mirror of https://git.ffmpeg.org/ffmpeg.git
lavfi: add flite audio source
This commit is contained in:
Stefano Sabatini
parent
fd2b6c1e4b
commit
66a5201819
|
|
@ -37,6 +37,7 @@ version next:
|
|||
- caca output device using libcaca
|
||||
- alphaextract and alphamerge filters
|
||||
- concat filter
|
||||
- flite filter
|
||||
|
||||
|
||||
version 0.11:
|
||||
|
|
|
|||
|
|
@ -180,6 +180,7 @@ External library support:
|
|||
and libraw1394 [no]
|
||||
--enable-libfaac enable AAC encoding via libfaac [no]
|
||||
--enable-libfdk-aac enable AAC encoding via libfdk-aac [no]
|
||||
--enable-libflite enable flite (voice synthesis) support via libflite [no]
|
||||
--enable-libfreetype enable libfreetype [no]
|
||||
--enable-libgsm enable GSM de/encoding via libgsm [no]
|
||||
--enable-libiec61883 enable iec61883 via libiec61883 [no]
|
||||
|
|
@ -1074,6 +1075,7 @@ CONFIG_LIST="
|
|||
libdc1394
|
||||
libfaac
|
||||
libfdk_aac
|
||||
libflite
|
||||
libfreetype
|
||||
libgsm
|
||||
libiec61883
|
||||
|
|
@ -1787,6 +1789,7 @@ cropdetect_filter_deps="gpl"
|
|||
delogo_filter_deps="gpl"
|
||||
deshake_filter_deps="avcodec"
|
||||
drawtext_filter_deps="libfreetype"
|
||||
flite_filter_deps="libflite"
|
||||
frei0r_filter_deps="frei0r dlopen"
|
||||
frei0r_filter_extralibs='$ldl'
|
||||
frei0r_src_filter_deps="frei0r dlopen"
|
||||
|
|
@ -3390,6 +3393,8 @@ enabled libcelt && require libcelt celt/celt.h celt_decode -lcelt0 &&
|
|||
enabled libcaca && require_pkg_config caca caca.h caca_create_canvas
|
||||
enabled libfaac && require2 libfaac "stdint.h faac.h" faacEncGetVersion -lfaac
|
||||
enabled libfdk_aac && require libfdk_aac fdk-aac/aacenc_lib.h aacEncOpen -lfdk-aac
|
||||
flite_libs="-lflite_cmu_time_awb -lflite_cmu_us_awb -lflite_cmu_us_kal -lflite_cmu_us_kal16 -lflite_cmu_us_rms -lflite_cmu_us_slt -lflite_usenglish -lflite_cmulex -lflite"
|
||||
enabled libflite && require2 libflite "flite/flite.h" flite_init $flite_libs
|
||||
enabled libfreetype && require_pkg_config freetype2 "ft2build.h freetype/freetype.h" FT_Init_FreeType
|
||||
enabled libgsm && require libgsm gsm/gsm.h gsm_create -lgsm
|
||||
enabled libilbc && require libilbc ilbc.h WebRtcIlbcfix_InitDecode -lilbc
|
||||
|
|
|
|||
|
|
@ -1026,6 +1026,65 @@ Channel layout of the audio data, in the form that can be accepted by
|
|||
|
||||
All the parameters need to be explicitly defined.
|
||||
|
||||
@section flite
|
||||
|
||||
Synthesize a voice utterance using the libflite library.
|
||||
|
||||
To enable compilation of this filter you need to configure FFmpeg with
|
||||
@code{--enable-libflite}.
|
||||
|
||||
The source accepts parameters as a list of @var{key}=@var{value} pairs,
|
||||
separated by ":".
|
||||
|
||||
The description of the accepted parameters follows.
|
||||
|
||||
@table @option
|
||||
|
||||
@item list_voices
|
||||
If set to 1, list the names of the available voices and exit
|
||||
immediately. Default value is 0.
