2012-03-23 22:42:17 +01:00
|
|
|
/*
|
|
|
|
* Copyright (c) 2012 Justin Ruggles <justin.ruggles@gmail.com>
|
|
|
|
*
|
|
|
|
* This file is part of Libav.
|
|
|
|
*
|
|
|
|
* Libav is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU Lesser General Public
|
|
|
|
* License as published by the Free Software Foundation; either
|
|
|
|
* version 2.1 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* Libav is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
* Lesser General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU Lesser General Public
|
|
|
|
* License along with Libav; if not, write to the Free Software
|
|
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifndef AVRESAMPLE_AVRESAMPLE_H
|
|
|
|
#define AVRESAMPLE_AVRESAMPLE_H
|
|
|
|
|
|
|
|
/**
|
|
|
|
* @file
|
2012-10-28 22:52:54 +01:00
|
|
|
* @ingroup lavr
|
2012-03-23 22:42:17 +01:00
|
|
|
* external API header
|
|
|
|
*/
|
|
|
|
|
2012-10-28 22:52:54 +01:00
|
|
|
/**
|
|
|
|
* @defgroup lavr Libavresample
|
|
|
|
* @{
|
|
|
|
*
|
|
|
|
* Libavresample (lavr) is a library that handles audio resampling, sample
|
|
|
|
* format conversion and mixing.
|
|
|
|
*
|
|
|
|
* Interaction with lavr is done through AVAudioResampleContext, which is
|
|
|
|
* allocated with avresample_alloc_context(). It is opaque, so all parameters
|
|
|
|
* must be set with the @ref avoptions API.
|
|
|
|
*
|
|
|
|
* For example the following code will setup conversion from planar float sample
|
|
|
|
* format to interleaved signed 16-bit integer, downsampling from 48kHz to
|
|
|
|
* 44.1kHz and downmixing from 5.1 channels to stereo (using the default mixing
|
|
|
|
* matrix):
|
|
|
|
* @code
|
|
|
|
* AVAudioResampleContext *avr = avresample_alloc_context();
|
|
|
|
* av_opt_set_int(avr, "in_channel_layout", AV_CH_LAYOUT_5POINT1, 0);
|
|
|
|
* av_opt_set_int(avr, "out_channel_layout", AV_CH_LAYOUT_STEREO, 0);
|
|
|
|
* av_opt_set_int(avr, "in_sample_rate", 48000, 0);
|
|
|
|
* av_opt_set_int(avr, "out_sample_rate", 44100, 0);
|
|
|
|
* av_opt_set_int(avr, "in_sample_fmt", AV_SAMPLE_FMT_FLTP, 0);
|
2013-01-03 14:17:29 +01:00
|
|
|
* av_opt_set_int(avr, "out_sample_fmt", AV_SAMPLE_FMT_S16, 0);
|
2012-10-28 22:52:54 +01:00
|
|
|
* @endcode
|
|
|
|
*
|
|
|
|
* Once the context is initialized, it must be opened with avresample_open(). If
|
|
|
|
* you need to change the conversion parameters, you must close the context with
|
|
|
|
* avresample_close(), change the parameters as described above, then reopen it
|
|
|
|
* again.
|
|
|
|
*
|
|
|
|
* The conversion itself is done by repeatedly calling avresample_convert().
|
|
|
|
* Note that the samples may get buffered in two places in lavr. The first one
|
|
|
|
* is the output FIFO, where the samples end up if the output buffer is not
|
|
|
|
* large enough. The data stored in there may be retrieved at any time with
|
|
|
|
* avresample_read(). The second place is the resampling delay buffer,
|
|
|
|
* applicable only when resampling is done. The samples in it require more input
|
|
|
|
* before they can be processed. Their current amount is returned by
|
|
|
|
* avresample_get_delay(). At the end of conversion the resampling buffer can be
|
|
|
|
* flushed by calling avresample_convert() with NULL input.
