Skip to content

Node reference library

Analysis

  • CrossCorrelate: Outputs the cross-correlation of the input signal with the given buffer. If hop_size is zero, calculates the cross-correlation every sample.
  • NearestNeighbour: Nearest Neighbour.
  • OnsetDetector: Simple time-domain onset detector. Outputs an impulse when an onset is detected in the input. Maintains short-time and long-time averages. An onset is registered when the short-time average is threshold x the long-time average. min_interval is the minimum interval between onsets, in seconds.
  • VampAnalysis: Feature extraction using the Vamp plugin toolkit.

Buffer

  • BeatCutter: Cuts a buffer into segment_count segments, and stutters/jumps with the given probabilities.
  • BufferLooper: Read and write from a buffer concurrently, with controllable overdub.
  • BufferPlayer: Plays the contents of the given buffer. start_time/end_time are in seconds. When a clock signal is received, rewinds to the start_time. Set clock to 0 to prevent the buffer from being triggered immediately.
  • BufferRecorder: Records the input to a buffer. feedback controls overdub.
  • FeedbackBufferReader: Counterpart to FeedbackBufferWriter.
  • FeedbackBufferWriter: Counterpart to FeedbackBufferReader.
  • HistoryBufferWriter: Writes a rolling history buffer of a given duration. At a given moment in time, the contents of the buffer will be equal to the past N seconds of the audio generated by input. This is useful for (e.g.) a visual display of a rolling waveform or LFO window. downsample can be used to downsample the input; for example, with downsample of 10, a 1-second buffer can be used to display 10 seconds of historical audio.
  • SegmentPlayer: Trigger segments of buffer at the given list of onsets positions, in seconds. index determines the index of the onset to play back at, which can also be passed as an argument to trigger(). rate determines the playback rate, and clock can be used to retrigger based on the output of another Node. If continue_after_segment is non-zero, playback will continue after the subsequent onset.

Buffer: Granulation

  • SegmentedGranulator: Segmented Granulator.
  • Granulator: Granulator. Generates a grain from the given buffer each time a trigger is received on the clock input. Each new grain uses the given duration, amplitude, pan and rate values presented at each input at the moment the grain is created. The input buffer can be mono or stereo. If wrap is true, grain playback can wrap around the end/start of the buffer.

Control

  • MouseX: Outputs the normalised cursor X position, from 0 to 1. Currently only supported on macOS.
  • MouseY: Outputs the normalised cursor Y position, from 0 to 1. Currently only supported on macOS.
  • MouseDown: Outputs 1 if the left mouse button is down, 0 otherwise. Currently only supported on macOS.

Envelope

  • Accumulator: Accumulator with decay.
  • ADSREnvelope: Attack-decay-sustain-release envelope. Sustain portion is held until gate is zero.
  • ASREnvelope: Attack-sustain-release envelope.
  • DetectSilence: Detects blocks of silence below the threshold value. Used as an auto-free node to terminate a Patch after processing is complete.
  • Envelope: Generic envelope constructor, given an array of levels, times and curves.
  • Line: Line segment with the given start/end values, and duration (in seconds). If loop is true, repeats indefinitely. Retriggers on a clock signal.
  • RectangularEnvelope: Rectangular envelope with the given sustain duration.

