FFmpeg
asrc_flite.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2012 Stefano Sabatini
3  *
4  * This file is part of FFmpeg.
5  *
6  * FFmpeg is free software; you can redistribute it and/or
7  * modify it under the terms of the GNU Lesser General Public
8  * License as published by the Free Software Foundation; either
9  * version 2.1 of the License, or (at your option) any later version.
10  *
11  * FFmpeg is distributed in the hope that it will be useful,
12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14  * Lesser General Public License for more details.
15  *
16  * You should have received a copy of the GNU Lesser General Public
17  * License along with FFmpeg; if not, write to the Free Software
18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19  */
20 
21 /**
22  * @file
23  * flite voice synth source
24  */
25 
26 #include <flite/flite.h>
28 #include "libavutil/file.h"
29 #include "libavutil/opt.h"
30 #include "avfilter.h"
31 #include "audio.h"
32 #include "formats.h"
33 #include "internal.h"
34 
35 typedef struct FliteContext {
36  const AVClass *class;
37  char *voice_str;
38  char *textfile;
39  char *text;
40  cst_wave *wave;
41  int16_t *wave_samples;
44  cst_voice *voice;
46  int64_t pts;
47  int frame_nb_samples; ///< number of samples per frame
48 } FliteContext;
49 
50 #define OFFSET(x) offsetof(FliteContext, x)
51 #define FLAGS AV_OPT_FLAG_AUDIO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
52 
53 static const AVOption flite_options[] = {
54  { "list_voices", "list voices and exit", OFFSET(list_voices), AV_OPT_TYPE_BOOL, {.i64=0}, 0, 1, FLAGS },
55  { "nb_samples", "set number of samples per frame", OFFSET(frame_nb_samples), AV_OPT_TYPE_INT, {.i64=512}, 0, INT_MAX, FLAGS },
56  { "n", "set number of samples per frame", OFFSET(frame_nb_samples), AV_OPT_TYPE_INT, {.i64=512}, 0, INT_MAX, FLAGS },
57  { "text", "set text to speak", OFFSET(text), AV_OPT_TYPE_STRING, {.str=NULL}, CHAR_MIN, CHAR_MAX, FLAGS },
58  { "textfile", "set filename of the text to speak", OFFSET(textfile), AV_OPT_TYPE_STRING, {.str=NULL}, CHAR_MIN, CHAR_MAX, FLAGS },
59  { "v", "set voice", OFFSET(voice_str), AV_OPT_TYPE_STRING, {.str="kal"}, CHAR_MIN, CHAR_MAX, FLAGS },
60  { "voice", "set voice", OFFSET(voice_str), AV_OPT_TYPE_STRING, {.str="kal"}, CHAR_MIN, CHAR_MAX, FLAGS },
61  { NULL }
62 };
63 
65 
66 static volatile int flite_inited = 0;
67 
68 /* declare functions for all the supported voices */
69 #define DECLARE_REGISTER_VOICE_FN(name) \
70  cst_voice *register_cmu_us_## name(const char *); \
71  void unregister_cmu_us_## name(cst_voice *);
77 
78 struct voice_entry {
79  const char *name;
80  cst_voice * (*register_fn)(const char *);
81  void (*unregister_fn)(cst_voice *);
82  cst_voice *voice;
83  unsigned usage_count;
84 } voice_entry;
85 
86 #define MAKE_VOICE_STRUCTURE(voice_name) { \
87  .name = #voice_name, \
88  .register_fn = register_cmu_us_ ## voice_name, \
89  .unregister_fn = unregister_cmu_us_ ## voice_name, \
90 }
91 static struct voice_entry voice_entries[] = {
94  MAKE_VOICE_STRUCTURE(kal16),
97 };
98 
99 static void list_voices(void *log_ctx, const char *sep)
100 {
101  int i, n = FF_ARRAY_ELEMS(voice_entries);
102  for (i = 0; i < n; i++)
103  av_log(log_ctx, AV_LOG_INFO, "%s%s",
104  voice_entries[i].name, i < (n-1) ? sep : "\n");
105 }
106 
107 static int select_voice(struct voice_entry **entry_ret, const char *voice_name, void *log_ctx)
108 {
109  int i;
110 
111  for (i = 0; i < FF_ARRAY_ELEMS(voice_entries); i++) {
112  struct voice_entry *entry = &voice_entries[i];
113  if (!strcmp(entry->name, voice_name)) {
114  if (!entry->voice)
115  entry->voice = entry->register_fn(NULL);
116  if (!entry->voice) {
117  av_log(log_ctx, AV_LOG_ERROR,
118  "Could not register voice '%s'\n", voice_name);
119  return AVERROR_UNKNOWN;
120  }
121  entry->usage_count++;
