1
0
mirror of https://github.com/cookiengineer/audacity synced 2025-04-30 07:39:42 +02:00
2013-10-24 04:32:13 +00:00

770 lines
24 KiB
C

/*
* TwoLAME: an optimized MPEG Audio Layer Two encoder
*
* Copyright (C) 2004-2007 The TwoLAME Project
*
* This library 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.
*
* This library 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 this library; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*
* $Id$
*
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <getopt.h>
#include <twolame.h>
#include <sndfile.h>
#include "frontend.h"
/*
Global Variables
*/
int use_raw = FALSE; // use raw input?
int sample_size = 16; // number of bits per sample for raw input
int single_frame_mode = FALSE; // only encode a single frame of MPEG audio ?
int byteswap = FALSE; // swap endian on input audio ?
int channelswap = FALSE; // swap left and right channels ?
SF_INFO sfinfo; // contains information about input file format
char inputfilename[MAX_NAME_SIZE] = "\0";
char outputfilename[MAX_NAME_SIZE] = "\0";
/*
new_extension()
Puts a new extension name on a file name <filename>.
Removes the last extension name, if any.
*/
static void new_extension(char *filename, char *extname, char *newname)
{
int found, dotpos;
// First, strip the old extension
dotpos = strlen(filename);
found = 0;
do {
switch (filename[dotpos]) {
case '.':
found = 1;
break;
case '\\':
case '/':
case ':':
found = -1;
break;
default:
dotpos--;
if (dotpos < 0)
found = -1;
break;
}
} while (found == 0);
if (found == -1) {
strncpy(newname, filename, MAX_NAME_SIZE);
}
if (found == 1) {
strncpy(newname, filename, dotpos);
newname[dotpos] = '\0';
}
// Make sure there is room in the string for the
// new filename and the extension
if (strlen(newname) + strlen(extname) + 1 < MAX_NAME_SIZE) {
strcat(newname, extname);
}
}
static char *format_filesize_string(int filesize)
{
static const int constKB = 1024; // Kilobyte
static const int constMB = 1024 * 1024; // Megabyte
static const int constGB = 1024 * 1024 * 1024; // Gigabyte
char *string = malloc(MAX_NAME_SIZE);
if (filesize < constKB) {
snprintf(string, MAX_NAME_SIZE, "%d bytes", filesize);
} else if (filesize < constMB) {
snprintf(string, MAX_NAME_SIZE, "%2.2f KB", (float) filesize / constKB);
} else if (filesize < constGB) {
snprintf(string, MAX_NAME_SIZE, "%2.2f MB", (float) filesize / constMB);
} else {
snprintf(string, MAX_NAME_SIZE, "%2.2f GB", (float) filesize / constGB);
}
return string;
}
/*
print_filenames()
Display the input and output filenames
*/
static void print_filenames(int verbosity)
{
char *ifn, *ofn;
if (strcmp(inputfilename, "-") == 0)
ifn = "STDIN";
else
ifn = inputfilename;
if (strcmp(outputfilename, "-") == 0)
ofn = "STDOUT";
else
ofn = outputfilename;
if (verbosity == 1) {
fprintf(stderr, "Encoding %s to %s\n", ifn, ofn);
} else if (verbosity > 1) {
fprintf(stderr, "---------------------------------------------------------\n");
fprintf(stderr, "Input Filename: %s\n", ifn);
fprintf(stderr, "Output Filename: %s\n", ofn);
}
}
/*
usage_long()
Display the extended usage information
*/
static void usage_long()
{
fprintf(stderr, "TwoLAME version %s (%s)\n", get_twolame_version(), get_twolame_url());
fprintf(stderr, "MPEG Audio Layer II (MP2) encoder\n");
fprintf(stderr, "Usage: \n");
fprintf(stderr, "\ttwolame [options] <infile> [outfile]\n");
fprintf(stderr, "\n");
fprintf(stderr, "Both input and output filenames can be set to - to use stdin/stdout.\n");