FFT

  • FFTContinuousPhaseVocoder: Continuous phase vocoder. Requires an FFT* input.
  • FFTConvolve: Frequency-domain convolution, using overlap-add. Useful for convolution reverb, with the input buffer containing an impulse response. Requires an FFT* input.
  • FFTBufferPlayer: FFTBufferPlayer. Plays from a buffer of audio spectra in mag/phase format.
  • FFTContrast: FFT Contrast. Requires an FFT* input.
  • FFTCrossFade: FFT FFTCrossFade. Requires two FFT* inputs.
  • FFTLFO: FFT LFO. Requires an FFT* input.
  • FFTMagnitudePhaseArray: Fixed mag/phase array.
  • FFTRandomPhase: Randomise phase values.
  • FFTScaleMagnitudes: Randomise phase values.
  • FFTTransform: Transforms the FFT magnitude spectrum in the X axis. Requires an FFT* input.
  • FFT: Fast Fourier Transform. Takes a time-domain input, and generates a frequency-domain (FFT) output.
  • FFTFindPeaks: Find peaks in the FFT magnitude spectrum. Requires an FFT* input.
  • IFFT: Inverse Fast Fourier Transform. Requires an FFT* input, generates a time-domain output.
  • FFTLPF: FFT-based brick wall low pass filter. Requires an FFT* input.
  • FFTNoiseGate: FFT-based noise gate. Requires an FFT* input.
  • FFTPhaseVocoder: Phase vocoder. Requires an FFT* input.
  • FFTTonality: Tonality filter. Requires an FFT* input.
  • FFTZeroPhase: Remove phase information from a frequency-domain input. Requires an FFT* input.

Operators

  • Add: Add each sample of a to each sample of b. Can also be written as a + b
  • AmplitudeToDecibels: Map a linear amplitude value to decibels.
  • DecibelsToAmplitude: DecibelsToAmplitude
  • Bus: Bus is a node with a fixed number of input channels and arbitrary number of inputs, used to aggregate multiple sources. It is similar to Sum, but with a defined channel count that does not adapt to its inputs.
  • ChannelArray: Takes an array of inputs and spreads them across multiple channels of output.
  • ChannelCrossfade: Given a multichannel input, crossfades between channels based on the given position within the virtual array, producing a single-channel output.
  • ChannelMixer: Downmix a multichannel input to a lower-channel output. If num_channels is greater than one, spreads the input channels across the field. If amplitude_compensation is enabled, scale down the amplitude based on the ratio of input to output channels.
  • ChannelOffset: Offsets the input by a specified number of channels. With an N-channel input and an offset of M, the output will have M+N channels.
  • ChannelSelect: Select a subset of channels from a multichannel input, starting at offset, up to a maximum of maximum, with the given step.
  • Equal: Compares the output of a to the output of b. Outputs 1 when equal, 0 otherwise. Can also be written as a == b
  • NotEqual: Compares the output of a to the output of b. Outputs 0 when equal, 1 otherwise. Can also be written as a != b
  • GreaterThan: Compares the output of a to the output of b. Outputs 1 when a > b, 0 otherwise. Can also be written as a > b
  • GreaterThanOrEqual: Compares the output of a to the output of b. Outputs 1 when a >= b, 0 otherwise. Can also be written as a >= b
  • LessThan: Compares the output of a to the output of b. Outputs 1 when a < b, 0 otherwise. Can also be written as a < b
  • LessThanOrEqual: Compares the output of a to the output of b. Outputs 1 when a <= b, 0 otherwise. Can also be written as a <= b
  • Modulo: Outputs the value of a modulo b, per sample. Supports fractional values. Can also be written as a % b
  • Abs: Outputs the absolute value of a, per sample. Can also be written as abs(a)
  • If: Outputs value_if_true for each non-zero value of a, value_if_false for all other values.
  • Divide: Divide each sample of a by each sample of b. Can also be written as a / b
  • FrequencyToMidiNote: Map a frequency to a MIDI note (where 440Hz = A4 = 69), with floating-point output.
  • MidiNoteToFrequency: Map a MIDI note to a frequency (where 440Hz = A4 = 69), supporting floating-point input.
  • Multiply: Multiply each sample of a by each sample of b. Can also be written as a * b
  • Pow: Outputs a to the power of b, per sample. Can also be written as a ** b
  • RoundToScale: Given a frequency input, generates a frequency output that is rounded to the nearest MIDI note. (TODO: Not very well named)
  • Round: Round the input to the nearest integer value.
  • ScaleLinExp: Scales the input from a linear range (between a and b) to an exponential range (between c and d).
  • ScaleLinLin: Scales the input from a linear range (between a and b) to a linear range (between c and d).
  • SelectInput: Pass through the output of one or more inputs, based on the integer input index specified in index. Unlike ChannelSelect, inputs may be multichannel, and index can be modulated in real time.
  • Subtract: Subtract each sample of b from each sample of a. Can also be written as a - b
  • Sum: Sums the output of all of the input nodes, by sample.
  • TimeShift: TimeShift
  • Sin: Outputs sin(a), per sample.
  • Cos: Outputs cos(a), per sample.
  • Tan: Outputs tan(a), per sample.
  • Tanh: Outputs tanh(a), per sample. Can be used as a soft clipper.