|
|
|
|
*
|
|
|
|
* The following code demonstrates the conversion loop assuming the parameters
|
|
|
|
* from above and caller-defined functions get_input() and handle_output():
|
|
|
|
* @code
|
|
|
|
* uint8_t **input;
|
|
|
|
* int in_linesize, in_samples;
|
|
|
|
*
|
|
|
|
* while (get_input(&input, &in_linesize, &in_samples)) {
|
|
|
|
* uint8_t *output
|
|
|
|
* int out_linesize;
|
|
|
|
* int out_samples = avresample_available(avr) +
|
|
|
|
* av_rescale_rnd(avresample_get_delay(avr) +
|
|
|
|
* in_samples, 44100, 48000, AV_ROUND_UP);
|
|
|
|
* av_samples_alloc(&output, &out_linesize, 2, out_samples,
|
|
|
|
* AV_SAMPLE_FMT_S16, 0);
|
|
|
|
* out_samples = avresample_convert(avr, &output, out_linesize, out_samples,
|
|
|
|
* input, in_linesize, in_samples);
|
|
|
|
* handle_output(output, out_linesize, out_samples);
|
|
|
|
* av_freep(&output);
|
|
|
|
* }
|
|
|
|
* @endcode
|
|
|
|
*
|
|
|
|
* When the conversion is finished and the FIFOs are flushed if required, the
|
|
|
|
* conversion context and everything associated with it must be freed with
|
|
|
|
* avresample_free().
|
|
|
|
*/
|
|
|
|
|
2012-03-23 22:42:17 +01:00
|
|
|
#include "libavutil/avutil.h"
|
2012-11-10 16:00:00 +01:00
|
|
|
#include "libavutil/channel_layout.h"
|
2012-03-23 22:42:17 +01:00
|
|
|
#include "libavutil/dict.h"
|
|
|
|
#include "libavutil/log.h"
|
|
|
|
|
|
|
|
#include "libavresample/version.h"
|
|
|
|
|
|
|
|
#define AVRESAMPLE_MAX_CHANNELS 32
|
|
|
|
|
|
|
|
typedef struct AVAudioResampleContext AVAudioResampleContext;
|
|
|
|
|
|
|
|
/** Mixing Coefficient Types */
|
|
|
|
enum AVMixCoeffType {
|
2012-04-26 22:02:37 +02:00
|
|
|
AV_MIX_COEFF_TYPE_Q8, /** 16-bit 8.8 fixed-point */
|
2012-03-23 22:42:17 +01:00
|
|
|
AV_MIX_COEFF_TYPE_Q15, /** 32-bit 17.15 fixed-point */
|
|
|
|
AV_MIX_COEFF_TYPE_FLT, /** floating-point */
|
|
|
|
AV_MIX_COEFF_TYPE_NB, /** Number of coeff types. Not part of ABI */
|
|
|
|
};
|
|
|
|
|
2012-05-26 20:50:02 +02:00
|
|
|
/** Resampling Filter Types */
|
|
|
|
enum AVResampleFilterType {
|
|
|
|
AV_RESAMPLE_FILTER_TYPE_CUBIC, /**< Cubic */
|
|
|
|
AV_RESAMPLE_FILTER_TYPE_BLACKMAN_NUTTALL, /**< Blackman Nuttall Windowed Sinc */
|
|
|
|
AV_RESAMPLE_FILTER_TYPE_KAISER, /**< Kaiser Windowed Sinc */
|
|
|
|
};
|
|
|
|
|
2012-10-31 20:40:12 +01:00
|
|
|
enum AVResampleDitherMethod {
|
|
|
|
AV_RESAMPLE_DITHER_NONE, /**< Do not use dithering */
|
|
|
|
AV_RESAMPLE_DITHER_RECTANGULAR, /**< Rectangular Dither */
|
|
|
|
AV_RESAMPLE_DITHER_TRIANGULAR, /**< Triangular Dither*/
|
|
|
|
AV_RESAMPLE_DITHER_TRIANGULAR_HP, /**< Triangular Dither with High Pass */
|
|
|
|
AV_RESAMPLE_DITHER_TRIANGULAR_NS, /**< Triangular Dither with Noise Shaping */
|
|
|
|
AV_RESAMPLE_DITHER_NB, /**< Number of dither types. Not part of ABI. */
|
|
|
|
};
|
|
|
|
|
2012-03-23 22:42:17 +01:00
|
|
|
/**
|
|
|
|
* Return the LIBAVRESAMPLE_VERSION_INT constant.
|
|
|
|
*/
|
|
|
|
unsigned avresample_version(void);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Return the libavresample build-time configuration.
|
|
|
|
* @return configure string
|
|
|
|
*/
|
|
|
|
const char *avresample_configuration(void);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Return the libavresample license.
|
|
|
|
*/
|
|
|
|
const char *avresample_license(void);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Get the AVClass for AVAudioResampleContext.
|
|
|
|
*
|
|
|
|
* Can be used in combination with AV_OPT_SEARCH_FAKE_OBJ for examining options
|
|
|
|
* without allocating a context.
