Analysis

Traits

trait GetFloat64Featurable

GetFloat64Featurable Required Methods

trait GetFloat64Featurable . fn get_features

Signature

def get_features(self) -> List[Float64]

Returns

Type: List[Float64]

---

trait GetBoolFeaturable

GetBoolFeaturable Required Methods

trait GetBoolFeaturable . fn get_features

Signature

def get_features(self) -> List[Bool]

Returns

Type: List[Bool]

---

Structs

struct YIN

Monophonic Frequency ('F0') Detection using the YIN algorithm (FFT-based, O(N log N) version).

Traits: AnyType, BufferedProcessable, Copyable, Deinitable, GetFloat64Featurable, Movable

---

YIN Functions

struct YIN . fn init

Initialize the YIN pitch detector.

fn init Signature

def __init__(out self, sr: Float64, window_size: Int = Int(1024), min_freq: Float64 = 20, max_freq: Float64 = 2.0E+4)

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate from the MMMWorld.
window_size	Int	Int(1024)	The size of the analysis window in samples.
min_freq	Float64	20	The minimum frequency to consider for pitch detection.
max_freq	Float64	2.0E+4	The maximum frequency to consider for pitch detection.

fn init Returns : Self An initialized YIN struct.

Static Method

This is a static method.

struct YIN . fn get_features

Get the pitch and confidence values.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current pitch and confidence values as a List[Float64].

struct YIN . fn next_window

Compute the YIN pitch estimate for the given frame of audio samples. Nothing is returned. The pitch and confidence are stored internally and can be accessed with .pitch and .confidence.

fn next_window Signature

def next_window(mut self, mut frame: List[Float64])

fn next_window Arguments

Name	Type	Default	Description
frame	List[Float64]	—	The input audio frame of size window_size. This List gets passed from BufferedProcess.

struct YIN . fn buf_analysis

Compute the YIN pitch and confidence values for a given audio buffer. This static method is useful when there is an audio buffer already loaded and you want to compute the YIN pitch over it.

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, window_size: Int = Int(1024), hop_size: Int = Int(512), min_freq: Float64 = 20, max_freq: Float64 = 2.0E+4, padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	The input audio buffer.
chan	Int	Int(0)	The channel index to analyze (default is 0).
start_frame	Int	Int(0)	The starting frame index in the buffer (default is 0).
num_frames	Optional[Int]	None	The number of frames to analyze (default is None, which means analyze until the end of the buffer).
window_size	Int	Int(1024)	The size of the analysis window in samples.
hop_size	Int	Int(512)	The hop size between windows in samples.
min_freq	Float64	20	The minimum frequency to consider for pitch detection.
max_freq	Float64	2.0E+4	The maximum frequency to consider for pitch detection.
padding	Padding	Padding.half_window	The padding strategy to use for the analysis.

fn buf_analysis Returns : List[List[Float64]] A List of Lists of Float64 containing the pitch and confidence values for each analyzed frame.

Raises An error if the analysis fails for any reason.

Static Method

This is a static method.

struct SpectralCentroid

Spectral Centroid analysis.

Based on the Peeters (2003)

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

SpectralCentroid Functions

struct SpectralCentroid . fn init

Initialize the Spectral Centroid analyzer.

fn init Signature

def __init__(out self, sr: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), power_mag: Bool = False)

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate from the MMMWorld.
min_freq	Float64	Float64("20")	The minimum frequency to consider when computing the spectral centroid.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider when computing the spectral centroid.
power_mag	Bool	False	Whether to use power magnitudes (mags^2) instead of linear magnitudes when computing the centroid.

fn init Returns : Self An initialized SpectralCentroid struct.

Static Method

This is a static method.

struct SpectralCentroid . fn get_features

Get the spectral centroid value as a List of Float64.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current spectral centroid feature value as a List[Float64] (with only the one element).

struct SpectralCentroid . fn next_frame

Compute the spectral centroid for a given FFT analysis. This function is to be used by FFTProcess if SpectralCentroid is passed as the "process".

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct SpectralCentroid . fn from_mags

Compute the spectral centroid for the given magnitudes of an FFT frame. This static method is useful when there is an FFT already computed, perhaps as part of a custom struct that implements the FFTProcessable trait.

fn from_mags Signature

def from_mags(mags: List[Float64], sample_rate: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), power_mag: Bool = False) -> Float64

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
sample_rate	Float64	—	The sample rate of the audio signal.
min_freq	Float64	Float64("20")	The minimum frequency to consider when computing the spectral centroid.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider when computing the spectral centroid.
power_mag	Bool	False	Whether to use power magnitudes (mags^2) instead of linear magnitudes when computing the centroid.

fn from_mags Returns : Float64 Float64. The spectral centroid value.

Static Method

This is a static method.

struct SpectralCentroid . fn buf_analysis

Compute the spectral centroid for a given audio buffer. This static method is useful when there is an audio buffer already loaded and you want to compute the spectral centroid over it.