Oscillators

  • Impulse: Produces a value of 1 at the given frequency, with output of 0 at all other times. If frequency is 0, produces a single impulse.
  • SawLFO: Produces a sawtooth LFO at the given frequency and phase offset, with output ranging from min to max.
  • SawOscillator: Produces a (non-band-limited) sawtooth wave, with the given frequency and phase offset. When a reset or trigger is received, resets the phase to zero.
  • SineLFO: Produces a sinusoidal LFO at the given frequency and phase offset, with output ranging from min to max.
  • SineOscillator: Produces a sine wave at the given frequency.
  • SquareLFO: Produces a pulse wave LFO with the given frequency and pulse width, ranging from min to max, where width of 0.5 is a square wave and other values produce a rectangular wave.
  • SquareOscillator: Produces a pulse wave with the given frequency and pulse width, where width of 0.5 is a square wave and other values produce a rectangular wave.
  • TriangleLFO: Produces a triangle LFO with the given frequency and phase offset, ranging from min to max.
  • TriangleOscillator: Produces a triangle wave with the given frequency.
  • Wavetable: Plays the wavetable stored in buffer at the given frequency and phase offset. sync can be used to provide a hard sync input, which resets the wavetable's phase at each zero-crossing.
  • Wavetable2D: Wavetable2D

Processors

  • Clip: Clip the input to min/max.
  • Fold: Fold the input beyond min/max, reflecting the excess back.
  • Smooth: Smooth the input with a given smoothing coefficient. When smooth = 0, applies no smoothing.
  • WetDry: Takes wet and dry inputs, and outputs a mix determined by wetness.
  • Wrap: Wrap the input beyond min/max.

Processors: Delays

  • AllpassDelay: All-pass delay, with feedback between 0 and 1. delay_time must be less than or equal to max_delay_time.
  • CombDelay: Comb delay, with feedback between 0 and 1. delay_time must be less than or equal to max_delay_time.
  • OneTapDelay: Single-tap delay line. delay_time must be less than or equal to max_delay_time.
  • Stutter: Stutters the input whenever a trigger is received on clock. Generates stutter_count repeats, with duration of stutter_time.

Processors: Distortion

  • Resample: Resampler and bit crusher. sample_rate is in Hz, bit_rate is an integer between 0 and 16.
  • SampleAndHold: Samples and holds the input each time a trigger is received on clock.
  • Squiz: Implementation of Dan Stowell's Squiz algorithm, a kind of downsampler.
  • WaveShaper: Applies wave-shaping as described in the WaveShaperBuffer buffer.

Processors: Dynamics

  • Compressor: Dynamic range compression, with optional sidechain input. When the input amplitude is above threshold, compresses the amplitude with the given ratio, following the given attack_time and release_time in seconds.
  • Gate: Outputs the input value when it is above the given threshold, otherwise zero.
  • Maximiser: Gain maximiser.
  • RMS: Outputs the root-mean-squared value of the input, in buffers equal to the graph's current buffer size.

Processors: Filters

  • BiquadFilter: Biquad filter. filter_type can be 'low_pass', 'band_pass', 'high_pass', 'notch', 'peak', 'low_shelf', 'high_shelf'. Not recommended for real-time modulation; for this, use SVFilter.
  • DCFilter: Remove low-frequency and DC content from a signal.
  • EQ: Three-band EQ.
  • MoogVCF: Simulation of the Moog ladder low-pass filter. cutoff sets the cutoff frequency; resonance should typically be between 0..1.
  • SVFilter: State variable filter. filter_type can be 'low_pass', 'band_pass', 'high_pass', 'notch', 'peak', 'low_shelf', 'high_shelf'. resonance should be between [0..1].