|
|
|
|
*
|
|
|
|
* @see av_opt_find().
|
|
|
|
*
|
|
|
|
* @return AVClass for AVAudioResampleContext
|
|
|
|
*/
|
|
|
|
const AVClass *avresample_get_class(void);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Allocate AVAudioResampleContext and set options.
|
|
|
|
*
|
|
|
|
* @return allocated audio resample context, or NULL on failure
|
|
|
|
*/
|
|
|
|
AVAudioResampleContext *avresample_alloc_context(void);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Initialize AVAudioResampleContext.
|
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
|
|
|
* @return 0 on success, negative AVERROR code on failure
|
|
|
|
*/
|
|
|
|
int avresample_open(AVAudioResampleContext *avr);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Close AVAudioResampleContext.
|
|
|
|
*
|
|
|
|
* This closes the context, but it does not change the parameters. The context
|
|
|
|
* can be reopened with avresample_open(). It does, however, clear the output
|
|
|
|
* FIFO and any remaining leftover samples in the resampling delay buffer. If
|
|
|
|
* there was a custom matrix being used, that is also cleared.
|
|
|
|
*
|
|
|
|
* @see avresample_convert()
|
|
|
|
* @see avresample_set_matrix()
|
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
|
|
|
*/
|
|
|
|
void avresample_close(AVAudioResampleContext *avr);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Free AVAudioResampleContext and associated AVOption values.
|
|
|
|
*
|
|
|
|
* This also calls avresample_close() before freeing.
|
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
|
|
|
*/
|
|
|
|
void avresample_free(AVAudioResampleContext **avr);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Generate a channel mixing matrix.
|
|
|
|
*
|
|
|
|
* This function is the one used internally by libavresample for building the
|
|
|
|
* default mixing matrix. It is made public just as a utility function for
|
|
|
|
* building custom matrices.
|
|
|
|
*
|
|
|
|
* @param in_layout input channel layout
|
|
|
|
* @param out_layout output channel layout
|
|
|
|
* @param center_mix_level mix level for the center channel
|
|
|
|
* @param surround_mix_level mix level for the surround channel(s)
|
|
|
|
* @param lfe_mix_level mix level for the low-frequency effects channel
|
|
|
|
* @param normalize if 1, coefficients will be normalized to prevent
|
|
|
|
* overflow. if 0, coefficients will not be
|
|
|
|
* normalized.
|
|
|
|
* @param[out] matrix mixing coefficients; matrix[i + stride * o] is
|
|
|
|
* the weight of input channel i in output channel o.
|
|
|
|
* @param stride distance between adjacent input channels in the
|
|
|
|
* matrix array
|
2012-06-09 22:45:49 +02:00
|
|
|
* @param matrix_encoding matrixed stereo downmix mode (e.g. dplii)
|
2012-03-23 22:42:17 +01:00
|
|
|
* @return 0 on success, negative AVERROR code on failure
|
|
|
|
*/
|
|
|
|
int avresample_build_matrix(uint64_t in_layout, uint64_t out_layout,
|
|
|
|
double center_mix_level, double surround_mix_level,
|
|
|
|
double lfe_mix_level, int normalize, double *matrix,
|
2012-06-09 22:45:49 +02:00
|
|
|
int stride, enum AVMatrixEncoding matrix_encoding);
|
2012-03-23 22:42:17 +01:00
|
|
|
|
|
|
|
/**
|
|
|
|
* Get the current channel mixing matrix.
|
|
|
|
*
|
2012-11-29 20:59:19 +01:00
|
|
|
* If no custom matrix has been previously set or the AVAudioResampleContext is
|
|
|
|
* not open, an error is returned.
|
|
|
|
*
|
2012-03-23 22:42:17 +01:00
|
|
|
* @param avr audio resample context
|
|
|
|
* @param matrix mixing coefficients; matrix[i + stride * o] is the weight of
|
|
|
|
* input channel i in output channel o.
|
|
|
|
* @param stride distance between adjacent input channels in the matrix array
|
|
|
|
* @return 0 on success, negative AVERROR code on failure
|
|
|
|
*/
|
|
|
|
int avresample_get_matrix(AVAudioResampleContext *avr, double *matrix,
|
|
|
|
int stride);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Set channel mixing matrix.