fprintf(stderr, " <infile> input sound file (any format supported by libsndfile)\n");
fprintf(stderr, " <outfile> output bit stream of encoded audio\n");
fprintf(stderr, "\nInput Options\n");
fprintf(stderr, "\t-r, --raw-input input is raw signed PCM audio\n");
fprintf(stderr, "\t-x, --byte-swap force byte-swapping of input\n");
fprintf(stderr, "\t-s, --samplerate srate sampling frequency of raw input (Hz)\n");
fprintf(stderr,
"\t --samplesize bits size of raw input samples in bits (default 16-bit)\n");
fprintf(stderr, "\t-N, --channels nch number of channels in raw input\n");
fprintf(stderr, "\t-g, --swap-channels swap channels of input file\n");
fprintf(stderr, "\t --scale value scale input (multiply PCM data)\n");
fprintf(stderr, "\t --scale-l value scale channel 0 (left) input\n");
fprintf(stderr, "\t --scale-r value scale channel 1 (right) input\n");
fprintf(stderr, "\nOutput Options\n");
fprintf(stderr, "\t-m, --mode mode (s)tereo, (j)oint, (d)ual, (m)ono or (a)uto\n");
fprintf(stderr,
"\t-a, --downmix downmix from stereo to mono file for mono encoding\n");
fprintf(stderr, "\t-b, --bitrate br total bitrate in kbps (default 192 for 44.1kHz)\n");
fprintf(stderr, "\t-P, --psyc-mode psyc psychoacoustic model -1 to 4 (default 3)\n");
fprintf(stderr, "\t-v, --vbr enable VBR mode\n");
fprintf(stderr,
"\t-V, --vbr-level lev enable VBR and set VBR level -50 to 50 (default 5)\n");
fprintf(stderr, "\t-B, --max-bitrate rate set the upper bitrate when in VBR mode\n");
fprintf(stderr, "\t-l, --ath lev ATH level (default 0.0)\n");
fprintf(stderr, "\t-q, --quick num only calculate psy model every num frames\n");
fprintf(stderr, "\t-S, --single-frame only encode a single frame of MPEG Audio\n");
fprintf(stderr, "\nMiscellaneous Options\n");
fprintf(stderr, "\t-c, --copyright mark as copyright\n");
fprintf(stderr, "\t --non-copyright mark as non-copyright (default)\n");
fprintf(stderr, "\t-o, --non-original mark as non-original\n");
fprintf(stderr, "\t --original mark as original (default)\n");
fprintf(stderr, "\t-p, --protect enable CRC error protection\n");
fprintf(stderr, "\t-d, --padding force padding bit/frame on\n");
fprintf(stderr, "\t-R, --reserve-bits num set number of reserved bits in each frame\n");
fprintf(stderr, "\t-e, --deemphasis emp de-emphasis n/5/c (default: (n)one)\n");
fprintf(stderr, "\t-E, --energy turn on energy level extensions\n");
fprintf(stderr, "\nVerbosity Options\n");
fprintf(stderr, "\t-t, --talkativity num talkativity 0-10 (default is 2)\n");
fprintf(stderr, "\t --quiet same as --talkativity=0\n");
fprintf(stderr, "\t --brief same as --talkativity=1\n");
fprintf(stderr, "\t --verbose same as --talkativity=4\n");
fprintf(stderr, "\n");
fprintf(stderr, "\nAllowable bitrates for 32, 44.1 and 48kHz sample input (MPEG-1)\n");
fprintf(stderr, " 32, 48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 256, 320, 384\n");
fprintf(stderr, "\nAllowable bitrates for 16, 22.05 and 24kHz sample input (MPEG-2)\n");
fprintf(stderr, " 8, 16, 24, 32, 40, 48, 56, 64, 80, 96, 112, 128, 144, 160\n");
fprintf(stderr, "\n");
exit(ERR_NO_ENCODE);
}
/*
usage_short()
Display the short usage information
*/
static void usage_short()
{
/* print a bit of info about the program */
fprintf(stderr, "TwoLAME version %s (%s)\n", get_twolame_version(), get_twolame_url());
fprintf(stderr, "MPEG Audio Layer II (MP2) encoder\n\n");