|
||||
|
||||
@item nb_samples, n
|
||||
Set the maximum number of samples per frame. Default value is 512.
|
||||
|
||||
@item textfile
|
||||
Set the filename containing the text to speak.
|
||||
|
||||
@item text
|
||||
Set the text to speak.
|
||||
|
||||
@item voice, v
|
||||
Set the voice to use for the speech synthesis. Default value is
|
||||
@code{kal}. See also the @var{list_voices} option.
|
||||
@end table
|
||||
|
||||
@section Examples
|
||||
|
||||
@itemize
|
||||
@item
|
||||
Read from file @file{speech.txt}, and synthetize the text using the
|
||||
standard flite voice:
|
||||
@example
|
||||
flite=textfile=speech.txt
|
||||
@end example
|
||||
|
||||
@item
|
||||
Read the specified text selecting the @code{slt} voice:
|
||||
@example
|
||||
flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
|
||||
@end example
|
||||
|
||||
@item
|
||||
Make @file{ffplay} speech the specified text, using @code{flite} and
|
||||
the @code{lavfi} device:
|
||||
@example
|
||||
ffplay -f lavfi flite='No more be grieved for which that thou hast done.'
|
||||
@end example
|
||||
@end itemize
|
||||
|
||||
For more information about libflite, check:
|
||||
@url{http://www.speech.cs.cmu.edu/flite/}
|
||||
|
||||
@c man end AUDIO SOURCES
|
||||
|
||||
@chapter Audio Sinks
|
||||
|
|
|
|||
|
|
@ -70,6 +70,7 @@ OBJS-$(CONFIG_VOLUME_FILTER) += af_volume.o
|
|||
|
||||
OBJS-$(CONFIG_AEVALSRC_FILTER) += asrc_aevalsrc.o
|
||||
OBJS-$(CONFIG_ANULLSRC_FILTER) += asrc_anullsrc.o
|
||||
OBJS-$(CONFIG_FLITE_FILTER) += asrc_flite.o
|
||||
|
||||
OBJS-$(CONFIG_ABUFFERSINK_FILTER) += sink_buffer.o
|
||||
OBJS-$(CONFIG_ANULLSINK_FILTER) += asink_anullsink.o
|
||||
|
|
|
|||
|
|
@ -59,6 +59,7 @@ void avfilter_register_all(void)
|
|||
|
||||
REGISTER_FILTER (AEVALSRC, aevalsrc, asrc);
|
||||
REGISTER_FILTER (ANULLSRC, anullsrc, asrc);
|
||||
REGISTER_FILTER (FLITE, flite, asrc);
|
||||
|
||||
REGISTER_FILTER (ABUFFERSINK, abuffersink, asink);
|
||||
REGISTER_FILTER (ANULLSINK, anullsink, asink);
|
||||
|
|
|
|||
|
|
@ -0,0 +1,274 @@
|
|||
/*
|
||||
* Copyright (c) 2012 Stefano Sabatini
|
||||
*
|
||||
* This file is part of FFmpeg.
|
||||
*
|
||||
* FFmpeg 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.