122  *entry_ret = entry;
123  return 0;
124  }
125  }
126 
127  av_log(log_ctx, AV_LOG_ERROR, "Could not find voice '%s'\n", voice_name);
128  av_log(log_ctx, AV_LOG_INFO, "Choose between the voices: ");
129  list_voices(log_ctx, ", ");
130 
131  return AVERROR(EINVAL);
132 }
133 
135 {
136  FliteContext *flite = ctx->priv;
137  int ret = 0;
138 
139  if (flite->list_voices) {
140  list_voices(ctx, "\n");
141  return AVERROR_EXIT;
142  }
143 
144  if (!flite_inited) {
145  if (flite_init() < 0) {
146  av_log(ctx, AV_LOG_ERROR, "flite initialization failed\n");
147  return AVERROR_UNKNOWN;
148  }
149  flite_inited++;
150  }
151 
152  if ((ret = select_voice(&flite->voice_entry, flite->voice_str, ctx)) < 0)
153  return ret;
154  flite->voice = flite->voice_entry->voice;
155 
156  if (flite->textfile && flite->text) {
157  av_log(ctx, AV_LOG_ERROR,
158  "Both text and textfile options set: only one must be specified\n");
159  return AVERROR(EINVAL);
160  }
161 
162  if (flite->textfile) {
163  uint8_t *textbuf;
164  size_t textbuf_size;
165 
166  if ((ret = av_file_map(flite->textfile, &textbuf, &textbuf_size, 0, ctx)) < 0) {
167  av_log(ctx, AV_LOG_ERROR,
168  "The text file '%s' could not be read: %s\n",
169  flite->textfile, av_err2str(ret));
170  return ret;
171  }
172 
173  if (!(flite->text = av_malloc(textbuf_size+1))) {
174  av_file_unmap(textbuf, textbuf_size);
175  return AVERROR(ENOMEM);
176  }
177  memcpy(flite->text, textbuf, textbuf_size);
178  flite->text[textbuf_size] = 0;
179  av_file_unmap(textbuf, textbuf_size);
180  }
181 
182  if (!flite->text) {
183  av_log(ctx, AV_LOG_ERROR,
184  "No speech text specified, specify the 'text' or 'textfile' option\n");
185  return AVERROR(EINVAL);
186  }
187 
188  /* synth all the file data in block */
189  flite->wave = flite_text_to_wave(flite->text, flite->voice);
190  flite->wave_samples = flite->wave->samples;
191  flite->wave_nb_samples = flite->wave->num_samples;
192  return 0;
193 }
194 
196 {
197  FliteContext *flite = ctx->priv;
198 
199  if (!--flite->voice_entry->usage_count)
200  flite->voice_entry->unregister_fn(flite->voice);
201  flite->voice = NULL;
202  flite->voice_entry = NULL;
203  delete_wave(flite->wave);
204  flite->wave = NULL;
205 }
206 
208 {
209  FliteContext *flite = ctx->priv;
210  int ret;
211 
212  AVFilterChannelLayouts *chlayouts = NULL;
213  int64_t chlayout = av_get_default_channel_layout(flite->wave->num_channels);
214  AVFilterFormats *sample_formats = NULL;
216 
217  if ((ret = ff_add_channel_layout (&chlayouts , chlayout )) < 0 ||
218  (ret = ff_set_common_channel_layouts (ctx , chlayouts )) < 0 ||
219  (ret = ff_add_format (&sample_formats, AV_SAMPLE_FMT_S16 )) < 0 ||
220  (ret = ff_set_common_formats (ctx , sample_formats )) < 0 ||
221  (ret = ff_add_format (&sample_rates , flite->wave->sample_rate)) < 0 ||
222  (ret = ff_set_common_samplerates (ctx , sample_rates )) < 0)
223  return ret;
224 
225  return 0;
226 }
227 
228 static int config_props(AVFilterLink *outlink)
229 {
230  AVFilterContext *ctx = outlink->src;
231  FliteContext *flite = ctx->priv;
232 
233  outlink->sample_rate = flite->wave->sample_rate;
234  outlink->time_base = (AVRational){1, flite->wave->sample_rate};
235 
236  av_log(ctx, AV_LOG_VERBOSE, "voice:%s fmt:%s sample_rate:%d\n",
237  flite->voice_str,
238  av_get_sample_fmt_name(outlink->format), outlink->sample_rate);
239  return 0;
240 }
241 
242 static int request_frame(AVFilterLink *outlink)
243 {
244  AVFrame *samplesref;
245  FliteContext *flite = outlink->src->priv;
246  int nb_samples = FFMIN(flite->wave_nb_samples, flite->frame_nb_samples);
247 
248  if (!nb_samples)
249  return AVERROR_EOF;
250 
251  samplesref = ff_get_audio_buffer(outlink, nb_samples);
252  if (!samplesref)
253  return AVERROR(ENOMEM);
254 
255  memcpy(samplesref->data[0], flite->wave_samples,