fprintf(stderr, "Usage: twolame [options] <infile> [outfile]\n\n");
fprintf(stderr, "Try \"twolame --help\" for more information.\n");
exit(ERR_NO_ENCODE);
}
/*
build_shortopt_string()
Creates a short args string from the options structure
for use with getopt_long
*/
static char *build_shortopt_string(struct option *opts)
{
int count = 0;
char *shortstr = NULL;
int c = 0, n = 0;
// Start by counting the number of options
while (opts[count].val != 0) {
count++;
}
// Allocate memory for the string
shortstr = malloc((count * 2) + 1);
// And loop through the options again
for (n = 0; opts[n].val != 0; n++) {
if (opts[n].val > 0 && opts[n].val < 127) {
shortstr[c++] = opts[n].val;
if (opts[n].has_arg == optional_argument) {
fprintf(stderr, "gah: can't do optional arguments\n");
} else if (opts[n].has_arg == required_argument) {
shortstr[c++] = ':';
}
}
}
// Finally - terminate the string
shortstr[c] = '\0';
return shortstr;
}
/*
parse_args()
Parse the command line arguments
*/
static void parse_args(int argc, char **argv, twolame_options * encopts)
{
int ch = 0;
// process args
struct option longopts[] = {
// Input
{"raw-input", no_argument, NULL, 'r'},
{"byte-swap", no_argument, NULL, 'x'},
{"samplerate", required_argument, NULL, 's'},
{"samplesize", required_argument, NULL, 1000},
{"channels", required_argument, NULL, 'N'},
{"swap-channels", no_argument, NULL, 'g'},
{"scale", required_argument, NULL, 1001},
{"scale-l", required_argument, NULL, 1002},
{"scale-r", required_argument, NULL, 1003},
// Output
{"mode", required_argument, NULL, 'm'},
{"downmix", no_argument, NULL, 'a'},
{"bitrate", required_argument, NULL, 'b'},
{"psyc-mode", required_argument, NULL, 'P'},
{"vbr", no_argument, NULL, 'v'},
{"vbr-level", required_argument, NULL, 'V'},
{"max-bitrate", required_argument, NULL, 'B'},
{"ath", required_argument, NULL, 'l'},
{"quick", required_argument, NULL, 'q'},
{"single-frame", no_argument, NULL, 'S'},
// Misc
{"copyright", no_argument, NULL, 'c'},
{"non-copyright", no_argument, NULL, 1004},
{"non-original", no_argument, NULL, 'o'},
{"original", no_argument, NULL, 1005},
{"protect", no_argument, NULL, 'p'},
{"padding", no_argument, NULL, 'd'},
{"reserve-bits", required_argument, NULL, 'R'},
{"deemphasis", required_argument, NULL, 'e'},
{"energy", no_argument, NULL, 'E'},
// Verbosity
{"talkativity", required_argument, NULL, 't'},
{"quiet", no_argument, NULL, 1006},
{"brief", no_argument, NULL, 1007},
{"verbose", no_argument, NULL, 1008},
{"help", no_argument, NULL, 'h'},
{NULL, 0, NULL, 0}
};
// Create a short options structure from the long one
char *shortopts = build_shortopt_string(longopts);
// fprintf(stderr,"shortopts: %s\n", shortopts);
// Input format defaults
memset(&sfinfo, 0, sizeof(sfinfo));
sfinfo.format = 0;
sfinfo.samplerate = DEFAULT_SAMPLERATE;
sfinfo.channels = DEFAULT_CHANNELS;
sfinfo.frames = 0;
while ((ch = getopt_long(argc, argv, shortopts, longopts, NULL)) != -1) {
switch (ch) {
// Input
case 'r':
use_raw = 1;
break;
case 'x':
byteswap = TRUE;
break;
case 's':
twolame_set_out_samplerate(encopts, atoi(optarg));
sfinfo.samplerate = atoi(optarg);
break;
case 1000: // --samplesize
sample_size = atoi(optarg);
break;
case 'N':
sfinfo.channels = atoi(optarg);
break;
case 'g':
channelswap = TRUE;
break;
case 1001: // --scale
twolame_set_scale(encopts, atof(optarg));
break;
case 1002: // --scale-l
twolame_set_scale_left(encopts, atof(optarg));
break;
case 1003: // --scale-r
