third_party_sonic/sonic.h

/* Sonic library
   Copyright 2010
   Bill Cox
   This file is part of the Sonic Library.

   This file is licensed under the Apache 2.0 license.
*/

/*
The Sonic Library implements a new algorithm invented by Bill Cox for the
specific purpose of speeding up speech by high factors at high quality.  It
generates smooth speech at speed up factors as high as 6X, possibly more.  It is
also capable of slowing down speech, and generates high quality results
regardless of the speed up or slow down factor.  For speeding up speech by 2X or
more, the following equation is used:

    newSamples = period/(speed - 1.0)
    scale = 1.0/newSamples;

where period is the current pitch period, determined using AMDF or any other
pitch estimator, and speed is the speedup factor.  If the current position in
the input stream is pointed to by "samples", and the current output stream
position is pointed to by "out", then newSamples number of samples can be
generated with:

    out[t] = (samples[t]*(newSamples - t) + samples[t + period]*t)/newSamples;

where t = 0 to newSamples - 1.

For speed factors < 2X, the PICOLA algorithm is used.  The above
algorithm is first used to double the speed of one pitch period.  Then, enough
input is directly copied from the input to the output to achieve the desired
speed up facter, where 1.0 < speed < 2.0.  The amount of data copied is derived:

    speed = (2*period + length)/(period + length)
    speed*length + speed*period = 2*period + length
    length(speed - 1) = 2*period - speed*period
    length = period*(2 - speed)/(speed - 1)

For slowing down speech where 0.5 < speed < 1.0, a pitch period is inserted into
the output twice, and length of input is copied from the input to the output
until the output desired speed is reached.  The length of data copied is:

    length = period*(speed - 0.5)/(1 - speed)

For slow down factors below 0.5, no data is copied, and an algorithm
similar to high speed factors is used.
*/

/* Uncomment this to use sin-wav based overlap add which in theory can improve
   sound quality slightly, at the expense of lots of floating point math. */
/* #define SONIC_USE_SIN */

#ifdef __cplusplus
extern "C" {
#endif

/* This specifies the range of voice pitches we try to match.
   Note that if we go lower than 65, we could overflow in findPitchInRange */
#define SONIC_MIN_PITCH 65
#define SONIC_MAX_PITCH 400

/* These are used to down-sample some inputs to improve speed */
#define SONIC_AMDF_FREQ 4000

struct sonicStreamStruct;
typedef struct sonicStreamStruct* sonicStream;

/* For all of the following functions, numChannels is multiplied by numSamples
   to determine the actual number of values read or returned. */

/* Create a sonic stream.  Return NULL only if we are out of memory and cannot
  allocate the stream. Set numChannels to 1 for mono, and 2 for stereo. */
sonicStream sonicCreateStream(int sampleRate, int numChannels);
/* Destroy the sonic stream. */
void sonicDestroyStream(sonicStream stream);
/* Use this to write floating point data to be speed up or down into the stream.
   Values must be between -1 and 1.  Return 0 if memory realloc failed,
   otherwise 1 */
int sonicWriteFloatToStream(sonicStream stream, float* samples, int numSamples);
/* Use this to write 16-bit data to be speed up or down into the stream.
   Return 0 if memory realloc failed, otherwise 1 */
int sonicWriteShortToStream(sonicStream stream, short* samples, int numSamples);
/* Use this to write 8-bit unsigned data to be speed up or down into the stream.
   Return 0 if memory realloc failed, otherwise 1 */
int sonicWriteUnsignedCharToStream(sonicStream stream, unsigned char* samples,
                                   int numSamples);
/* Use this to read floating point data out of the stream.  Sometimes no data
   will be available, and zero is returned, which is not an error condition. */
int sonicReadFloatFromStream(sonicStream stream, float* samples,
                             int maxSamples);
/* Use this to read 16-bit data out of the stream.  Sometimes no data will
   be available, and zero is returned, which is not an error condition. */
int sonicReadShortFromStream(sonicStream stream, short* samples,
                             int maxSamples);
/* Use this to read 8-bit unsigned data out of the stream.  Sometimes no data
   will be available, and zero is returned, which is not an error condition. */
int sonicReadUnsignedCharFromStream(sonicStream stream, unsigned char* samples,
                                    int maxSamples);
/* Force the sonic stream to generate output using whatever data it currently
   has.  No extra delay will be added to the output, but flushing in the middle
   of words could introduce distortion. */
int sonicFlushStream(sonicStream stream);
/* Return the number of samples in the output buffer */
int sonicSamplesAvailable(sonicStream stream);
/* Get the speed of the stream. */
float sonicGetSpeed(sonicStream stream);
/* Set the speed of the stream. */
void sonicSetSpeed(sonicStream stream, float speed);
/* Get the pitch of the stream. */
float sonicGetPitch(sonicStream stream);
/* Set the pitch of the stream. */
void sonicSetPitch(sonicStream stream, float pitch);
/* Get the rate of the stream. */
float sonicGetRate(sonicStream stream);
/* Set the rate of the stream. */
void sonicSetRate(sonicStream stream, float rate);
/* Get the scaling factor of the stream. */
float sonicGetVolume(sonicStream stream);
/* Set the scaling factor of the stream. */
void sonicSetVolume(sonicStream stream, float volume);
/* Get the chord pitch setting. */
int sonicGetChordPitch(sonicStream stream);
/* Set chord pitch mode on or off.  Default is off.  See the documentation
   page for a description of this feature. */
void sonicSetChordPitch(sonicStream stream, int useChordPitch);
/* Get the quality setting. */
int sonicGetQuality(sonicStream stream);
/* Set the "quality".  Default 0 is virtually as good as 1, but very much
 * faster. */
void sonicSetQuality(sonicStream stream, int quality);
/* Get the sample rate of the stream. */
int sonicGetSampleRate(sonicStream stream);
/* Set the sample rate of the stream.  This will drop any samples that have not
 * been read. */
void sonicSetSampleRate(sonicStream stream, int sampleRate);
/* Get the number of channels. */
int sonicGetNumChannels(sonicStream stream);
/* Set the number of channels.  This will drop any samples that have not been
 * read. */
void sonicSetNumChannels(sonicStream stream, int numChannels);
/* This is a non-stream oriented interface to just change the speed of a sound
   sample.  It works in-place on the sample array, so there must be at least
   speed*numSamples available space in the array. Returns the new number of
   samples. */
int sonicChangeFloatSpeed(float* samples, int numSamples, float speed,
                          float pitch, float rate, float volume,
                          int useChordPitch, int sampleRate, int numChannels);
/* This is a non-stream oriented interface to just change the speed of a sound
   sample.  It works in-place on the sample array, so there must be at least
   speed*numSamples available space in the array. Returns the new number of
   samples. */
int sonicChangeShortSpeed(short* samples, int numSamples, float speed,
                          float pitch, float rate, float volume,
                          int useChordPitch, int sampleRate, int numChannels);