256  nb_samples * flite->wave->num_channels * 2);
257  samplesref->pts = flite->pts;
258  samplesref->pkt_pos = -1;
259  samplesref->sample_rate = flite->wave->sample_rate;
260  flite->pts += nb_samples;
261  flite->wave_samples += nb_samples * flite->wave->num_channels;
262  flite->wave_nb_samples -= nb_samples;
263 
264  return ff_filter_frame(outlink, samplesref);
265 }
266 
267 static const AVFilterPad flite_outputs[] = {
268  {
269  .name = "default",
270  .type = AVMEDIA_TYPE_AUDIO,
271  .config_props = config_props,
272  .request_frame = request_frame,
273  },
274  { NULL }
275 };
276 
278  .name = "flite",
279  .description = NULL_IF_CONFIG_SMALL("Synthesize voice from text using libflite."),
280  .query_formats = query_formats,
281  .init = init,
282  .uninit = uninit,
283  .priv_size = sizeof(FliteContext),
284  .inputs = NULL,
285  .outputs = flite_outputs,
286  .priv_class = &flite_class,
287 };
static const AVOption flite_options[]
Definition: asrc_flite.c:53
#define NULL
Definition: coverity.c:32
int ff_set_common_channel_layouts(AVFilterContext *ctx, AVFilterChannelLayouts *layouts)
A helper for query_formats() which sets all links to the same list of channel layouts/sample rates...
Definition: formats.c:550
This structure describes decoded (raw) audio or video data.
Definition: frame.h:295
#define FLAGS
Definition: asrc_flite.c:51
AVOption.
Definition: opt.h:246
int64_t pkt_pos
reordered pos from the last AVPacket that has been input into the decoder
Definition: frame.h:566
Main libavfilter public API header.
static av_cold void uninit(AVFilterContext *ctx)
Definition: asrc_flite.c:195
int wave_nb_samples
Definition: asrc_flite.c:42
AVFILTER_DEFINE_CLASS(flite)
static int request_frame(AVFilterLink *outlink)
Definition: asrc_flite.c:242
const char * name
Definition: asrc_flite.c:79
const char * name
Pad name.
Definition: internal.h:60
cst_wave * wave
Definition: asrc_flite.c:40
struct voice_entry * voice_entry
Definition: asrc_flite.c:45
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
Send a frame of data to the next filter.
Definition: avfilter.c:1093
uint8_t
#define av_cold
Definition: attributes.h:82
#define av_malloc(s)
int list_voices
Definition: asrc_flite.c:43
AVOptions.
static int query_formats(AVFilterContext *ctx)
Definition: asrc_flite.c:207
int64_t pts
Presentation timestamp in time_base units (time when frame should be shown to user).
Definition: frame.h:388
AVFilter ff_asrc_flite
Definition: asrc_flite.c:277
Misc file utilities.
#define AVERROR_EOF
End of file.
Definition: error.h:55
#define AV_LOG_VERBOSE
Detailed information.
Definition: log.h:192
#define OFFSET(x)
Definition: asrc_flite.c:50
char * text
Definition: asrc_flite.c:39
#define av_log(a,...)
unsigned usage_count
Definition: asrc_flite.c:83
int frame_nb_samples
number of samples per frame
Definition: asrc_flite.c:47
A filter pad used for either input or output.
Definition: internal.h:54
#define i(width, name, range_min, range_max)
Definition: cbs_h2645.c:259
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
Definition: log.h:176
void av_file_unmap(uint8_t *bufptr, size_t size)
Unmap or free the buffer bufptr created by av_file_map().
Definition: file.c:144
int ff_set_common_formats(AVFilterContext *ctx, AVFilterFormats *formats)
A helper for query_formats() which sets all links to the same list of formats.
Definition: formats.c:569
int av_file_map(const char *filename, uint8_t **bufptr, size_t *size, int log_offset, void *log_ctx)
Read the file with name filename, and put its content in a newly allocated buffer or map it with mmap...
Definition: file.c:53
int ff_add_channel_layout(AVFilterChannelLayouts **l, uint64_t channel_layout)
Definition: formats.c:342
AVFrame * ff_get_audio_buffer(AVFilterLink *link, int nb_samples)
Request an audio samples buffer with a specific set of permissions.