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), power_mag: Bool = False, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	The input audio buffer.
chan	Int	Int(0)	The channel index to analyze (default is 0).
start_frame	Int	Int(0)	The starting frame index in the buffer (default is 0).
num_frames	Optional[Int]	None	The number of frames to analyze (default is None, which means analyze until the end of the buffer).
min_freq	Float64	Float64("20")	The minimum frequency to consider when computing the spectral centroid.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider when computing the spectral centroid.
power_mag	Bool	False	Whether to use power magnitudes (mags^2) instead of linear magnitudes when computing the centroid.
window_size	Int	Int(1024)	The size of the analysis window in samples.
hop_size	Int	Int(512)	The hop size between windows in samples.
padding	Padding	Padding.half_window	The padding strategy to use for the analysis.

fn buf_analysis Returns : List[List[Float64]] A List of Lists of Float64 containing the spectral centroid values for each analyzed frame.

Raises An error if the analysis fails for any reason.

Static Method

This is a static method.

struct SpectralSpread

Spectral Spread analysis.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

SpectralSpread Functions

struct SpectralSpread . fn init

Initialize the Spectral Spread analyzer.

fn init Signature

def __init__(out self, sr: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False)

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate from the MMMWorld.
min_freq	Float64	Float64("20")	The minimum frequency to consider.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2).

fn init Returns : Self

Static Method

This is a static method.

struct SpectralSpread . fn get_features

Get the current spectral spread value.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current spectral spread feature value as a List[Float64] (with only the one element).

struct SpectralSpread . fn next_frame

Compute the spectral spread for a given FFT analysis. This function is to be used by FFTProcess if SpectralSpread is passed as the "process".

Nothing is returned. The spread is stored internally and can be accessed with .spread.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct SpectralSpread . fn from_mags

fn from_mags Signature

def from_mags(mags: List[Float64], sample_rate: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False) -> Float64

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	—
sample_rate	Float64	—	—
min_freq	Float64	Float64("20")	—
max_freq	Float64	Float64("2.0E+4")	—
log_freq	Bool	False	—
power_mag	Bool	False	—

fn from_mags Returns : Float64

Static Method

This is a static method.

struct SpectralSpread . fn buf_analysis

Compute the spectral spread for a given audio buffer.

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	The input audio buffer.
chan	Int	Int(0)	The channel index to analyze (default is 0).
start_frame	Int	Int(0)	The starting frame index in the buffer (default is 0).
num_frames	Optional[Int]	None	The number of frames to analyze (default is None, which means analyze until the end of the buffer).
min_freq	Float64	Float64("20")	The minimum frequency to consider when computing the spectral spread.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider when computing the spectral spread.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2) instead of linear magnitudes when computing the spread.
window_size	Int	Int(1024)	The size of the analysis window in samples.
hop_size	Int	Int(512)	The hop size between windows in samples.
padding	Padding	Padding.half_window	The padding strategy to use for the analysis.

fn buf_analysis Returns : List[List[Float64]] A List of Lists of Float64 containing the spectral spread values for each analyzed frame.

Raises An error if the analysis fails for any reason.

Static Method

This is a static method.

struct SpectralSkewness

Spectral Skewness analysis.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

SpectralSkewness Functions

struct SpectralSkewness . fn init

Initialize the Spectral Skewness analyzer.

fn init Signature

def __init__(out self, sr: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False)

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate from the MMMWorld.
min_freq	Float64	Float64("20")	The minimum frequency to consider.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2).

fn init Returns : Self

Static Method

This is a static method.

struct SpectralSkewness . fn get_features

Return the current spectral skewness value as a List of Float64.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current spectral skewness feature value as a List[Float64] (with only the one element).

struct SpectralSkewness . fn next_frame

Compute the spectral skewness for a given FFT analysis. Nothing is returned. The skewness is stored internally and can be accessed with .skewness.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct SpectralSkewness . fn from_mags

fn from_mags Signature

def from_mags(mags: List[Float64], sample_rate: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False) -> Float64

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	—
sample_rate	Float64	—	—
min_freq	Float64	Float64("20")	—
max_freq	Float64	Float64("2.0E+4")	—
log_freq	Bool	False	—
power_mag	Bool	False	—

fn from_mags Returns : Float64

Static Method

This is a static method.

struct SpectralSkewness . fn buf_analysis

Compute the spectral skewness for a given audio buffer.

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	The input audio buffer.
chan	Int	Int(0)	The channel index to analyze (default is 0).
start_frame	Int	Int(0)	The starting frame index in the buffer (default is 0).
num_frames	Optional[Int]	None	The number of frames to analyze (default is None, which means analyze until the end of the buffer).
min_freq	Float64	Float64("20")	The minimum frequency to consider when computing the spectral skewness.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider when computing the spectral skewness.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2) instead of linear magnitudes when computing the skewness.
window_size	Int	Int(1024)	The size of the analysis window in samples.
hop_size	Int	Int(512)	The hop size between windows in samples.
padding	Padding	Padding.half_window	The padding strategy to use for the analysis.

fn buf_analysis Returns : List[List[Float64]] A List of Lists of Float64 containing the spectral skewness values for each analyzed frame.