|
||||
*
|
||||
* FFmpeg 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 FFmpeg; if not, write to the Free Software
|
||||
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
||||
*/
|
||||
|
||||
/**
|
||||
* @file
|
||||
* flite voice synth source
|
||||
*/
|
||||
|
||||
#include <flite/flite.h>
|
||||
#include "libavutil/audioconvert.h"
|
||||
#include "libavutil/file.h"
|
||||
#include "libavutil/opt.h"
|
||||
#include "avfilter.h"
|
||||
#include "audio.h"
|
||||
#include "formats.h"
|
||||
#include "internal.h"
|
||||
|
||||
typedef struct {
|
||||
const AVClass *class;
|
||||
char *voice_str;
|
||||
char *textfile;
|
||||
char *text;
|
||||
cst_wave *wave;
|
||||
int16_t *wave_samples;
|
||||
int wave_nb_samples;
|
||||
int list_voices;
|
||||
cst_voice *voice;
|
||||
int64_t pts;
|
||||
int frame_nb_samples; ///< number of samples per frame
|
||||
} FliteContext;
|
||||
|
||||
#define OFFSET(x) offsetof(FliteContext, x)
|
||||
|
||||
static const AVOption flite_options[] = {
|
||||
{ "list_voices", "list voices and exit", OFFSET(list_voices), AV_OPT_TYPE_INT, {.dbl=0}, 0, 1 },
|
||||
{ "nb_samples", "set number of samples per frame", OFFSET(frame_nb_samples), AV_OPT_TYPE_INT, {.dbl=512}, 0, INT_MAX },
|
||||
{ "n", "set number of samples per frame", OFFSET(frame_nb_samples), AV_OPT_TYPE_INT, {.dbl=512}, 0, INT_MAX },
|
||||
{ "text", "set text to speak", OFFSET(text), AV_OPT_TYPE_STRING, {.str=NULL}, CHAR_MIN, CHAR_MAX },
|
||||
{ "textfile", "set filename of the text to speak", OFFSET(textfile), AV_OPT_TYPE_STRING, {.str=NULL}, CHAR_MIN, CHAR_MAX },
|
||||
{ "v", "set voice", OFFSET(voice_str), AV_OPT_TYPE_STRING, {.str="kal"}, CHAR_MIN, CHAR_MAX },
|
||||
{ "voice", "set voice", OFFSET(voice_str), AV_OPT_TYPE_STRING, {.str="kal"}, CHAR_MIN, CHAR_MAX },
|
||||
{ NULL }
|
||||
};
|
||||
|
||||
AVFILTER_DEFINE_CLASS(flite);
|
||||
|
||||
static volatile int flite_inited = 0;
|
||||
|
||||
/* declare functions for all the supported voices */
|
||||
#define DECLARE_REGISTER_VOICE_FN(name) cst_voice *register_cmu_us_## name(const char *)
|
||||
DECLARE_REGISTER_VOICE_FN(awb);
|
||||
DECLARE_REGISTER_VOICE_FN(kal);
|
||||
DECLARE_REGISTER_VOICE_FN(kal16);
|
||||
DECLARE_REGISTER_VOICE_FN(rms);
|
||||
DECLARE_REGISTER_VOICE_FN(slt);
|
||||
|
||||
struct voice_entry {
|
||||
const char *name;
|
||||
cst_voice * (*register_fn)(const char *);
|
||||
} voice_entry;
|
||||
|
||||
static struct voice_entry voice_entries[] = {
|
||||
{ "awb", register_cmu_us_awb },
|
||||
{ "kal", register_cmu_us_kal },
|
||||
{ "kal16", register_cmu_us_kal16 },
|
||||
{ "rms", register_cmu_us_rms },
|
||||
{ "slt", register_cmu_us_slt },
|
||||
};
|
||||
|
||||
static void list_voices(void *log_ctx, const char *sep)
|
||||
{
|
||||
int i, n = FF_ARRAY_ELEMS(voice_entries);
|
||||
for (i = 0; i < n; i++)
|
||||
av_log(log_ctx, AV_LOG_INFO, "%s%s",
|
||||
voice_entries[i].name, i < (n-1) ? sep : "\n");
|
||||
}
|
||||
|
||||
static int select_voice(cst_voice **voice, const char *voice_name, void *log_ctx)