twolame_set_scale_right(encopts, atof(optarg));
break;
// Output
case 'm':
if (*optarg == 's') {
twolame_set_mode(encopts, TWOLAME_STEREO);
} else if (*optarg == 'd') {
twolame_set_mode(encopts, TWOLAME_DUAL_CHANNEL);
} else if (*optarg == 'j') {
twolame_set_mode(encopts, TWOLAME_JOINT_STEREO);
} else if (*optarg == 'm') {
twolame_set_mode(encopts, TWOLAME_MONO);
} else if (*optarg == 'a') {
twolame_set_mode(encopts, TWOLAME_AUTO_MODE);
} else {
fprintf(stderr, "Error: mode must be a/s/d/j/m not '%s'\n\n", optarg);
usage_long();
}
break;
case 'a': // downmix
twolame_set_mode(encopts, TWOLAME_MONO);
break;
case 'b':
twolame_set_bitrate(encopts, atoi(optarg));
break;
case 'P':
twolame_set_psymodel(encopts, atoi(optarg));
break;
case 'v':
twolame_set_VBR(encopts, TRUE);
break;
case 'V':
twolame_set_VBR(encopts, TRUE);
twolame_set_VBR_level(encopts, atof(optarg));
break;
case 'B':
twolame_set_VBR_max_bitrate_kbps(encopts, atoi(optarg));
break;
case 'l':
twolame_set_ATH_level(encopts, atof(optarg));
break;
case 'q':
twolame_set_quick_mode(encopts, TRUE);
twolame_set_quick_count(encopts, atoi(optarg));
break;
case 'S':
single_frame_mode = TRUE;
break;
// Miscellaneous
case 'c':
twolame_set_copyright(encopts, TRUE);
break;
case 1004: // --non-copyright
twolame_set_copyright(encopts, FALSE);
break;
case 'o': // --non-original
twolame_set_original(encopts, FALSE);
break;
case 1005: // --original
twolame_set_original(encopts, TRUE);
break;
case 'p':
twolame_set_error_protection(encopts, TRUE);
break;
case 'd':
twolame_set_padding(encopts, TWOLAME_PAD_ALL);
break;
case 'R':
twolame_set_num_ancillary_bits(encopts, atoi(optarg));
break;
case 'e':
if (*optarg == 'n')
twolame_set_emphasis(encopts, TWOLAME_EMPHASIS_N);
else if (*optarg == '5')
twolame_set_emphasis(encopts, TWOLAME_EMPHASIS_5);
else if (*optarg == 'c')
twolame_set_emphasis(encopts, TWOLAME_EMPHASIS_C);
else {
fprintf(stderr, "Error: emphasis must be n/5/c not '%s'\n\n", optarg);
usage_long();
}
break;
case 'E':
twolame_set_energy_levels(encopts, TRUE);
break;
// Verbosity
case 't':
twolame_set_verbosity(encopts, atoi(optarg));
break;
case 1006: // --quiet
twolame_set_verbosity(encopts, 0);
break;
case 1007: // --brief
twolame_set_verbosity(encopts, 1);
break;
case 1008: // --verbose
twolame_set_verbosity(encopts, 4);
break;
case 'h':
usage_long();
break;
default:
usage_short();
break;
}
}
// Look for the input and output file names
argc -= optind;
argv += optind;
while (argc) {
if (inputfilename[0] == '\0')
strncpy(inputfilename, *argv, MAX_NAME_SIZE);
else if (outputfilename[0] == '\0')
strncpy(outputfilename, *argv, MAX_NAME_SIZE);
else {
fprintf(stderr, "excess argument: %s\n", *argv);
usage_short();
}
argv++;
argc--;
}
// Check that we now have input and output file names ok
if (inputfilename[0] == '\0') {
fprintf(stderr, "Missing input filename.\n");
usage_short();
}
if (outputfilename[0] == '\0' && strcmp(inputfilename, "-") != 0) {
// Create output filename from the inputfilename
// and change the suffix
new_extension(inputfilename, OUTPUT_SUFFIX, outputfilename);
}
if (outputfilename[0] == '\0') {
fprintf(stderr, "Missing output filename.\n");
usage_short();
}
// Check -r is supplied when reading from STDIN
if (strcmp(inputfilename, "-") == 0 && !use_raw) {
fprintf(stderr, "Error: please use RAW audio '-r' switch when reading from STDIN.\n");
usage_short();
}
}
static FILE *open_output_file(char *filename)
{
FILE *file;
// Do they want STDOUT ?