Raises Raises an error if the analysis fails for any reason.

Static Method

This is a static method.

struct SpectralKurtosis

Spectral Kurtosis analysis.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

SpectralKurtosis Functions

struct SpectralKurtosis . fn init

Initialize the Spectral Kurtosis analyzer.

fn init Signature

def __init__(out self, sr: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False)

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate from the MMMWorld.
min_freq	Float64	Float64("20")	The minimum frequency to consider.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2).

fn init Returns : Self

Static Method

This is a static method.

struct SpectralKurtosis . fn get_features

Get the current spectral kurtosis value as a List of Float64.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current spectral kurtosis feature value.

struct SpectralKurtosis . fn next_frame

Compute the spectral kurtosis for a given FFT analysis. This function is to be used by FFTProcess if SpectralKurtosis is passed as the "process".

Nothing is returned. The kurtosis is stored internally and can be accessed with .kurtosis.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct SpectralKurtosis . fn from_mags

fn from_mags Signature

def from_mags(mags: List[Float64], sample_rate: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False) -> Float64

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	—
sample_rate	Float64	—	—
min_freq	Float64	Float64("20")	—
max_freq	Float64	Float64("2.0E+4")	—
log_freq	Bool	False	—
power_mag	Bool	False	—

fn from_mags Returns : Float64

Static Method

This is a static method.

struct SpectralKurtosis . fn buf_analysis

Compute the spectral kurtosis for a given audio buffer.

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	The input audio buffer.
chan	Int	Int(0)	The channel index to analyze (default is 0).
start_frame	Int	Int(0)	The starting frame index in the buffer (default is 0).
num_frames	Optional[Int]	None	The number of frames to analyze (default is None, which means analyze until the end of the buffer).
min_freq	Float64	Float64("20")	The minimum frequency to consider when computing the spectral kurtosis.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider when computing the spectral kurtosis.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2) instead of linear magnitudes when computing the kurtosis.
window_size	Int	Int(1024)	The size of the analysis window in samples.
hop_size	Int	Int(512)	The hop size between windows in samples.
padding	Padding	Padding.half_window	The padding strategy to use for the analysis.

fn buf_analysis Returns : List[List[Float64]] A List of Lists of Float64 containing the spectral kurtosis values for each analyzed frame.

Raises Error: If analysis fails.

Static Method

This is a static method.

struct SpectralRolloff

Spectral Rolloff analysis.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

SpectralRolloff Functions

struct SpectralRolloff . fn init

Initialize the Spectral Rolloff analyzer.

fn init Signature

def __init__(out self, sr: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), rolloff_target: Float64 = 95, log_freq: Bool = False, power_mag: Bool = False)

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate from the MMMWorld.
min_freq	Float64	Float64("20")	The minimum frequency to consider.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider.
rolloff_target	Float64	95	Percentage of spectral energy for rolloff.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2).

fn init Returns : Self

Static Method

This is a static method.

struct SpectralRolloff . fn get_features

Get the current spectral rolloff value.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current spectral rolloff feature value as a List[Float64] (with only the one element).

struct SpectralRolloff . fn next_frame

Compute the spectral rolloff for a given FFT analysis. This function is to be used by FFTProcess if SpectralRolloff is passed as the "process".

Nothing is returned. The rolloff is stored internally and can be accessed with .rolloff.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct SpectralRolloff . fn from_mags

fn from_mags Signature

def from_mags(mags: List[Float64], sample_rate: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), rolloff_target: Float64 = 95, log_freq: Bool = False, power_mag: Bool = False) -> Float64

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	—
sample_rate	Float64	—	—
min_freq	Float64	Float64("20")	—
max_freq	Float64	Float64("2.0E+4")	—
rolloff_target	Float64	95	—
log_freq	Bool	False	—
power_mag	Bool	False	—

fn from_mags Returns : Float64

Static Method

This is a static method.

struct SpectralRolloff . fn buf_analysis

Compute the spectral rolloff for a given audio buffer.

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), rolloff_target: Float64 = 95, log_freq: Bool = False, power_mag: Bool = False, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	The input audio buffer.
chan	Int	Int(0)	The channel index to analyze (default is 0).
start_frame	Int	Int(0)	The starting frame index in the buffer (default is 0).
num_frames	Optional[Int]	None	The number of frames to analyze (default is None, which means analyze until the end of the buffer).
min_freq	Float64	Float64("20")	The minimum frequency to consider when computing the spectral rolloff.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider when computing the spectral rolloff.
rolloff_target	Float64	95	Percentage of spectral energy for rolloff.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2) instead of linear magnitudes when computing the rolloff.
window_size	Int	Int(1024)	The size of the analysis window in samples.
hop_size	Int	Int(512)	The hop size between windows in samples.
padding	Padding	Padding.half_window	The padding strategy to use for the analysis.

fn buf_analysis Returns : List[List[Float64]] A List of Lists of Float64 containing the spectral rolloff values for each analyzed frame.

