

Parameter Type Default Description
length float required Duration in seconds
frequency float required Note frequency in Hz
intensity float 1.0 Pluck amplitude
damping float 0.5 0=bright, 1=dark (0-1)
brightness float 0.5 Filter brightness (0-1)
mute_strength float 0.5 Palm muting amount (0-1)
pitch_bend float 0.5 Pitch envelope amount (0-1)
pluck_position float 0.5 Position along string (0-1)
plectrum_size float 0.5 Pick size (0=small, 1=large)

Example Usage

from audio_dsp.synth.pluck import generate_string_pluck

# Bright acoustic guitar pluck
generate_string_pluck(
    output_file="guitar_pluck.wav",
    length=2.0,
    frequency=329.63,  # E4
    damping=0.3,
    brightness=0.7,
    pluck_position=0.2
)

# Muted bass string
generate_string_pluck(
    output_file="bass_muted.wav",
    length=1.0,
    frequency=82.41,  # E2
    damping=0.7,
    mute_strength=0.8,
    plectrum_size=0.8
)

Physical Modeling Synth

PhysicalModelingSynth

synth.physical_modeling_synth

Excite spectral models with extracted transients for realistic instrument synthesis.

Parameters

Parameter Type Default Description
frequency float required Target frequency in Hz
length float required Duration in seconds
transient_file str required Path to .transient JSON file
spectral_file str required Path to .spectral JSON file
decay_rate float 0.5 Body resonance decay speed

Dialup Modem Synth

generate_56k_dialup

synth.dialup_synth

Generate randomized 56k modem handshake sounds with authentic phases.

Parameters

Parameter Type Default Description
filename str "56k_dialup_random.wav" Output file path
duration float 8.0 Total duration in seconds
sample_rate int 44100 Sample rate in Hz

Example Usage

from audio_dsp.synth.dialup_synth import generate_56k_dialup

# Generate nostalgic dialup sounds
generate_56k_dialup(
    filename="dialup.wav",
    duration=10.0
)

Chip Tone Synths

ChipTone

synth.chip_tone

8-bit retro chiptune synthesis with square waves, arpeggiation, and classic game sounds.

Example Usage

from audio_dsp.synth.chip_tone import ChipTone

chip = ChipTone()

# Generate 8-bit melody
chip.generate_melody(
    notes=[440, 523, 587, 659],
    durations=[0.2, 0.2, 0.2, 0.4],
    output_file="chiptune.wav"
)