aus.operations¶

adjust_level(audio: np.ndarray, max_level: float)¶

Adjusts the level of audio to a specified dB level

Parameters:

audio (np.ndarray) – The audio samples as a NumPy array
max_level (float) – The max level for the audio

Returns:

The scaled audio

Return type:

np.ndarray

beat_envelope(tempo, sample_rate, beat_sequence, envelope_type='hanning', envelope_duration=1000)¶

Generates a beat envelope. You specify the tempo, and provide a list of beat durations in order. This function will generate a volume envelope that can be applied to a sound file to make it pulse with these beat durations.

Parameters:

tempo (float) – The tempo
sample_rate (int) – The audio sample rate
beat_sequence (list) – A sequence of Fractions representing beat durations
envelope_type (str) – The envelope used for fade-in and fade-out on each beat
envelope_duration (int) – The envelope duration. The envelope will be split in half and applied to the beginning and end of the beat. If the beat is too short, a shorter envelope will be generated for that beat.

Returns:

An array with the beat envelope

Return type:

np.ndarray

beat_envelope_multichannel(tempo: float, sample_rate: int, beat_sequence: list, num_channels: int, channel_levels: list, envelope_type='hanning', envelope_duration: int = 1000)¶

Generates a multichannel beat envelope. You specify the tempo, and provide a list of N beat durations in order. You also specify the number of output channels M and a list of level tuples N*x*M - one level coefficient for each channel, for each beat. This function will generate a volume envelope that can be applied to a sound file to make it pulse with these beat durations. You will need to provide a multichannel sound file with the correct number of channels when you apply the envelope.

Parameters:

tempo (float) – The tempo
sample_rate (int) – The audio sample rate
beat_sequence (list) – A sequence of Fractions representing beat durations
num_channels (int) – The number of channels in the envelope
channel_levels (list) – A list of channel level tuples (or lists). Each sublist corresponds to the level coefficients for the current beat.
envelope_type (str) – The envelope used for fade-in and fade-out on each beat
envelope_duration (int) – The envelope duration. The envelope will be split in half and applied to the beginning and end of the beat. If the beat is too short, a shorter envelope will be generated for that beat.

Returns:

An array with the beat envelope

Return type:

np.ndarray

calculate_dc_bias(audio: np.ndarray)¶

Calculates DC bias of an audio signal

Parameters:: audio (np.ndarray) – The audio signal
Returns:: The DC bias
Return type:: float

cpsmidi(freq: float)¶

Calculates the MIDI note of a provided frequency

Parameters:: freq (float) – The frequency in Hz
Returns:: The MIDI note
Return type:: float

dbfs(audio)¶

Calculates dbfs (decibels full scale) for an audio sequence or sample. This function assumes that the audio is in float format where 1 is the highest possible peak.

Parameters:: audio (np.ndarray) – The audio array or sample to calculate dbfs for
Returns:: A float value representing the dbfs
Return type:: float

dbfs_audio(audio: np.ndarray)¶

Calculates dbfs (decibels full scale) for a chunk of audio. This function will use the RMS method, and assumes that the audio is in float format where 1 is the highest possible peak.

Parameters:: audio (np.ndarray) – The audio to calculate dbfs for
Returns:: A float value representing the dbfs
Return type:: float

dbfs_max_local(audio: np.ndarray, chunk_size: int = 10, hop_size: int = 5)¶

Checks the maximum local dbfs (decibels full scale) of an audio file

Parameters:

audio (np.ndarray) – The audio
chunk_size (int) – The chunk size to check
hop_size (int) – The number of frames to hop from chunk center to chunk center

Returns:

The max local dbfs

Return type:

float

dbfs_min_local(audio: np.ndarray, chunk_size: int = 10, hop_size: int = 5)¶

Checks the minimum local dbfs (decibels full scale) of an audio file

Parameters:

audio (np.ndarray) – The audio
chunk_size (int) – The chunk size to check
hop_size (int) – The number of frames to hop from chunk center to chunk center

Returns:

The min local dbfs

Return type:

float

exchanger(data: np.ndarray, hop: int)¶

Exchanges samples in an audio file or STFT frames in a spectrum. Each sample (or STFT frame) is swapped with the sample (or STFT frame) hop steps ahead or hop steps behind. If audio is being processed, it should be in the shape (channels, samples). If STFT data is being processed, it should be in the shape (channels, frames, bins).

Parameters:

data (np.ndarray) – The audio (or spectrum) to process
hop (int) – The hop size

Returns:

The exchanged audio (or spectrum)

Return type:

np.ndarray

fade_in(audio: np.ndarray, envelope='hanning', duration: int = 100)¶

Implements a fade-in on an array of audio samples.

Parameters:

audio (np.ndarray) – The array of audio samples (may have multiple channels; the fade-in will be applied to all channels)
envelope (str) – The shape of the fade-in envelope. Must be a NumPy envelope. The envelope will be divided in half, and only the first half will be used.
duration (int) – The duration (in frames) of the fade-in envelope half. If the duration is longer than the audio, it will be truncated.

Returns:

The audio with a fade in applied.

Return type:

np.ndarray

fade_out(audio: np.ndarray, envelope='hanning', duration: int = 100)¶

Implements a fade-out on an array of audio samples.