Raises Error: If analysis fails.

Static Method

This is a static method.

struct SpectralFlatness

Spectral Flatness analysis.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

SpectralFlatness Functions

struct SpectralFlatness . fn init

Initialize the Spectral Flatness analyzer.

fn init Signature

def __init__(out self, sr: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False)

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate from the MMMWorld.
min_freq	Float64	Float64("20")	The minimum frequency to consider.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2).

fn init Returns : Self

Static Method

This is a static method.

struct SpectralFlatness . fn get_features

Get the current spectral flatness value (dB).

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current spectral flatness feature value as a List[Float64] (with only the one element).

struct SpectralFlatness . fn next_frame

Compute the spectral flatness for a given FFT analysis. This function is to be used by FFTProcess if SpectralFlatness is passed as the "process".

Nothing is returned. The flatness is stored internally and can be accessed with .flatness.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct SpectralFlatness . fn from_mags

fn from_mags Signature

def from_mags(mags: List[Float64], sample_rate: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False) -> Float64

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	—
sample_rate	Float64	—	—
min_freq	Float64	Float64("20")	—
max_freq	Float64	Float64("2.0E+4")	—
log_freq	Bool	False	—
power_mag	Bool	False	—

fn from_mags Returns : Float64

Static Method

This is a static method.

struct SpectralFlatness . fn buf_analysis

Compute the spectral flatness for a given audio buffer.

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	The input audio buffer.
chan	Int	Int(0)	The channel index to analyze (default is 0).
start_frame	Int	Int(0)	The starting frame index in the buffer (default is 0).
num_frames	Optional[Int]	None	The number of frames to analyze (default is None, which means analyze until the end of the buffer).
min_freq	Float64	Float64("20")	The minimum frequency to consider when computing the spectral flatness.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider when computing the spectral flatness.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2) instead of linear magnitudes when computing the flatness.
window_size	Int	Int(1024)	The size of the analysis window in samples.
hop_size	Int	Int(512)	The hop size between windows in samples.
padding	Padding	Padding.half_window	The padding strategy to use for the analysis.

fn buf_analysis Returns : List[List[Float64]] A List of Lists of Float64 containing the spectral flatness values for each analyzed frame.

Raises Error: If analysis fails.

Static Method

This is a static method.

struct SpectralCrest

Spectral Crest analysis.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

SpectralCrest Functions

struct SpectralCrest . fn init

Initialize the Spectral Crest analyzer.

fn init Signature

def __init__(out self, sr: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False)

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate from the MMMWorld.
min_freq	Float64	Float64("20")	The minimum frequency to consider.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2).

fn init Returns : Self

Static Method

This is a static method.

struct SpectralCrest . fn get_features

Get the current spectral crest value (dB).

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current spectral crest feature value as a List[Float64] (with only the one element).

struct SpectralCrest . fn next_frame

Compute the spectral crest for a given FFT analysis. This function is to be used by FFTProcess if SpectralCrest is passed as the "process".

Nothing is returned. The crest is stored internally and can be accessed with .crest.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct SpectralCrest . fn from_mags

fn from_mags Signature

def from_mags(mags: List[Float64], sample_rate: Float64, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False) -> Float64

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	—
sample_rate	Float64	—	—
min_freq	Float64	Float64("20")	—
max_freq	Float64	Float64("2.0E+4")	—
log_freq	Bool	False	—
power_mag	Bool	False	—

fn from_mags Returns : Float64

Static Method

This is a static method.

struct SpectralCrest . fn buf_analysis

Compute the spectral crest for a given audio buffer. This static method is useful when there is an audio buffer already loaded and you want to compute the spectral crest over it.

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, min_freq: Float64 = Float64("20"), max_freq: Float64 = Float64("2.0E+4"), log_freq: Bool = False, power_mag: Bool = False, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	The input audio buffer.
chan	Int	Int(0)	The channel index to analyze (default is 0).
start_frame	Int	Int(0)	The starting frame index in the buffer (default is 0).
num_frames	Optional[Int]	None	The number of frames to analyze (default is None, which means analyze until the end of the buffer).
min_freq	Float64	Float64("20")	The minimum frequency to consider when computing the spectral crest.
max_freq	Float64	Float64("2.0E+4")	The maximum frequency to consider when computing the spectral crest.
log_freq	Bool	False	Whether to use log-frequency (MIDI) bins.
power_mag	Bool	False	Whether to use power magnitudes (mags^2) instead of linear magnitudes when computing the crest.
window_size	Int	Int(1024)	The size of the analysis window in samples.
hop_size	Int	Int(512)	The hop size between windows in samples.
padding	Padding	Padding.half_window	The padding strategy to use for the analysis.

fn buf_analysis Returns : List[List[Float64]] A List of Lists of Float64 containing the spectral crest values for each analyzed frame.

Raises An error if the analysis fails for any reason.

Static Method

This is a static method.

struct RMS

Root Mean Square (RMS) amplitude analysis.

