Direct processing

Functions
void	bigvoice_ProcessAudio (struct bigvoice_CInstance *instance, const float *input, float *output)
void	bigvoice_ProcessAudio_Planar (struct bigvoice_CInstance *instance, float const *const *input, float *const *output)
float *	bigvoice_GetBufferIn (struct bigvoice_CInstance *instance)
float *	bigvoice_GetBufferOut (struct bigvoice_CInstance *instance)
unsigned int	bigvoice_GetEstimatedDelay (struct bigvoice_CInstance *instance)

Detailed Description

Optimized processing with direct in->out.

Function Documentation

bigvoice_GetBufferIn()

float * bigvoice_GetBufferIn

(

struct bigvoice_CInstance *

instance

)

Get preallocated aligned buffer for audio input for bigvoice_ProcessAudio()

The buffer is a float[256 * bigvoice_GetAudioInputCount() * bigvoice_GetChannelCount()]

So for this process/version : float[256*1*1]

Note

The calling process can allocate the buffer itself, but it must be aligned.

Parameters

instance

the processing instance

Returns

An audio input dedicated buffer.

bigvoice_GetBufferOut()

float * bigvoice_GetBufferOut

(

struct bigvoice_CInstance *

instance

)

Get preallocated aligned buffer for audio output for bigvoice_ProcessAudio()

The buffer is a float[256 * bigvoice_GetAudioOutputCount() * bigvoice_GetChannelCount()]

So for this process/version : float[256*1*1]

Note

The calling process can allocate the buffer itself, but it must be aligned.

Parameters

instance

the processing instance

Returns

An audio output dedicated buffer.

bigvoice_GetEstimatedDelay()

unsigned int bigvoice_GetEstimatedDelay

(

struct bigvoice_CInstance *

instance

)

Get the delay (estimated) of current processing algorithm

Remarks

This is not very accurate, it does not take into account group delays.

Returns

Estimated delay in frames

bigvoice_ProcessAudio()

void bigvoice_ProcessAudio	(	struct bigvoice_CInstance *	instance,
		const float *	input,
		float *	output
	)

Process an audio chunk

A chunk contains bigvoice_GetProcessChunkFrames frames, set at init by frames_per_chunk in bigvoice_InitProcess3

Input has bigvoice_GetAudioInputCount() * bigvoice_GetChannelCount() channels

Ouput has bigvoice_GetAudioOutputCount() * bigvoice_GetChannelCount() channels

Each sample is a native 32-bit floating-point number where 1.0 is full scale. Channels are interlaced, so the buffers are [L0,R0,L1,R1,...,Ln,Rn] (L=Left, R=Right)

In case of multiple audio input/output, buffer is [aL0,aR0,bL0,bR0,...,aLn,aRn,bLn,bRn]

Warning

Buffers must be 64-byte aligned

Note

With 64 frames per chunk and 2 channels and 1 audio input, input buffer size is 64*1*2*4=512 bytes

Parameters

instance	the processing instance
input	audio input buffer (64 byte aligned) or NULL to use internal buffer bigvoice_GetBufferIn
output	audio output buffer (64 byte aligned) or NULL to use internal buffer bigvoice_GetBufferOut

bigvoice_ProcessAudio_Planar()

void bigvoice_ProcessAudio_Planar	(	struct bigvoice_CInstance *	instance,
		float const *const *	input,
		float *const *	output
	)

Process an audio chunk of planar audio

A chunk contains bigvoice_GetProcessChunkFrames frames, set at init by frames_per_chunk in bigvoice_InitProcess3

Input has bigvoice_GetAudioInputCount() * bigvoice_GetChannelCount() channels

Ouput has bigvoice_GetAudioOutputCount() * bigvoice_GetChannelCount() channels

Each sample is a native 32-bit floating-point number where 1.0 is full scale. Channels are NOT interlaced, so the buffers are arrays of planar audio : [ [L0,L1,...,Ln] , [R0,R1,...,Rn] ] (L=Left, R=Right)

In case of multiple audio input/output, buffer are [ [aL...], [aR...], [bL...], [bR...], ... ]

Parameters

instance	the processing instance
input	array of audio input buffers for each input channel (no alignment required)
output	array of audio output buffers for each input channel (no alignment required)