/* * Asterisk -- An open source telephony toolkit. * * Copyright (C) 2010, Digium, Inc. * * David Vossel * * See http://www.asterisk.org for more information about * the Asterisk project. Please do not directly contact * any of the maintainers of this project for assistance; * the project provides a web site, mailing lists and IRC * channels for your use. * * This program is free software, distributed under the terms of * the GNU General Public License Version 2. See the LICENSE file * at the top of the source tree. */ /*! \file * * \brief Pitch Shift Audio Effect * * \author David Vossel * * \ingroup functions */ /************************* SMB FUNCTION LICENSE ********************************* * * SYNOPSIS: Routine for doing pitch shifting while maintaining * duration using the Short Time Fourier Transform. * * DESCRIPTION: The routine takes a pitchShift factor value which is between 0.5 * (one octave down) and 2. (one octave up). A value of exactly 1 does not change * the pitch. num_samps_to_process tells the routine how many samples in indata[0... * num_samps_to_process-1] should be pitch shifted and moved to outdata[0 ... * num_samps_to_process-1]. The two buffers can be identical (ie. it can process the * data in-place). fft_frame_size defines the FFT frame size used for the * processing. Typical values are 1024, 2048 and 4096. It may be any value <= * MAX_FRAME_LENGTH but it MUST be a power of 2. osamp is the STFT * oversampling factor which also determines the overlap between adjacent STFT * frames. It should at least be 4 for moderate scaling ratios. A value of 32 is * recommended for best quality. sampleRate takes the sample rate for the signal * in unit Hz, ie. 44100 for 44.1 kHz audio. The data passed to the routine in * indata[] should be in the range [-1.0, 1.0), which is also the output range * for the data, make sure you scale the data accordingly (for 16bit signed integers * you would have to divide (and multiply) by 32768). * * COPYRIGHT 1999-2009 Stephan M. Bernsee * * The Wide Open License (WOL) * * Permission to use, copy, modify, distribute and sell this software and its * documentation for any purpose is hereby granted without fee, provided that * the above copyright notice and this license appear in all source copies. * THIS SOFTWARE IS PROVIDED "AS IS" WITHOUT EXPRESS OR IMPLIED WARRANTY OF * ANY KIND. See http://www.dspguru.com/wol.htm for more information. * *****************************************************************************/ #include "asterisk.h" ASTERISK_FILE_VERSION(__FILE__, "$Revision$") #include "asterisk/module.h" #include "asterisk/channel.h" #include "asterisk/pbx.h" #include "asterisk/utils.h" #include "asterisk/audiohook.h" #include /*** DOCUMENTATION Pitch shift both tx and rx audio streams on a channel. Direction can be either rx, tx, or both. The direction can either be set to a valid floating point number between 0.1 and 4.0 or one of the enum values listed below. A value of 1.0 has no effect. Greater than 1 raises the pitch. Lower than 1 lowers the pitch. The pitch amount can also be set by the following values Examples: exten => 1,1,Set(PITCH_SHIFT(tx)=highest); raises pitch an octave exten => 1,1,Set(PITCH_SHIFT(rx)=higher) ; raises pitch more exten => 1,1,Set(PITCH_SHIFT(both)=high) ; raises pitch exten => 1,1,Set(PITCH_SHIFT(rx)=low) ; lowers pitch exten => 1,1,Set(PITCH_SHIFT(tx)=lower) ; lowers pitch more exten => 