Traits: AnyType, BufferedProcessable, Copyable, Deinitable, GetFloat64Featurable, Movable

---

RMS Functions

struct RMS . fn init

Initialize the RMS analyzer.

fn init Signature

def __init__(out self)

fn init Returns : Self

Static Method

This is a static method.

struct RMS . fn get_features

Get the current RMS value.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current RMS feature value as a List[Float64] (with only the one element).

struct RMS . fn next_window

Compute the RMS for the given window of audio samples. This function is to be used with a BufferedProcess.

The computed RMS value is stored in self.rms. fn next_window Signature

def next_window(mut self, mut input: List[Float64])

fn next_window Arguments

Name	Type	Default	Description
input	List[Float64]	—	The input audio frame of samples. This List gets passed from BufferedProcess.

struct RMS . fn from_window

Compute the RMS for the given window of audio samples. This static method is useful when there is an audio frame already available, perhaps as part of a custom struct that implements the BufferedProcessable trait.

fn from_window Signature

def from_window(mut frame: List[Float64]) -> Float64

fn from_window Arguments

Name	Type	Default	Description
frame	List[Float64]	—	The input audio frame of samples.

fn from_window Returns : Float64 Float64. The computed RMS value.

Static Method

This is a static method.

struct RMS . fn buf_analysis

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	—
chan	Int	Int(0)	—
start_frame	Int	Int(0)	—
num_frames	Optional[Int]	None	—
window_size	Int	Int(1024)	—
hop_size	Int	Int(512)	—
padding	Padding	Padding.half_window	—

fn buf_analysis Returns : List[List[Float64]]

Raises

Static Method

This is a static method.

struct MelBands

Mel Bands analysis.

This implementation follows the approach used in the Librosa library.

The Mel scale is a perceptual scale of pitches that approximates the human ear's response more closely than the linear frequency scale. Mel Bands analysis involves mapping the FFT frequency bins to the Mel scale and computing the energy in each Mel band. This way the "magnitudes" of each Mel band represent a, roughly, equal amount of perceptual frequency space (unlike the FFT).

Because the definition of the mel scale is conditioned by a finite number of subjective psychoacoustical experiments, several implementations coexist in the audio signal processing literature. MMMAudio replicates the default of Librosa, which replicates the behavior of the well-established MATLAB "Auditory Toolbox" of Slaney (citation below). According to this implementation, the conversion from Hertz to mel is linear below 1 kHz and logarithmic above 1 kHz. Additionally, the weights are normalized such that the area under each mel filter is equal. Slaney mel filter "triangles" all have equal area (visualization), which helps to ensure that the energy in each mel band is comparable.

Slaney, M. Auditory Toolbox: A MATLAB Toolbox for Auditory Modeling Work. Technical Report, version 2, Interval Research Corporation, 1998.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

MelBands Functions

struct MelBands . fn init

Initialize the Mel Bands analyzer.

fn init Signature

def __init__(out self, sr: Float64, num_bands: Int = Int(40), min_freq: Float64 = 20, max_freq: Float64 = 2.0E+4, fft_size: Int = Int(1024), power: Float64 = 2)

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate from the MMMWorld.
num_bands	Int	Int(40)	The number of mel bands to compute.
min_freq	Float64	20	The minimum frequency (in Hz) to consider when computing the mel bands.
max_freq	Float64	2.0E+4	The maximum frequency (in Hz) to consider when computing the mel bands.
fft_size	Int	Int(1024)	The size of the FFT being used to compute the mel bands.
power	Float64	2	Exponent applied to magnitudes before mel filtering (librosa default is 2.0 for power).

fn init Returns : Self An initialized MelBands struct.

Static Method

This is a static method.

struct MelBands . fn get_features

Get the current mel band values.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current mel band feature vector.

struct MelBands . fn next_frame

Compute the mel bands for a given FFT analysis. This function is to be used by FFTProcess if MelBands is passed as the "process".

Nothing is returned from this function, but the computed mel band values are stored in self.bands.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct MelBands . fn from_mags

Compute the mel bands for a given list of magnitudes. This function is useful when there is an FFT already computed, perhaps as part of a custom struct that implements the FFTProcessable trait.

fn from_mags Signature

def from_mags(mut self, ref mags: List[Float64])

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.

struct MelBands . fn mel_frequencies

Compute an array of acoustic frequencies tuned to the mel scale. This implementation is based on Librosa's eponymous function. For more information on mel frequencies space see the MelBands documentation.

fn mel_frequencies Signature

def mel_frequencies(n_mels: Int = Int(128), fmin: Float64 = 0, fmax: Float64 = 2.0E+4) -> List[Float64]

fn mel_frequencies Arguments

Name	Type	Default	Description
n_mels	Int	Int(128)	The number of mel bands to generate.
fmin	Float64	0	The lowest frequency (in Hz).
fmax	Float64	2.0E+4	The highest frequency (in Hz).

fn mel_frequencies Returns : List[Float64] A List of Float64 representing the center frequencies of each mel band.