if (strncmp(filename, "-", 1) == 0) {
file = stdout;
} else {
file = fopen(filename, "wb");
}
// Check for errors
if (file == NULL) {
perror("Failed to open output file");
exit(ERR_OPENING_OUTPUT);
}
return file;
}
int main(int argc, char **argv)
{
twolame_options *encopts = NULL;
audioin_t *inputfile = NULL;
FILE *outputfile = NULL;
short int *pcmaudio = NULL;
unsigned int frame_count = 0;
unsigned int total_frames = 0;
unsigned int frame_len = 0;
unsigned int total_bytes = 0;
unsigned char *mp2buffer = NULL;
int samples_read = 0;
int mp2fill_size = 0;
int audioReadSize = 0;
// Allocate memory for the PCM audio data
if ((pcmaudio = (short int *) calloc(AUDIO_BUF_SIZE, sizeof(short int))) == NULL) {
fprintf(stderr, "Error: pcmaudio memory allocation failed\n");
exit(ERR_MEM_ALLOC);
}
// Allocate memory for the encoded MP2 audio data
if ((mp2buffer = (unsigned char *) calloc(MP2_BUF_SIZE, sizeof(unsigned char))) == NULL) {
fprintf(stderr, "Error: mp2buffer memory allocation failed\n");
exit(ERR_MEM_ALLOC);
}
// Initialise Encoder Options Structure
encopts = twolame_init();
if (encopts == NULL) {
fprintf(stderr, "Error: initializing libtwolame encoder failed.\n");
exit(ERR_MEM_ALLOC);
}
// Get options and parameters from the command line
parse_args(argc, argv, encopts);
// Display the filenames
print_filenames(twolame_get_verbosity(encopts));
// Open the input file
if (use_raw) {
// use raw input handler
inputfile = open_audioin_raw(inputfilename, &sfinfo, sample_size);
} else {
// use libsndfile
inputfile = open_audioin_sndfile(inputfilename, &sfinfo);
}
// Display input information
if (twolame_get_verbosity(encopts) > 1) {
inputfile->print_info(inputfile);
}
// Use information from input file to configure libtwolame
twolame_set_num_channels(encopts, sfinfo.channels);
twolame_set_in_samplerate(encopts, sfinfo.samplerate);
// Open the output file
outputfile = open_output_file(outputfilename);
// initialise twolame with this set of options
if (twolame_init_params(encopts) != 0) {
fprintf(stderr, "Error: configuring libtwolame encoder failed.\n");
exit(ERR_INVALID_PARAM);
}
// display encoder settings
twolame_print_config(encopts);
// Only encode a single frame of mpeg audio ?