|
|
|
|
*
|
|
|
|
* Allows for setting a custom mixing matrix, overriding the default matrix
|
|
|
|
* generated internally during avresample_open(). This function can be called
|
|
|
|
* anytime on an allocated context, either before or after calling
|
2012-11-29 20:59:19 +01:00
|
|
|
* avresample_open(), as long as the channel layouts have been set.
|
|
|
|
* avresample_convert() always uses the current matrix.
|
2012-03-23 22:42:17 +01:00
|
|
|
* Calling avresample_close() on the context will clear the current matrix.
|
|
|
|
*
|
|
|
|
* @see avresample_close()
|
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
|
|
|
* @param matrix mixing coefficients; matrix[i + stride * o] is the weight of
|
|
|
|
* input channel i in output channel o.
|
|
|
|
* @param stride distance between adjacent input channels in the matrix array
|
|
|
|
* @return 0 on success, negative AVERROR code on failure
|
|
|
|
*/
|
|
|
|
int avresample_set_matrix(AVAudioResampleContext *avr, const double *matrix,
|
|
|
|
int stride);
|
|
|
|
|
2012-12-19 20:58:57 +01:00
|
|
|
/**
|
|
|
|
* Set a customized input channel mapping.
|
|
|
|
*
|
|
|
|
* This function can only be called when the allocated context is not open.
|
|
|
|
* Also, the input channel layout must have already been set.
|
|
|
|
*
|
|
|
|
* Calling avresample_close() on the context will clear the channel mapping.
|
|
|
|
*
|
|
|
|
* The map for each input channel specifies the channel index in the source to
|
|
|
|
* use for that particular channel, or -1 to mute the channel. Source channels
|
|
|
|
* can be duplicated by using the same index for multiple input channels.
|
|
|
|
*
|
|
|
|
* Examples:
|
|
|
|
*
|
|
|
|
* Reordering 5.1 AAC order (C,L,R,Ls,Rs,LFE) to Libav order (L,R,C,LFE,Ls,Rs):
|
|
|
|
* { 1, 2, 0, 5, 3, 4 }
|
|
|
|
*
|
|
|
|
* Muting the 3rd channel in 4-channel input:
|
|
|
|
* { 0, 1, -1, 3 }
|
|
|
|
*
|
|
|
|
* Duplicating the left channel of stereo input:
|
|
|
|
* { 0, 0 }
|
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
|
|
|
* @param channel_map customized input channel mapping
|
|
|
|
* @return 0 on success, negative AVERROR code on failure
|
|
|
|
*/
|
|
|
|
int avresample_set_channel_mapping(AVAudioResampleContext *avr,
|
|
|
|
const int *channel_map);
|
|
|
|
|
2012-03-23 22:42:17 +01:00
|
|
|
/**
|
|
|
|
* Set compensation for resampling.
|
|
|
|
*
|
2012-11-30 02:58:05 +01:00
|
|
|
* This can be called anytime after avresample_open(). If resampling is not
|
|
|
|
* automatically enabled because of a sample rate conversion, the
|
|
|
|
* "force_resampling" option must have been set to 1 when opening the context
|
|
|
|
* in order to use resampling compensation.
|
2012-03-23 22:42:17 +01:00
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
|
|
|
* @param sample_delta compensation delta, in samples
|
|
|
|
* @param compensation_distance compensation distance, in samples
|
|
|
|
* @return 0 on success, negative AVERROR code on failure
|
|
|
|
*/
|
|
|
|
int avresample_set_compensation(AVAudioResampleContext *avr, int sample_delta,
|
|
|
|
int compensation_distance);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Convert input samples and write them to the output FIFO.
|
|
|
|
*
|
2012-10-28 22:52:55 +01:00
|
|
|
* The upper bound on the number of output samples is given by
|
|
|
|
* avresample_available() + (avresample_get_delay() + number of input samples) *
|
|
|
|
* output sample rate / input sample rate.
|
|
|
|
*
|
2012-03-23 22:42:17 +01:00
|
|
|
* The output data can be NULL or have fewer allocated samples than required.
|
|
|
|
* In this case, any remaining samples not written to the output will be added
|
|
|
|
* to an internal FIFO buffer, to be returned at the next call to this function
|
|
|
|
* or to avresample_read().
|
|
|
|
*
|
|
|
|
* If converting sample rate, there may be data remaining in the internal
|
|
|
|
* resampling delay buffer. avresample_get_delay() tells the number of remaining
|
|
|
|
* samples. To get this data as output, call avresample_convert() with NULL
|
|
|
|
* input.