Static Method

This is a static method.

struct MelBands . fn hz_to_mel

Convert Hz to Mels. This implementation is based on Librosa's eponymous function. For more information on mel frequencies space see the MelBands documentation.

fn hz_to_mel Signature

def hz_to_mel[num_chans: SIMDLength = SIMDLength(1)](freq: SIMD[DType.float64, num_chans]) -> SIMD[DType.float64, num_chans]

fn hz_to_mel Parameters

Name	Type	Default	Description
num_chans	SIMDLength	SIMDLength(1)	Size of the SIMD vector. This parameter is inferred by the values passed to the function.

fn hz_to_mel Arguments

Name	Type	Default	Description
freq	SIMD[DType.float64, num_chans]	—	The frequencies in Hz to convert.

fn hz_to_mel Returns : SIMD[DType.float64, num_chans] The corresponding mel frequencies.

Static Method

This is a static method.

struct MelBands . fn mel_to_hz

Convert mel bin numbers to frequencies. This implementation is based on Librosa's eponymous function. For more information on mel frequencies space see the MelBands documentation.

fn mel_to_hz Signature

def mel_to_hz[num_chans: SIMDLength = SIMDLength(1)](mel: SIMD[DType.float64, num_chans]) -> SIMD[DType.float64, num_chans]

fn mel_to_hz Parameters

Name	Type	Default	Description
num_chans	SIMDLength	SIMDLength(1)	Number of SIMD channels in the mel vector.

fn mel_to_hz Arguments

Name	Type	Default	Description
mel	SIMD[DType.float64, num_chans]	—	Mel values to convert to Hertz.

fn mel_to_hz Returns : SIMD[DType.float64, num_chans] Frequencies in Hertz for the input mel values.

Static Method

This is a static method.

struct MelBands . fn buf_analysis

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, num_bands: Int = Int(40), min_freq: Float64 = 20, max_freq: Float64 = 2.0E+4, power: Float64 = 2, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	—
chan	Int	Int(0)	—
start_frame	Int	Int(0)	—
num_frames	Optional[Int]	None	—
num_bands	Int	Int(40)	—
min_freq	Float64	20	—
max_freq	Float64	2.0E+4	—
power	Float64	2	—
window_size	Int	Int(1024)	—
hop_size	Int	Int(512)	—
padding	Padding	Padding.half_window	—

fn buf_analysis Returns : List[List[Float64]]

Raises

Static Method

This is a static method.

struct MFCC

Mel-Frequency Cepstral Coefficients (MFCC) analysis.

This implementation follows the approach used in the Librosa library.

Learn more about MFCCs on the FluCoMa learn page and with this interactive demonstration.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

MFCC Functions

struct MFCC . fn init

Initialize the MFCC analyzer.

fn init Signature

def __init__(out self, sr: Float64, num_coeffs: Int = Int(13), num_bands: Int = Int(40), min_freq: Float64 = 20, max_freq: Float64 = 2.0E+4, fft_size: Int = Int(1024))

fn init Arguments

Name	Type	Default	Description
sr	Float64	—	The sample rate for the mel band computation.
num_coeffs	Int	Int(13)	The number of MFCC coefficients to compute (including the 0th coefficient).
num_bands	Int	Int(40)	The number of mel bands to use when computing the MFCCs.
min_freq	Float64	20	The minimum frequency (in Hz) to consider when computing the mel bands for the MFCCs.
max_freq	Float64	2.0E+4	The maximum frequency (in Hz) to consider when computing the mel bands for the MFCCs.
fft_size	Int	Int(1024)	The size of the FFT being used to compute the mel bands for the MFCCs.

fn init Returns : Self An initialized MFCC struct.

Static Method

This is a static method.

struct MFCC . fn get_features

Get the current MFCC values.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current MFCC feature vector as a List[Float64].

struct MFCC . fn next_frame

Compute the MFCCs for a given FFT analysis. This function is to be used by FFTProcess if MFCC is passed as the "process".

Nothing is returned from this function, but the computed MFCC values are stored in self.coeffs.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct MFCC . fn from_mags

Compute the MFCCs for a given list of magnitudes. This function is useful when there is an FFT already computed, perhaps as part of a custom struct that implements the FFTProcessable trait.

Nothing is returned from this function, but the computed MFCC values are stored in self.coeffs.

fn from_mags Signature

def from_mags(mut self, ref mags: List[Float64])

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.

struct MFCC . fn from_mel_bands

Compute the MFCCs for a given list of mel band energies. This function is useful when there is a mel band analysis already computed, perhaps as part of a custom struct that implements the FFTProcessable trait.