if (single_frame_mode)
audioReadSize = TWOLAME_SAMPLES_PER_FRAME;
else
audioReadSize = AUDIO_BUF_SIZE;
// Calculate the size and number of frames we are going to encode
frame_len = twolame_get_framelength(encopts);
if (sfinfo.frames)
total_frames = sfinfo.frames / TWOLAME_SAMPLES_PER_FRAME;
// Now do the reading/encoding/writing
while ((samples_read = inputfile->read(inputfile, pcmaudio, audioReadSize)) > 0) {
int bytes_out = 0;
// Force byte swapping if requested
if (byteswap) {
int i;
for (i = 0; i < samples_read; i++) {
short tmp = pcmaudio[i];
char *src = (char *) &tmp;
char *dst = (char *) &pcmaudio[i];
dst[0] = src[1];
dst[1] = src[0];
}
}
// Calculate the number of samples we have (per channel)
samples_read /= sfinfo.channels;
// Do swapping of left and right channels if requested
if (channelswap && sfinfo.channels == 2) {
int i;
for (i = 0; i < samples_read; i++) {
short tmp = pcmaudio[(2 * i)];
pcmaudio[(2 * i)] = pcmaudio[(2 * i) + 1];
pcmaudio[(2 * i) + 1] = tmp;
}
}
// Encode the audio to MP2
mp2fill_size =
twolame_encode_buffer_interleaved(encopts, pcmaudio, samples_read, mp2buffer,
MP2_BUF_SIZE);
// Stop if we don't have any bytes (probably don't have enough audio for a full frame of
// mpeg audio)
if (mp2fill_size == 0)
break;
if (mp2fill_size < 0) {
fprintf(stderr, "error while encoding audio: %d\n", mp2fill_size);
exit(ERR_ENCODING);
}
// Check that a whole number of frame was written
// if (mp2fill_size % frame_len != 0) {
// fprintf(stderr,"error while encoding audio: non-whole number of frames written\n");
// exit(ERR_ENCODING);
// }
// Write the encoded audio out
bytes_out = fwrite(mp2buffer, sizeof(unsigned char), mp2fill_size, outputfile);
if (bytes_out != mp2fill_size) {
perror("error while writing to output file");
exit(ERR_WRITING_OUTPUT);
}
total_bytes += bytes_out;
// Only single frame ?
if (single_frame_mode)
break;
// Display Progress
frame_count += (mp2fill_size / frame_len);
if (twolame_get_verbosity(encopts) > 0) {
fprintf(stderr, "\rEncoding frame: %i", frame_count);
if (total_frames) {
fprintf(stderr, "/%i (%i%%)", total_frames, (frame_count * 100) / total_frames);
}
fflush(stderr);
}
}
// Was there an error reading the audio?
if (inputfile->error_str(inputfile)) {
fprintf(stderr, "Error reading from input file: %s\n", inputfile->error_str(inputfile));
}
//
// flush any remaining audio. (don't send any new audio data) There
// should only ever be a max of 1 frame on a flush. There may be zero
// frames if the audio data was an exact multiple of 1152
//
mp2fill_size = twolame_encode_flush(encopts, mp2buffer, MP2_BUF_SIZE);
if (mp2fill_size > 0) {
frame_count++;
int bytes_out = fwrite(mp2buffer, sizeof(unsigned char), mp2fill_size, outputfile);
if (bytes_out <= 0) {
perror("error while writing to output file");
exit(ERR_WRITING_OUTPUT);
}
total_bytes += bytes_out;
}
if (twolame_get_verbosity(encopts) > 1) {
char *filesize = format_filesize_string(total_bytes);
fprintf(stderr, "\nEncoding Finished.\n");
fprintf(stderr, "Total bytes written: %s.\n", filesize);
free(filesize);
}
// Close input and output streams
inputfile->close(inputfile);
fclose(outputfile);
// Close the libtwolame encoder
twolame_close(&encopts);
// Free up memory
free(pcmaudio);
free(mp2buffer);
return (ERR_NO_ERROR);
}
/* vim:ts=4:sw=4:nowrap: */