Parameters:

audio (np.ndarray) – The array of audio samples (may have multiple channels; the fade-out will be applied to all channels)
envelope (str) – The shape of the fade-out envelope. Must be a NumPy envelope. The envelope will be divided in half, and only the second half will be used.
duration (int) – The duration (in frames) of the fade-out envelope half. If the duration is longer than the audio, it will be truncated.

Returns:

The audio with a fade-out applied.

Return type:

np.ndarray

force_equal_energy(audio: np.ndarray, dbfs: float = -6.0, window_size: int = 8192)¶

An algorithm that forces equal energy on a mono signal over time. For example, if a signal initially has high energy, and gets less energetic, this will adjust the energy level so that it does not decrease. Better results come with using a larger window size, so the energy changes more gradually.

Parameters:

audio (np.ndarray) – The array of audio samples
dbfs (float) – The target level of the entire signal, in dbfs
window_size (int) – The window size to consider when detecting RMS energy

Returns:

An adjusted version of the signal

Return type:

np.ndarray

leak_dc_bias_averager(audio: np.ndarray)¶

Leaks DC bias of an audio signal

Parameters:: audio (np.ndarray) – The audio signal
Returns:: The bias-free signal
Return type:: np.ndarray

leak_dc_bias_filter(audio: np.ndarray)¶

Leaks DC bias of an audio signal using a highpass filter, described on pp. 762-763 of “Understanding Digital Signal Processing,” 3rd edition, by Richard G. Lyons

Parameters:: audio (np.ndarray) – The audio signal
Returns:: The bias-free signal
Return type:: np.ndarray

midicps(midi_note: float)¶

Calculates the frequency of a specified MIDI note

Parameters:: midi_note (float) – The MIDI note
Returns:: The frequency in Hz
Return type:: float

midiratio(interval: float)¶

Calculates the MIDI ratio of a specified MIDI interval

Parameters:: interval (freq) – The MIDI interval in half steps
Returns:: The ratio
Return type:: float

mixdown(audio: np.ndarray)¶

Mixes a multichannel signal to a mono signal.

Parameters:: audio (np.ndarray) – The audio to mix if it isn’t mono
Returns:: The mixed audio
Return type:: np.ndarray

pan_level_adjuster(pan_levels: np.ndarray)¶

Adjusts pan levels in a list for a power sum of 1. The values in the list should be fractional volume levels that sum to 1. After applying this operations, the values in the list will be adjusted so that their squares now sum to 1. The levels are adjusted in place.

The idea is that you might want to arbitrarily divide the total volume of a sound over several channels. However, you want the sum of the signal power to equal to 1. So you need to adjust these fractional levels so that the power sum is correct. This function computes a scalar that is applied to all of the pan levels to make the summed power level equal to 1.

Parameters:: pan_levels (np.ndarray) – A list of pan levels (one level for each channel)

pan_mapper(pan_coefficients: np.ndarray, mapper: np.ndarray)¶

Maps pan positions to actual speaker positions. You pass a mapping array that lists the speaker numbers in panning order.

This is useful if you want to use a different numbering system for your pan positions than the numbering system used for the actual output channels. For example, you might want to pan in a circle for a quad-channel setup, but the hardware is set up for stereo pairs.

Example: Suppose you have a quad setup. Your mapper would be [0, 1, 3, 2] if you are thinking clockwise, or [1, 0, 2, 3] if you are thinking counterclockwise. If you have an 8-channel setup, your mapper would be [0, 1, 3, 5, 7, 6, 4, 2] for clockwise and [1, 0, 2, 4, 6, 7, 5, 3] for counterclockwise.

Parameters:

pan_coefficients (np.ndarray) – A list of pan coefficient lists
mapper (np.ndarray) – The mapper for reordering the pan coefficients

Returns:

A new, mapped pan coefficient list

Return type:

np.ndarray

panner(num_channels: int, start_pos: int, end_pos: int, num_iterations: int, pan_law: str = 'constant_power')¶

Multichannel panner, moving from start_pos to end_pos over num_iterations. It generates a list of pan coefficients (each coefficient is the volume coefficient for the corresponding channel). (https://www.cs.cmu.edu/~music/icm-online/readings/panlaws/panlaws.pdf)

Parameters:

num_channels (int) – The number of channels
start_pos (float) – The start panning position
end_pos (float) – The end panning position
num_iterations (int) – The number of steps to take to move from start_pos to end_pos
pan_law (str) – The pan law (“linear”, “constant_power”, “neg_4_5_db”)

Returns:

An array of pan coefficients

Return type:

np.ndarray

stochastic_exchanger(data: np.ndarray, max_hop: int)¶

Stochastically exchanges samples in an audio file or STFT frames in a spectrum. Each sample (or STFT frame) is swapped with the sample (or STFT frame) up to max_hop steps ahead or max_hop steps behind. If audio is being processed, it should be in the shape (channels, samples). If STFT data is being processed, it should be in the shape (channels, frames, bins).

Warning

If you try to run this on sampled audio rather than STFT data, this will take a very long time!

Parameters:

data (np.ndarray) – The audio (or spectrum) to process
hop (int) – The hop size

Returns:

The exchanged audio (or spectrum)

Return type:

np.ndarray

aus.operations¶

aus

Navigation

Related Topics