Nothing is returned from this function, but the computed MFCC values are stored in self.coeffs.

fn from_mel_bands Signature

def from_mel_bands(mut self, ref mbands: List[Float64])

fn from_mel_bands Arguments

Name	Type	Default	Description
mbands	List[Float64]	—	The input mel band energies as a List of Float64.

struct MFCC . fn buf_analysis

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, num_coeffs: Int = Int(13), num_bands: Int = Int(40), min_freq: Float64 = 20, max_freq: Float64 = 2.0E+4, window_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	—
chan	Int	Int(0)	—
start_frame	Int	Int(0)	—
num_frames	Optional[Int]	None	—
num_coeffs	Int	Int(13)	—
num_bands	Int	Int(40)	—
min_freq	Float64	20	—
max_freq	Float64	2.0E+4	—
window_size	Int	Int(1024)	—
hop_size	Int	Int(512)	—
padding	Padding	Padding.half_window	—

fn buf_analysis Returns : List[List[Float64]]

Raises

Static Method

This is a static method.

struct DCT

Compute the Discrete Cosine Transform (DCT).

Traits: AnyType, Copyable, Deinitable, Movable

---

DCT Functions

struct DCT . fn init

fn init Signature

def __init__(out self, input_size: Int, output_size: Int)

fn init Arguments

Name	Type	Default	Description
input_size	Int	—	—
output_size	Int	—	—

fn init Returns : Self

Static Method

This is a static method.

struct DCT . fn process

Compute the first output_size DCT-II coefficients for input. Nothing is returned from this function, but the computed DCT coefficients are stored in the output List passed as an argument.

fn process Signature

def process(mut self, ref input: List[Float64], mut output: List[Float64])

fn process Arguments

Name	Type	Default	Description
input	List[Float64]	—	Input vector of length input_size.
output	List[Float64]	—	Output vector of length output_size.

struct SpectralFlux

Spectral Flux analysis.

This implementation computes the squared difference between the magnitudes of the current frame and the previous frame, summed across all frequency bins.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

SpectralFlux Functions

struct SpectralFlux . fn init

Initialize the Spectral Flux analyzer.

fn init Signature

def __init__(out self, num_mags: Int, positive_only: Bool = False)

fn init Arguments

Name	Type	Default	Description
num_mags	Int	—	The number of magnitude bins in the input to expect. This is typically the FFT size divided by 2, but could also be the number of mel bands or another spectral summary that produces a list of values.
positive_only	Bool	False	Whether to only consider positive differences (increases in energy) when computing the spectral flux. If False, spectral flux is the average of squared differences between the magnitudes. If True, spectral flux is the average of (non-squared) differences between the magnitudes, but negative differences are set to 0. Using positive_only=True is a common approach when using spectral flux for onset detection, as onsets are typically characterized by increases in energy.

fn init Returns : Self

Static Method

This is a static method.

struct SpectralFlux . fn next_frame

Compute the spectral flux onset value for a given FFT analysis. This function is to be used by FFTProcess if SpectralFlux is passed as the "process".

Nothing is returned from this function, but the computed spectral flux value is stored in self.flux.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.
phases	List[Float64]	—	The input phases as a List of Float64.

struct SpectralFlux . fn get_features

Get the current spectral flux value.

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64] The current spectral flux feature value as a List[Float64] (with only the one element).

struct SpectralFlux . fn from_mags

Compute the spectral flux onset value for a given list of magnitudes. This function is useful when there is an FFT already computed, perhaps as part of a custom struct that implements the FFTProcessable trait.

Nothing is returned from this function, but the computed spectral flux value is stored in self.flux.

fn from_mags Signature

def from_mags(mut self, ref mags: List[Float64]) -> Float64

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The input magnitudes as a List of Float64.

fn from_mags Returns : Float64 The computed spectral flux value.

struct TopNFreqs

An FFTProcessable that identifies the top N frequency peaks in each FFT frame and provides their frequencies and amplitudes as output.

Args:

sample_rate: The sample rate of the audio signal.
window_size: The size of the FFT window. This determines the frequency resolution and the maximum number of frequency bins (window_size // 2 + 1).
num_peaks: The number of top peaks to identify in each FFT frame.
sort_by_freq: Whether to sort the output freq, amp pairs by frequency (True) or by amplitude (False).
thresh: The minimum amplitude threshold (in dB) for a peak to be considered. Peaks below this threshold will be ignored.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

TopNFreqs Functions

struct TopNFreqs . fn init

Initialize the TopNFreqs process.

fn init Signature

def __init__(out self, sample_rate: Float64, window_size: Int, num_peaks: Int = Int(5), sort_by_freq: Bool = True, thresh: Float64 = -30)

fn init Arguments

Name	Type	Default	Description
sample_rate	Float64	—	The sample rate of the audio signal.
window_size	Int	—	The size of the FFT window. This determines the frequency resolution and the maximum number of frequency bins (window_size // 2 + 1).
num_peaks	Int	Int(5)	The number of top peaks to identify in each FFT frame.
sort_by_freq	Bool	True	Whether to sort the output freq, amp pairs by frequency (True) or by amplitude (False).
thresh	Float64	-30	The minimum amplitude threshold (in dB) for a peak to be considered. Peaks below this threshold will be ignored.

fn init Returns : Self

Static Method

This is a static method.

struct TopNFreqs . fn get_features

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64]

struct TopNFreqs . fn get_features_ptr

Get a pointer to the current List of freq, amp pairs.