#ifdef SONIC_SPECTROGRAM
/*
This code generates high quality spectrograms from sound samples, using
Time-Aliased-FFTs as described at:

    https://github.com/waywardgeek/spectrogram

Basically, two adjacent pitch periods are overlap-added to create a sound
sample that accurately represents the speech sound at that moment in time.
This set of samples is converted to a spetral line using an FFT, and the result
is saved as a single spectral line at that moment in time.  The resulting
spectral lines vary in resolution (it is equal to the number of samples in the
pitch period), and the spacing of spectral lines also varies (proportional to
the numver of samples in the pitch period).

To generate a bitmap, linear interpolation is used to render the grayscale
value at any particular point in time and frequency.
*/

#define SONIC_MAX_SPECTRUM_FREQ 5000

struct sonicSpectrogramStruct;
struct sonicBitmapStruct;
typedef struct sonicSpectrogramStruct* sonicSpectrogram;
typedef struct sonicBitmapStruct* sonicBitmap;

/* sonicBitmap objects represent spectrograms as grayscale bitmaps where each
   pixel is from 0 (black) to 255 (white).  Bitmaps are rows*cols in size.
   Rows are indexed top to bottom and columns are indexed left to right */
struct sonicBitmapStruct {
  unsigned char* data;
  int numRows;
  int numCols;
};

typedef struct sonicBitmapStruct* sonicBitmap;

/* Enable coomputation of a spectrogram on the fly. */
void sonicComputeSpectrogram(sonicStream stream);

/* Get the spectrogram. */
sonicSpectrogram sonicGetSpectrogram(sonicStream stream);

/* Create an empty spectrogram. Called automatically if sonicComputeSpectrogram
   has been called. */
sonicSpectrogram sonicCreateSpectrogram(int sampleRate);

/* Destroy the spectrotram.  This is called automatically when calling
   sonicDestroyStream. */
void sonicDestroySpectrogram(sonicSpectrogram spectrogram);

/* Convert the spectrogram to a bitmap. Caller must destroy bitmap when done. */
sonicBitmap sonicConvertSpectrogramToBitmap(sonicSpectrogram spectrogram,
                                            int numRows, int numCols);

/* Destroy a bitmap returned by sonicConvertSpectrogramToBitmap. */
void sonicDestroyBitmap(sonicBitmap bitmap);

int sonicWritePGM(sonicBitmap bitmap, char* fileName);

/* Add two pitch periods worth of samples to the spectrogram.  There must be
   2*period samples.  Time should advance one pitch period for each call to
   this function. */
void sonicAddPitchPeriodToSpectrogram(sonicSpectrogram spectrogram,
                                      short* samples, int period,
                                      int numChannels);
#endif  /* SONIC_SPECTROGRAM */

#ifdef __cplusplus
}
#endif