Definition: audio.c:86
#define NULL_IF_CONFIG_SMALL(x)
Return NULL if CONFIG_SMALL is true, otherwise the argument without modification. ...
Definition: internal.h:186
void * priv
private data for use by the filter
Definition: avfilter.h:353
static struct voice_entry voice_entries[]
Definition: asrc_flite.c:91
#define MAKE_VOICE_STRUCTURE(voice_name)
Definition: asrc_flite.c:86
int ff_add_format(AVFilterFormats **avff, int64_t fmt)
Add fmt to the list of media formats contained in *avff.
Definition: formats.c:336
const char * av_get_sample_fmt_name(enum AVSampleFormat sample_fmt)
Return the name of sample_fmt, or NULL if sample_fmt is not recognized.
Definition: samplefmt.c:49
Definition: asrc_flite.c:78
static int config_props(AVFilterLink *outlink)
Definition: asrc_flite.c:228
void(* unregister_fn)(cst_voice *)
Definition: asrc_flite.c:81
audio channel layout utility functions
#define FFMIN(a, b)
Definition: common.h:96
static volatile int flite_inited
Definition: asrc_flite.c:66
char * voice_str
Definition: asrc_flite.c:37
typedef void(APIENTRY *FF_PFNGLACTIVETEXTUREPROC)(GLenum texture)
#define av_err2str(errnum)
Convenience macro, the return value should be used only directly in function arguments but never stan...
Definition: error.h:119
AVFormatContext * ctx
Definition: movenc.c:48
cst_voice * voice
Definition: asrc_flite.c:44
static int select_voice(struct voice_entry **entry_ret, const char *voice_name, void *log_ctx)
Definition: asrc_flite.c:107
int n
Definition: avisynth_c.h:760
#define AVERROR_EXIT
Immediate exit was requested; the called function should not be restarted.
Definition: error.h:56
static const AVFilterPad outputs[]
Definition: af_acontrast.c:203
#define FF_ARRAY_ELEMS(a)
A list of supported channel layouts.
Definition: formats.h:85
int64_t pts
Definition: asrc_flite.c:46
#define AV_LOG_INFO
Standard information.
Definition: log.h:187
int16_t * wave_samples
Definition: asrc_flite.c:41
these buffered frames must be flushed immediately if a new input produces new the filter must not call request_frame to get more It must just process the frame or queue it The task of requesting more frames is left to the filter s request_frame method or the application If a filter has several inputs
Describe the class of an AVClass context structure.
Definition: log.h:67
int sample_rate
Sample rate of the audio data.
Definition: frame.h:467
Filter definition.
Definition: avfilter.h:144
Rational number (pair of numerator and denominator).
Definition: rational.h:58
cst_voice * voice
Definition: asrc_flite.c:82
const char * name
Filter name.
Definition: avfilter.h:148
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:309
static av_cold int init(AVFilterContext *ctx)
Definition: asrc_flite.c:134
sample_rates
signed 16 bits
Definition: samplefmt.h:61
#define AVERROR_UNKNOWN
Unknown error, typically from an external library.
Definition: error.h:71
A list of supported formats for one end of a filter link.
Definition: formats.h:64
An instance of a filter.
Definition: avfilter.h:338
int64_t av_get_default_channel_layout(int nb_channels)
Return default channel layout for a given number of channels.
#define DECLARE_REGISTER_VOICE_FN(name)
Definition: asrc_flite.c:69
char * textfile
Definition: asrc_flite.c:38
static const AVFilterPad flite_outputs[]
Definition: asrc_flite.c:267
internal API functions
Filter the word “frame” indicates either a video frame or a group of audio as stored in an AVFrame structure Format for each input and each output the list of supported formats For video that means pixel format For audio that means channel sample they are references to shared objects When the negotiation mechanism computes the intersection of the formats supported at each end of a all references to both lists are replaced with a reference to the intersection And when a single format is eventually chosen for a link amongst the remaining all references to the list are updated That means that if a filter requires that its input and output have the same format amongst a supported all it has to do is use a reference to the same list of formats query_formats can leave some formats unset and return AVERROR(EAGAIN) to cause the negotiation mechanism toagain later.That can be used by filters with complex requirements to use the format negotiated on one link to set the formats supported on another.Frame references ownership and permissions
int ff_set_common_samplerates(AVFilterContext *ctx, AVFilterFormats *samplerates)
Definition: formats.c:557
cst_voice *(* register_fn)(const char *)
Definition: asrc_flite.c:80
const char * name
Definition: opengl_enc.c:102