fn get_features_ptr Signature

def get_features_ptr(self) -> Pointer[List[Tuple[Float64, Float64]], origin_of(self.freq_amp_pairs)]

fn get_features_ptr Returns : Pointer[List[Tuple[Float64, Float64]], origin_of(self.freq_amp_pairs)] A pointer to the current frequency and amplitude pairs.

struct TopNFreqs . fn get_messages

fn get_messages Signature

def get_messages(mut self)

struct TopNFreqs . fn next_frame

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	—
phases	List[Float64]	—	—

struct TopNFreqs . fn sort_pairs_by_freq

fn sort_pairs_by_freq Signature

def sort_pairs_by_freq(mut self)

struct Chroma

A struct for computing chroma features.

Chroma features are a representation of the spectral content of an audio signal in terms of musical pitch classes. It can indicate how much of each pitch class is present in the audio signal.

This implementation imitates the Librosa library's chroma_stft.

Traits: AnyType, Copyable, Deinitable, FFTProcessable, GetFloat64Featurable, Movable

---

Chroma Functions

struct Chroma . fn init

Initialize the Chroma struct.

fn init Signature

def __init__(out self, sample_rate: Float64, window_size: Int, n_chroma: Int = Int(12), tuning: Float64 = 0, norm: Float64 = inf[DType.float64](), power: Float64 = 2, ctroct: Float64 = 5, octwidth: Float64 = 2, base_c: Bool = True)

fn init Arguments

Name	Type	Default	Description
sample_rate	Float64	—	The sample rate of the audio signal.
window_size	Int	—	The size of the FFT window.
n_chroma	Int	Int(12)	The number of chroma bins (divisions per octave).
tuning	Float64	0	The tuning deviation of the reference pitch as a fraction of a division per octave (when n_chroma = 12, these are semitones).
norm	Float64	inf[DType.float64]()	The normalization to apply to the chroma values. If inf the largest absolute chroma becomes 1 and everything is normalized to that. If 1, the sum of the chroma values is 1. If 0, no normalization is applied.
power	Float64	2	The power to which to raise the magnitude values before computing the chroma.
ctroct	Float64	5	Center of the octave weighting, measured in octaves above A0 = 27.5 Hz. FFT bins near this octave contribute most strongly to the chroma. For example, 4.0 centers the weighting near A4 (440 Hz) and 5.0 near A5 (880 Hz). Lower values emphasize lower-frequency octaves; higher values emphasize higher-frequency octaves.
octwidth	Float64	2	Width of the Gaussian octave weighting, in octaves. Smaller values give stronger octave focus; larger values approach flat weighting. 0 disables the octave weighting.
base_c	Bool	True	Whether to use C as the base pitch (True) or A as the base pitch (False).

fn init Returns : Self

Static Method

This is a static method.

struct Chroma . fn get_features

fn get_features Signature

def get_features(self) -> List[Float64]

fn get_features Returns : List[Float64]

struct Chroma . fn from_mags

Compute the chroma features from the magnitude values. Nothing is returned. The chroma values are updated internally and can be accessed via .chroma.

fn from_mags Signature

def from_mags(mut self, mags: List[Float64])

fn from_mags Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The magnitude values of the current FFT frame.

struct Chroma . fn next_frame

Process the next FFT frame. Nothing is returned. The chroma values are updated internally and can be accessed via .chroma.

fn next_frame Signature

def next_frame(mut self, mut mags: List[Float64], mut phases: List[Float64])

fn next_frame Arguments

Name	Type	Default	Description
mags	List[Float64]	—	The magnitude values of the current FFT frame.
phases	List[Float64]	—	The phase values of the current FFT frame.

struct Chroma . fn buf_analysis

fn buf_analysis Signature

def buf_analysis(buf: Buffer, chan: Int = Int(0), start_frame: Int = Int(0), var num_frames: Optional[Int] = None, n_chroma: Int = Int(12), tuning: Float64 = 0, norm: Float64 = inf[DType.float64](), power: Float64 = 2, ctroct: Float64 = 5, octwidth: Float64 = 2, base_c: Bool = True, fft_size: Int = Int(1024), hop_size: Int = Int(512), padding: Padding = Padding.half_window) -> List[List[Float64]]

fn buf_analysis Arguments

Name	Type	Default	Description
buf	Buffer	—	—
chan	Int	Int(0)	—
start_frame	Int	Int(0)	—
num_frames	Optional[Int]	None	—
n_chroma	Int	Int(12)	—
tuning	Float64	0	—
norm	Float64	inf[DType.float64]()	—
power	Float64	2	—
ctroct	Float64	5	—
octwidth	Float64	2	—
base_c	Bool	True	—
fft_size	Int	Int(1024)	—
hop_size	Int	Int(512)	—
padding	Padding	Padding.half_window	—

fn buf_analysis Returns : List[List[Float64]]

Raises

Static Method

This is a static method.

---

Documentation generated with mojo doc from Mojo version 1.0.0