speech-tools/main/pda_main.cc
2015-09-19 10:52:26 +02:00

273 lines
9.1 KiB
C++

/*************************************************************************/
/* */
/* Centre for Speech Technology Research */
/* University of Edinburgh, UK */
/* Copyright (c) 1996 */
/* All Rights Reserved. */
/* */
/* Permission is hereby granted, free of charge, to use and distribute */
/* this software and its documentation without restriction, including */
/* without limitation the rights to use, copy, modify, merge, publish, */
/* distribute, sublicense, and/or sell copies of this work, and to */
/* permit persons to whom this work is furnished to do so, subject to */
/* the following conditions: */
/* 1. The code must retain the above copyright notice, this list of */
/* conditions and the following disclaimer. */
/* 2. Any modifications must be clearly marked as such. */
/* 3. Original authors' names are not deleted. */
/* 4. The authors' names are not used to endorse or promote products */
/* derived from this software without specific prior written */
/* permission. */
/* */
/* THE UNIVERSITY OF EDINBURGH AND THE CONTRIBUTORS TO THIS WORK */
/* DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING */
/* ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT */
/* SHALL THE UNIVERSITY OF EDINBURGH NOR THE CONTRIBUTORS BE LIABLE */
/* FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES */
/* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN */
/* AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, */
/* ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF */
/* THIS SOFTWARE. */
/* */
/*************************************************************************/
/* Author : Paul Taylor */
/* Date : May 1994 */
/*-----------------------------------------------------------------------*/
/* Pitch Detection Algorithm Main routine */
/* */
/*=======================================================================*/
#include <fstream>
#include "EST.h"
#include "sigpr/EST_sigpr_utt.h"
#include "EST_cmd_line_options.h"
void set_parameters(EST_Features &a_list, EST_Option &al);
void option_override(EST_Features &op, EST_Option al,
const EST_String &option, const EST_String &arg);
static int save_pm(EST_String filename, EST_Track fz);
/** @name <command>pda</command> <emphasis>Pitch Detection Algorithm</emphasis>
@id pda-manual
* @toc
*/
//@{
/**@name Synopsis
*/
//@{
//@synopsis
/**
pda is a pitch detection algorithm that produces a fundamental frequency
contour from a speech waveform file. At present only the
super resolution pitch determination algorithm is implemented.
See (Medan, Yair, and Chazan, 1991) and (Bagshaw et al., 1993) for a detailed
description of the algorithm.
</para><para>
The default values given below were found to optimise the performance
of the pitch determination algorithm for speech data sampled at 20kHz
using a 16\-bit waveform and low pass filter with a 600Hz cut-off
frequency and more than \-85dB rejection above 700Hz. The best
performances occur if the [\-p] flag is passed. </para><para>
*/
//@}
/**@name Options
*/
//@{
//@options
//@}
int main (int argc, char *argv[])
{
EST_Track fz;
EST_Wave sig;
EST_Option al;
EST_Features op;
EST_String out_file("-");
EST_StrList files;
parse_command_line
(argc, argv,
EST_String("[input file] -o [output file] [options]\n")+
"Summary: pitch track waveform files\n"
"use \"-\" to make input and output files stdin/out\n"
"-h Options help\n\n"+
options_wave_input()+
options_pda_general()+
options_pda_srpd()+
options_track_output(),
files, al);
default_pda_options(op);
set_parameters(op, al);
if (read_wave(sig, files.first(), al) != format_ok)
exit(-1);
out_file = al.present("-o") ? al.val("-o") : (EST_String)"-";
pda(sig, fz, op); // do f0 tracking
if (al.present("-pm"))
save_pm(out_file, fz);
else
fz.save(out_file, op.S("f0_file_type", "0"));
if (al.present("-diff"))
{
fz = differentiate(fz);
fz.save(out_file + ".diff", op.S("f0_file_type", "0"));
}
return 0;
}
void set_parameters(EST_Features &op, EST_Option &al)
{
op.set("srpd_resize", 1);
// general options
option_override(op, al, "pda_frame_shift", "-shift");
option_override(op, al, "pda_frame_length", "-length");
option_override(op, al, "max_pitch", "-fmax");
option_override(op, al, "min_pitch", "-fmin");
// low pass filtering options.