|
|
|
|
*
|
|
|
|
* At the end of the conversion process, there may be data remaining in the
|
|
|
|
* internal FIFO buffer. avresample_available() tells the number of remaining
|
|
|
|
* samples. To get this data as output, either call avresample_convert() with
|
|
|
|
* NULL input or call avresample_read().
|
|
|
|
*
|
|
|
|
* @see avresample_available()
|
|
|
|
* @see avresample_read()
|
|
|
|
* @see avresample_get_delay()
|
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
|
|
|
* @param output output data pointers
|
|
|
|
* @param out_plane_size output plane size, in bytes.
|
|
|
|
* This can be 0 if unknown, but that will lead to
|
|
|
|
* optimized functions not being used directly on the
|
|
|
|
* output, which could slow down some conversions.
|
|
|
|
* @param out_samples maximum number of samples that the output buffer can hold
|
|
|
|
* @param input input data pointers
|
|
|
|
* @param in_plane_size input plane size, in bytes
|
|
|
|
* This can be 0 if unknown, but that will lead to
|
|
|
|
* optimized functions not being used directly on the
|
|
|
|
* input, which could slow down some conversions.
|
|
|
|
* @param in_samples number of input samples to convert
|
|
|
|
* @return number of samples written to the output buffer,
|
|
|
|
* not including converted samples added to the internal
|
|
|
|
* output FIFO
|
|
|
|
*/
|
2012-10-05 06:56:00 +02:00
|
|
|
int avresample_convert(AVAudioResampleContext *avr, uint8_t **output,
|
|
|
|
int out_plane_size, int out_samples, uint8_t **input,
|
2012-03-23 22:42:17 +01:00
|
|
|
int in_plane_size, int in_samples);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Return the number of samples currently in the resampling delay buffer.
|
|
|
|
*
|
|
|
|
* When resampling, there may be a delay between the input and output. Any
|
|
|
|
* unconverted samples in each call are stored internally in a delay buffer.
|
|
|
|
* This function allows the user to determine the current number of samples in
|
|
|
|
* the delay buffer, which can be useful for synchronization.
|
|
|
|
*
|
|
|
|
* @see avresample_convert()
|
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
|
|
|
* @return number of samples currently in the resampling delay buffer
|
|
|
|
*/
|
|
|
|
int avresample_get_delay(AVAudioResampleContext *avr);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Return the number of available samples in the output FIFO.
|
|
|
|
*
|
|
|
|
* During conversion, if the user does not specify an output buffer or
|
|
|
|
* specifies an output buffer that is smaller than what is needed, remaining
|
|
|
|
* samples that are not written to the output are stored to an internal FIFO
|
|
|
|
* buffer. The samples in the FIFO can be read with avresample_read() or
|
|
|
|
* avresample_convert().
|
|
|
|
*
|
|
|
|
* @see avresample_read()
|
|
|
|
* @see avresample_convert()
|
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
|
|
|
* @return number of samples available for reading
|
|
|
|
*/
|
|
|
|
int avresample_available(AVAudioResampleContext *avr);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Read samples from the output FIFO.
|
|
|
|
*
|
|
|
|
* During conversion, if the user does not specify an output buffer or
|
|
|
|
* specifies an output buffer that is smaller than what is needed, remaining
|
|
|
|
* samples that are not written to the output are stored to an internal FIFO
|
|
|
|
* buffer. This function can be used to read samples from that internal FIFO.
|
|
|
|
*
|
|
|
|
* @see avresample_available()
|
|
|
|
* @see avresample_convert()
|
|
|
|
*
|
|
|
|
* @param avr audio resample context
|
2012-05-09 06:35:13 +02:00
|
|
|
* @param output output data pointers. May be NULL, in which case
|
|
|
|
* nb_samples of data is discarded from output FIFO.
|
2012-03-23 22:42:17 +01:00
|
|
|
* @param nb_samples number of samples to read from the FIFO
|
|
|
|
* @return the number of samples written to output
|
|
|
|
*/
|
2012-10-05 06:56:00 +02:00
|
|
|
int avresample_read(AVAudioResampleContext *avr, uint8_t **output, int nb_samples);
|
2012-03-23 22:42:17 +01:00
|
|
|
|
2012-10-28 22:52:54 +01:00
|
|
|
/**
|
|
|
|
* @}
|
|
|
|
*/
|
|
|
|
|
2012-03-23 22:42:17 +01:00
|
|
|
#endif /* AVRESAMPLE_AVRESAMPLE_H */
|