|
||||
{
|
||||
int i;
|
||||
|
||||
for (i = 0; i < FF_ARRAY_ELEMS(voice_entries); i++) {
|
||||
struct voice_entry *entry = &voice_entries[i];
|
||||
if (!strcmp(entry->name, voice_name)) {
|
||||
*voice = entry->register_fn(NULL);
|
||||
if (!*voice) {
|
||||
av_log(log_ctx, AV_LOG_ERROR,
|
||||
"Could not register voice '%s'\n", voice_name);
|
||||
return AVERROR_UNKNOWN;
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
}
|
||||
|
||||
av_log(log_ctx, AV_LOG_ERROR, "Could not find voice '%s'\n", voice_name);
|
||||
av_log(log_ctx, AV_LOG_INFO, "Choose between the voices: ");
|
||||
list_voices(log_ctx, ", ");
|
||||
|
||||
return AVERROR(EINVAL);
|
||||
}
|
||||
|
||||
static av_cold int init(AVFilterContext *ctx, const char *args)
|
||||
{
|
||||
FliteContext *flite = ctx->priv;
|
||||
int ret = 0;
|
||||
|
||||
flite->class = &flite_class;
|
||||
av_opt_set_defaults(flite);
|
||||
|
||||
if ((ret = av_set_options_string(flite, args, "=", ":")) < 0) {
|
||||
av_log(ctx, AV_LOG_ERROR, "Error parsing options string: '%s'\n", args);
|
||||
return ret;
|
||||
}
|
||||
|
||||
if (flite->list_voices) {
|
||||
list_voices(ctx, "\n");
|
||||
return AVERROR_EXIT;
|
||||
}
|
||||
|
||||
if (!flite_inited) {
|
||||
if (flite_init() < 0) {
|
||||
av_log(ctx, AV_LOG_ERROR, "flite initialization failed\n");
|
||||
return AVERROR_UNKNOWN;
|
||||
}
|
||||
flite_inited++;
|
||||
}
|
||||
|
||||
if ((ret = select_voice(&flite->voice, flite->voice_str, ctx)) < 0)
|
||||
return ret;
|
||||
|
||||
if (flite->textfile && flite->text) {
|
||||
av_log(ctx, AV_LOG_ERROR,
|
||||
"Both text and textfile options set: only one must be specified\n");
|
||||
return AVERROR(EINVAL);
|
||||
}
|
||||
|
||||
if (flite->textfile) {
|
||||
uint8_t *textbuf;
|
||||
size_t textbuf_size;
|
||||
|
||||
if ((ret = av_file_map(flite->textfile, &textbuf, &textbuf_size, 0, ctx)) < 0) {
|
||||
av_log(ctx, AV_LOG_ERROR,
|
||||
"The text file '%s' could not be read: %s\n",
|
||||
flite->textfile, av_err2str(ret));
|
||||
return ret;
|
||||
}
|
||||
|
||||
if (!(flite->text = av_malloc(textbuf_size+1)))
|
||||
return AVERROR(ENOMEM);
|
||||
memcpy(flite->text, textbuf, textbuf_size);
|
||||
flite->text[textbuf_size] = 0;
|
||||
av_file_unmap(textbuf, textbuf_size);
|
||||
}
|
||||
|
||||
if (!flite->text) {
|
||||
av_log(ctx, AV_LOG_ERROR,
|
||||
"No speech text specified, specify the 'text' or 'textfile' option\n");
|
||||
return AVERROR(EINVAL);
|
||||
}
|
||||
|
||||
/* synth all the file data in block */
|
||||
flite->wave = flite_text_to_wave(flite->text, flite->voice);
|
||||
flite->wave_samples = flite->wave->samples;
|
||||
flite->wave_nb_samples = flite->wave->num_samples;
|
||||
return 0;
|
||||
}
|
||||
|
||||
static av_cold void uninit(AVFilterContext *ctx)
|
||||
{
|
||||
FliteContext *flite = ctx->priv;
|
||||
|
||||
av_opt_free(flite);
|
||||
|
||||
delete_voice(flite->voice);
|
||||
flite->voice = NULL;
|
||||
delete_wave(flite->wave);
|
||||
flite->wave = NULL;
|
||||
}
|
||||
|
||||
static int query_formats(AVFilterContext *ctx)