option_override(op, al, "lpf_cutoff", "-u");
option_override(op, al, "lpf_order", "-forder");
option_override(op, al, "decimation", "-d");
option_override(op, al, "noise_floor", "-n");
option_override(op, al, "min_v2uv_coef_thresh", "-m");
option_override(op, al, "v2uv_coef_thresh_ratio", "-R");
option_override(op, al, "v2uv_coef_thresh", "-H");
option_override(op, al, "anti_doubling_thresh", "-t");
option_override(op, al, "peak_tracking", "-P");
option_override(op, al, "f0_file_type", "-otype");
option_override(op, al, "wave_file_type", "-itype");
if (al.val("-L", 0) == "true")
op.set("do_low_pass", "true");
if (al.val("-R", 0) == "true")
op.set("do_low_pass", "false");
/* op.set("lpf_cutoff",al.val("-u", 0));
op.set("lpf_order",al.val("-forder", 0));
//sprd options
op.set("decimation", al.val("-d", 0));
op.set("noise_floor", al.val("-n", 0));
op.set("min_v2uv_coef_thresh", al.val("-m", 0));
op.set("v2uv_coef_thresh_ratio", al.val("-r", 0));
op.set("v2uv_coef_thresh", al.val("-H", 0));
op.set("anti_doubling_thresh", al.val("-t", 0));
op.set("peak_tracking", al.val("-P", 0));
if (al.val("-L", 0) == "true")
op.set("do_low_pass", "true");
if (al.val("-R", 0) == "true")
op.set("do_low_pass", "false");
op.set("f0_file_type", al.val("-otype", 0));
op.set("wave_file_type", al.val("-itype", 0));
*/
}
/* a_list.override_val("sample_rate", al.val("-f", 0));
a_list.override_val("min_pitch", al.val("-fmin", 0));
a_list.override_val("max_pitch", al.val("-fmax", 0));
a_list.override_val("pda_frame_shift", al.val("-s", 0));
a_list.override_val("pda_frame_length",al.val("-l", 0));
// low pass filtering options.
a_list.override_val("lpf_cutoff",al.val("-u", 0));
a_list.override_val("lpf_order",al.val("-forder", 0));
//sprd options
a_list.override_val("decimation", al.val("-d", 0));
a_list.override_val("noise_floor", al.val("-n", 0));
a_list.override_val("min_v2uv_coef_thresh", al.val("-m", 0));
a_list.override_val("v2uv_coef_thresh_ratio", al.val("-r", 0));
a_list.override_val("v2uv_coef_thresh", al.val("-H", 0));
a_list.override_val("anti_doubling_thresh", al.val("-t", 0));
a_list.override_val("peak_tracking", al.val("-P", 0));
if (al.val("-L", 0) == "true")
a_list.override_val("do_low_pass", "true");
if (al.val("-R", 0) == "true")
a_list.override_val("do_low_pass", "false");
a_list.override_val("f0_file_type", al.val("-otype", 0));
a_list.override_val("wave_file_type", al.val("-itype", 0));
*/
static int save_pm(EST_String filename, EST_Track fz)
{
ostream *outf;
float position, period;
if (filename == "-")
outf = &cout;
else
outf = new ofstream(filename);
if (!(*outf))
{
cerr << "save_pm: can't write to file \"" << filename << "\"" << endl;
return -1;
}
*outf << "XAO1\n\n"; // xmg header identifier.
*outf << "LineType bars \n";
*outf << "LineStyle solid \n";
*outf << "LineWidth 0 \n";
*outf << "Freq 16\n";
*outf << "Format Binary \n";
*outf << char(12) << "\n"; // control L character
position = 0.0;
int gap = 0;
for (int i = 0; i < fz.num_frames(); ++i)
{
if (fz.val(i))
{
if (gap)
{
position = fz.t(i);
gap = 0;
}
period = 1.0 / fz.a(i);
*outf << (position + period) * 1000.0 << endl;
position += period;
}
else
gap = 1;
}
if (outf != &cout)
delete outf;
return 0;
}
/**@name Examples
Pitch detection on typical male voice, using low pass filtering:
<screen>
$ pda kdt_010.wav -o kdt_010.f0 -fmin 80 -fmax 200 -L
</screen>
*/
//@{
//@}
//@}