1,1,Set(PITCH_SHIFT(both)=lowest) ; lowers pitch an octave exten => 1,1,Set(PITCH_SHIFT(rx)=0.8) ; lowers pitch exten => 1,1,Set(PITCH_SHIFT(tx)=1.5) ; raises pitch ***/ #ifndef M_PI #define M_PI 3.14159265358979323846 #endif #define MAX_FRAME_LENGTH 256 #define HIGHEST 2 #define HIGHER 1.5 #define HIGH 1.25 #define LOW .85 #define LOWER .7 #define LOWEST .5 struct fft_data { float in_fifo[MAX_FRAME_LENGTH]; float out_fifo[MAX_FRAME_LENGTH]; float fft_worksp[2*MAX_FRAME_LENGTH]; float last_phase[MAX_FRAME_LENGTH/2+1]; float sum_phase[MAX_FRAME_LENGTH/2+1]; float output_accum[2*MAX_FRAME_LENGTH]; float ana_freq[MAX_FRAME_LENGTH]; float ana_magn[MAX_FRAME_LENGTH]; float syn_freq[MAX_FRAME_LENGTH]; float sys_magn[MAX_FRAME_LENGTH]; long gRover; float shift_amount; }; struct pitchshift_data { struct ast_audiohook audiohook; struct fft_data rx; struct fft_data tx; }; static void smb_fft(float *fft_buffer, long fft_frame_size, long sign); static void smb_pitch_shift(float pitchShift, long num_samps_to_process, long fft_frame_size, long osamp, float sample_rate, int16_t *indata, int16_t *outdata, struct fft_data *fft_data); static int pitch_shift(struct ast_frame *f, float amount, struct fft_data *fft_data); static void destroy_callback(void *data) { struct pitchshift_data *shift = data; ast_audiohook_destroy(&shift->audiohook); ast_free(shift); }; static const struct ast_datastore_info pitchshift_datastore = { .type = "pitchshift", .destroy = destroy_callback }; static int pitchshift_cb(struct ast_audiohook *audiohook, struct ast_channel *chan, struct ast_frame *f, enum ast_audiohook_direction direction) { struct ast_datastore *datastore = NULL; struct pitchshift_data *shift = NULL; if (!f) { return 0; } if ((audiohook->status == AST_AUDIOHOOK_STATUS_DONE) || (f->frametype != AST_FRAME_VOICE) || ((f->subclass.codec != AST_FORMAT_SLINEAR) && (f->subclass.codec != AST_FORMAT_SLINEAR16))) { return -1; } if (!(datastore = ast_channel_datastore_find(chan, &pitchshift_datastore, NULL))) { return -1; } shift = datastore->data; if (direction == AST_AUDIOHOOK_DIRECTION_WRITE) { pitch_shift(f, shift->tx.shift_amount, &shift->tx); } else { pitch_shift(f, shift->rx.shift_amount, &shift->rx); } return 0; } static int pitchshift_helper(struct ast_channel *chan, const char *cmd, char *data, const char *value) { struct ast_datastore *datastore = NULL; struct pitchshift_data *shift = NULL; int new = 0; float amount = 0; ast_channel_lock(chan); if (!(datastore = ast_channel_datastore_find(chan, &pitchshift_datastore, NULL))) { ast_channel_unlock(chan); if (!(datastore = ast_datastore_alloc(&pitchshift_datastore, NULL))) { return 0; } if (!(shift = ast_calloc(1, sizeof(*shift)))) { ast_datastore_free(datastore); return 0; } ast_audiohook_init(&shift->audiohook, AST_AUDIOHOOK_TYPE_MANIPULATE, "pitch_shift"); shift->audiohook.manipulate_callback = pitchshift_cb; datastore->data = shift; new = 1; } else { ast_channel_unlock(chan); shift = datastore->data; } if (!strcasecmp(value, "highest")) { amount = HIGHEST; } else if (!strcasecmp(value, "higher")) { amount = HIGHER; } else if (!strcasecmp(value, "high")) { amount = HIGH; } else if (!strcasecmp(value, "lowest")) { amount = LOWEST; } else if (!strcasecmp(value, "lower")) { amount = LOWER; } else if (!strcasecmp(value, "low")) { amount = LOW; } else { if (!sscanf(value, "%30f", &amount) || (amount <= 0) || (amount > 4)) { goto cleanup_error; } } if (!strcasecmp(data, "rx")) { shift->rx.shift_amount = amount; } else if (!strcasecmp(data, "tx")) { shift->tx.shift_amount = amount; } else if (!strcasecmp(data, "both")) { shift->rx.shift_amount = amount; shift->tx.shift_amount = amount; } else { goto cleanup_error; } if (new) { ast_channel_lock(chan); ast_channel_datastore_add(chan, datastore); ast_channel_unlock(chan); ast_audiohook_attach(chan, &shift->audiohook); } return 0; cleanup_error: ast_log(LOG_ERROR, "Invalid argument provided to the %s function\n", cmd); if (new) { ast_datastore_free(datastore); } return -1; } static void smb_fft(float *fft_buffer, long fft_frame_size, long sign) { float wr, wi, arg, *p1, *p2, temp; float tr, ti, ur, ui, *p1r, *p1i, *p2r, *p2i; long i, bitm, j, le, le2, k; for (i = 2; i < 2 * fft_frame_size - 2; i += 2) { for (bitm = 2, j = 0; bitm < 2 * fft_frame_size; bitm <<= 1) { if (i & bitm) { j++; } j <<= 1; } if (i < j) { p1 = fft_buffer + i; p2 = fft_buffer + j; temp = *p1; *(p1++) = *p2; *(p2++) = temp; temp = *p1; *p1 = *p2; *p2 = temp; } } for (k = 0, le = 2; k < (long) (log(fft_frame_size) / log(2.) + .5); k++) { le <<= 1; le2 = le>>1; ur = 1.0; ui = 0.0; arg = M_PI / (le2>>1); wr = cos(arg); wi = sign * sin(arg); for (j = 0; j < le2; j += 2) { p1r = fft_buffer+j; p1i = p1r + 1; p2r = p1r + le2; p2i = p2r + 1; for (i = j; i < 2 * fft_frame_size; i += le) { tr = *p2r * ur - *p2i * ui; ti = *p2r * ui + *p2i * ur; *p2r = *p1r - tr; *p2i = *p1i - ti; *p1r += tr; *p1i += ti; p1r += le; p1i += le; p2r += le; p2i += le; } tr = ur * wr - ui * wi; ui = ur * wi + ui * wr; ur = tr; } } } static void smb_pitch_shift(float pitchShift, long num_samps_to_process, long fft_frame_size, long osamp, float sample_rate, int16_t *indata, int16_t *outdata, struct fft_data *fft_data) { float *in_fifo = fft_data->in_fifo; float *out_fifo = fft_data->out_fifo; float *fft_worksp = fft_data->fft_worksp; float *last_phase = fft_data->last_phase; float *sum_phase = fft_data->sum_phase; float *output_accum = fft_data->output_accum; float *ana_freq = fft_data->ana_freq; float *ana_magn = fft_data->ana_magn; float *syn_freq = fft_data->syn_freq; float *sys_magn = fft_data->sys_magn; double magn, phase, tmp, window, real, imag; double freq_per_bin, expct; long i,k, qpd, index, in_fifo_latency, step_size, fft_frame_size2; /* set up some handy variables */ fft_frame_size2 = fft_frame_size / 2; step_size = fft_frame_size / osamp; freq_per_bin = sample_rate / (double) fft_frame_size; expct = 2. * M_PI * (double) step_size / (double) fft_frame_size; in_fifo_latency = fft_frame_size-step_size; if (fft_data->gRover == 0) { fft_data->gRover = in_fifo_latency; } /* main processing loop */ for (i = 0; i < num_samps_to_process; i++){ /* As long as we have not yet collected enough data just read in */ in_fifo[fft_data->gRover] = indata[i]; outdata[i] = out_fifo[fft_data->gRover - in_fifo_latency]; fft_data->gRover++; /* now we have enough data for processing */ if (fft_data->gRover >= fft_frame_size) { fft_data->gRover = in_fifo_latency; /* do windowing and re,im interleave */ for (k = 0; k < fft_frame_size;k++) { window = -.5 * cos(2. * M_PI * (double) k / (double) fft_frame_size) + .5; fft_worksp[2*k] = in_fifo[k] * window; fft_worksp[2*k+1] = 0.; } /* ***************** ANALYSIS ******************* */ /* do transform */ smb_fft(fft_worksp, fft_frame_size, -1); /* this is the analysis step */ for (k = 0; k <= fft_frame_size2; k++) { /* de-interlace