|
||||
{
|
||||
FliteContext *flite = ctx->priv;
|
||||
|
||||
AVFilterChannelLayouts *chlayouts = NULL;
|
||||
int64_t chlayout = av_get_default_channel_layout(flite->wave->num_channels);
|
||||
AVFilterFormats *sample_formats = NULL;
|
||||
AVFilterFormats *sample_rates = NULL;
|
||||
|
||||
ff_add_channel_layout(&chlayouts, chlayout);
|
||||
ff_set_common_channel_layouts(ctx, chlayouts);
|
||||
ff_add_format(&sample_formats, AV_SAMPLE_FMT_S16);
|
||||
ff_set_common_formats(ctx, sample_formats);
|
||||
ff_add_format(&sample_rates, flite->wave->sample_rate);
|
||||
ff_set_common_samplerates (ctx, sample_rates);
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int config_props(AVFilterLink *outlink)
|
||||
{
|
||||
AVFilterContext *ctx = outlink->src;
|
||||
FliteContext *flite = ctx->priv;
|
||||
|
||||
outlink->sample_rate = flite->wave->sample_rate;
|
||||
outlink->time_base = (AVRational){1, flite->wave->sample_rate};
|
||||
|
||||
av_log(ctx, AV_LOG_VERBOSE, "voice:%s fmt:%s sample_rate:%d\n",
|
||||
flite->voice_str,
|
||||
av_get_sample_fmt_name(outlink->format), outlink->sample_rate);
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int request_frame(AVFilterLink *outlink)
|
||||
{
|
||||
AVFilterBufferRef *samplesref;
|
||||
FliteContext *flite = outlink->src->priv;
|
||||
int nb_samples = FFMIN(flite->wave_nb_samples, flite->frame_nb_samples);
|
||||
|
||||
if (!nb_samples)
|
||||
return AVERROR_EOF;
|
||||
|
||||
samplesref = ff_get_audio_buffer(outlink, AV_PERM_WRITE, nb_samples);
|
||||
if (!samplesref)
|
||||
return AVERROR(ENOMEM);
|
||||
|
||||
memcpy(samplesref->data[0], flite->wave_samples,
|
||||
nb_samples * flite->wave->num_channels * 2);
|
||||
samplesref->pts = flite->pts;
|
||||
samplesref->pos = -1;
|
||||
samplesref->audio->sample_rate = flite->wave->sample_rate;
|
||||
flite->pts += nb_samples;
|
||||
flite->wave_samples += nb_samples * flite->wave->num_channels;
|
||||
flite->wave_nb_samples -= nb_samples;
|
||||
|
||||
return ff_filter_samples(outlink, samplesref);
|
||||
}
|
||||
|
||||
AVFilter avfilter_asrc_flite = {
|
||||
.name = "flite",
|
||||
.description = NULL_IF_CONFIG_SMALL("Synthesize voice from text using libflite."),
|
||||
.query_formats = query_formats,
|
||||
.init = init,
|
||||
.uninit = uninit,
|
||||
.priv_size = sizeof(FliteContext),
|
||||
|
||||
.inputs = (const AVFilterPad[]) {{ .name = NULL}},
|
||||
|
||||
.outputs = (const AVFilterPad[]) {
|
||||
{
|
||||
.name = "default",
|
||||
.type = AVMEDIA_TYPE_AUDIO,
|
||||
.config_props = config_props,
|
||||
.request_frame = request_frame,
|
||||
},
|
||||
{ .name = NULL }
|
||||
},
|
||||
};
|
||||
|
|
@ -29,7 +29,7 @@
|
|||
#include "libavutil/avutil.h"
|
||||
|
||||
#define LIBAVFILTER_VERSION_MAJOR 3
|
||||
#define LIBAVFILTER_VERSION_MINOR 3
|
||||
#define LIBAVFILTER_VERSION_MINOR 4
|
||||
#define LIBAVFILTER_VERSION_MICRO 100
|
||||
|
||||
#define LIBAVFILTER_VERSION_INT AV_VERSION_INT(LIBAVFILTER_VERSION_MAJOR, \
|
||||
|
|
|
|||
Loading…
Reference in New Issue