Processors: Panning

  • AzimuthPanner: Pan input around an equally-spaced ring of num_channels speakers. pan is the pan position from -1..+1, where 0 = centre front. width is the source's width, where 1.0 spans exactly between an adjacent pair of channels.
  • ChannelPanner: Pan the input between a linear series of channels, where pan 0 = channel 0, 1 = channel 1, etc. No wrapping is applied.
  • SpatialPanner: Implements a spatial panning algorithm, applied to a given SpatialEnvironment. Currently, only DBAP is supported.
  • StereoBalance: Takes a stereo input and rebalances it, where balance of 0 is unchanged, -1 is hard left, and 1 is hard right.
  • StereoPanner: Pans a mono input to a stereo output. pan should be between -1 (hard left) to +1 (hard right), with 0 = centre.
  • StereoWidth: Reduces the width of a stereo signal. When width = 1, input is unchanged. When width = 0, outputs a pair of identical channels both containing L+R.

Sequencing

  • ClockDivider: When given a clock input (e.g., an Impulse), divides the clock by the given factor. factor must be an integer greater than or equal to 1.
  • Counter: Count upwards from min to max, driven by clock.
  • Euclidean: Euclidean rhythm as described by Toussaint, with sequence_length (n) and num_events (k), driven by clock.
  • FlipFlop: Flips from 0/1 on each clock.
  • ImpulseSequence: Each time a clock or trigger is received, outputs the next value in sequence. At all other times, outputs zero.
  • Index: Outputs the value in list corresponding to index.
  • Latch: Initially outputs 0. When a trigger is received at set, outputs 1. When a trigger is subsequently received at reset, outputs 0, until the next set.
  • Sequence: Outputs the elements in sequence, incrementing position on each clock.
  • TriggerMult: Distribute any triggers to all output nodes.
  • TriggerRoundRobin: Relay trigger() events to a single node from the list of connected outputs, with direction determining the direction: 1 (or above) = move forwards by N, -1 = move backwards by N, 0 = stationary.

Stochastic

  • Logistic: Logistic noise.
  • PinkNoise: Pink noise, with specified low/high cutoffs.
  • RandomBrownian: Outputs Brownian noise between min/max, with a mean change of delta between samples. If a clock is passed, only generates a new value on a clock tick.
  • RandomChoice: Pick a random value from the given array. If a clock is passed, only picks a new value on a clock tick.
  • RandomCoin: Flip a coin with the given probability. If a clock is passed, only picks a new value on a clock tick.
  • RandomExponentialDist: Generate an random value following the exponential distribution. If a clock is passed, only picks a new value on a clock tick.
  • RandomExponential: Generate an random exponential value between min/max. If a clock is passed, only picks a new value on a clock tick.
  • RandomGaussian: Generate an random Gaussian value, with given mean and sigma. If a clock is passed, only picks a new value on a clock tick.
  • RandomImpulseSequence: Generates a random sequence of 0/1 bits with the given length, and the given probability each each bit = 1. The position of the sequence is incremented on each clock signal. explore and generate are trigger inputs which cause the sequence to mutate and re-generate respectively.
  • RandomImpulse: Generate random impulses at the given frequency, with either uniform or poisson distribution.
  • RandomUniform: Generates a uniformly random value between min/max. If a clock is passed, only picks a new value on a clock tick.
  • WhiteNoise: Generates whitenoise between min/max. If frequency is zero, generates at audio rate. For frequencies lower than audio rate, interpolate applies linear interpolation between values, and random_interval specifies whether new random values should be equally-spaced or randomly-spaced.
Last update: 2024-09-11
Created: 2021-06-08