FFT buffer */ real = fft_worksp[2*k]; imag = fft_worksp[2*k+1]; /* compute magnitude and phase */ magn = 2. * sqrt(real * real + imag * imag); phase = atan2(imag, real); /* compute phase difference */ tmp = phase - last_phase[k]; last_phase[k] = phase; /* subtract expected phase difference */ tmp -= (double) k * expct; /* map delta phase into +/- Pi interval */ qpd = tmp / M_PI; if (qpd >= 0) { qpd += qpd & 1; } else { qpd -= qpd & 1; } tmp -= M_PI * (double) qpd; /* get deviation from bin frequency from the +/- Pi interval */ tmp = osamp * tmp / (2. * M_PI); /* compute the k-th partials' true frequency */ tmp = (double) k * freq_per_bin + tmp * freq_per_bin; /* store magnitude and true frequency in analysis arrays */ ana_magn[k] = magn; ana_freq[k] = tmp; } /* ***************** PROCESSING ******************* */ /* this does the actual pitch shifting */ memset(sys_magn, 0, fft_frame_size * sizeof(float)); memset(syn_freq, 0, fft_frame_size * sizeof(float)); for (k = 0; k <= fft_frame_size2; k++) { index = k * pitchShift; if (index <= fft_frame_size2) { sys_magn[index] += ana_magn[k]; syn_freq[index] = ana_freq[k] * pitchShift; } } /* ***************** SYNTHESIS ******************* */ /* this is the synthesis step */ for (k = 0; k <= fft_frame_size2; k++) { /* get magnitude and true frequency from synthesis arrays */ magn = sys_magn[k]; tmp = syn_freq[k]; /* subtract bin mid frequency */ tmp -= (double) k * freq_per_bin; /* get bin deviation from freq deviation */ tmp /= freq_per_bin; /* take osamp into account */ tmp = 2. * M_PI * tmp / osamp; /* add the overlap phase advance back in */ tmp += (double) k * expct; /* accumulate delta phase to get bin phase */ sum_phase[k] += tmp; phase = sum_phase[k]; /* get real and imag part and re-interleave */ fft_worksp[2*k] = magn * cos(phase); fft_worksp[2*k+1] = magn * sin(phase); } /* zero negative frequencies */ for (k = fft_frame_size + 2; k < 2 * fft_frame_size; k++) { fft_worksp[k] = 0.; } /* do inverse transform */ smb_fft(fft_worksp, fft_frame_size, 1); /* do windowing and add to output accumulator */ for (k = 0; k < fft_frame_size; k++) { window = -.5 * cos(2. * M_PI * (double) k / (double) fft_frame_size) + .5; output_accum[k] += 2. * window * fft_worksp[2*k] / (fft_frame_size2 * osamp); } for (k = 0; k < step_size; k++) { out_fifo[k] = output_accum[k]; } /* shift accumulator */ memmove(output_accum, output_accum+step_size, fft_frame_size * sizeof(float)); /* move input FIFO */ for (k = 0; k < in_fifo_latency; k++) { in_fifo[k] = in_fifo[k+step_size]; } } } } static int pitch_shift(struct ast_frame *f, float amount, struct fft_data *fft) { int16_t *fun = (int16_t *) f->data.ptr; int samples; /* an amount of 1 has no effect */ if (!amount || amount == 1 || !fun || (f->samples % 32)) { return 0; } for (samples = 0; samples < f->samples; samples += 32) { smb_pitch_shift(amount, 32, MAX_FRAME_LENGTH, 32, ast_format_rate(f->subclass.codec), fun+samples, fun+samples, fft); } return 0; } static struct ast_custom_function pitch_shift_function = { .name = "PITCH_SHIFT", .write = pitchshift_helper, }; static int unload_module(void) { return ast_custom_function_unregister(&pitch_shift_function); } static int load_module(void) { int res = ast_custom_function_register(&pitch_shift_function); return res ? AST_MODULE_LOAD_DECLINE : AST_MODULE_LOAD_SUCCESS; } AST_MODULE_INFO_STANDARD(ASTERISK_GPL_KEY, "Audio Effects Dialplan Functions");