|
|
|
@ -37,10 +37,6 @@ |
|
|
|
|
* The float version (default) can be downloaded from: |
|
|
|
|
* http://www.3gpp.org/ftp/Specs/archive/26_series/26.104/26104-610.zip
|
|
|
|
|
* |
|
|
|
|
* \subsection Fixed-point |
|
|
|
|
* The fixed-point (TS26.073) can be downloaded from: |
|
|
|
|
* http://www.3gpp.org/ftp/Specs/archive/26_series/26.073/26073-600.zip
|
|
|
|
|
* |
|
|
|
|
* \subsection Specification |
|
|
|
|
* The specification for AMR-NB can be found in TS 26.071 |
|
|
|
|
* (http://www.3gpp.org/ftp/Specs/html-info/26071.htm) and some other
|
|
|
|
@ -52,10 +48,6 @@ |
|
|
|
|
* The reference code can be downloaded from: |
|
|
|
|
* http://www.3gpp.org/ftp/Specs/archive/26_series/26.204/26204-600.zip
|
|
|
|
|
* |
|
|
|
|
* \subsection Fixed-point |
|
|
|
|
* If someone wants to use the fixed point version it can be downloaded from: |
|
|
|
|
* http://www.3gpp.org/ftp/Specs/archive/26_series/26.173/26173-571.zip.
|
|
|
|
|
* |
|
|
|
|
* \subsection Specification |
|
|
|
|
* The specification for AMR-WB can be found in TS 26.171 |
|
|
|
|
* (http://www.3gpp.org/ftp/Specs/html-info/26171.htm) and some other
|
|
|
|
@ -65,21 +57,8 @@ |
|
|
|
|
|
|
|
|
|
#include "avcodec.h" |
|
|
|
|
|
|
|
|
|
#if CONFIG_LIBAMR_NB_FIXED |
|
|
|
|
|
|
|
|
|
#define MMS_IO |
|
|
|
|
|
|
|
|
|
#include "amr/sp_dec.h" |
|
|
|
|
#include "amr/d_homing.h" |
|
|
|
|
#include "amr/typedef.h" |
|
|
|
|
#include "amr/sp_enc.h" |
|
|
|
|
#include "amr/sid_sync.h" |
|
|
|
|
#include "amr/e_homing.h" |
|
|
|
|
|
|
|
|
|
#else |
|
|
|
|
#include <amrnb/interf_dec.h> |
|
|
|
|
#include <amrnb/interf_enc.h> |
|
|
|
|
#endif |
|
|
|
|
|
|
|
|
|
static const char nb_bitrate_unsupported[] = |
|
|
|
|
"bitrate not supported: use one of 4.75k, 5.15k, 5.9k, 6.7k, 7.4k, 7.95k, 10.2k or 12.2k\n"; |
|
|
|
@ -137,225 +116,7 @@ static void amr_decode_fix_avctx(AVCodecContext * avctx) |
|
|
|
|
avctx->sample_fmt = SAMPLE_FMT_S16; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
#if CONFIG_LIBAMR_NB_FIXED |
|
|
|
|
/* fixed point version*/ |
|
|
|
|
/* frame size in serial bitstream file (frame type + serial stream + flags) */ |
|
|
|
|
#define SERIAL_FRAMESIZE (1+MAX_SERIAL_SIZE+5) |
|
|
|
|
|
|
|
|
|
typedef struct AMRContext { |
|
|
|
|
int frameCount; |
|
|
|
|
Speech_Decode_FrameState *speech_decoder_state; |
|
|
|
|
enum RXFrameType rx_type; |
|
|
|
|
enum Mode mode; |
|
|
|
|
Word16 reset_flag; |
|
|
|
|
Word16 reset_flag_old; |
|
|
|
|
|
|
|
|
|
int enc_bitrate; |
|
|
|
|
Speech_Encode_FrameState *enstate; |
|
|
|
|
sid_syncState *sidstate; |
|
|
|
|
enum TXFrameType tx_frametype; |
|
|
|
|
} AMRContext; |
|
|
|
|
|
|
|
|
|
static av_cold int amr_nb_decode_init(AVCodecContext * avctx) |
|
|
|
|
{ |
|
|
|
|
AMRContext *s = avctx->priv_data; |
|
|
|
|
|
|
|
|
|
s->frameCount=0; |
|
|
|
|
s->speech_decoder_state=NULL; |
|
|
|
|
s->rx_type = (enum RXFrameType)0; |
|
|
|
|
s->mode= (enum Mode)0; |
|
|
|
|
s->reset_flag=0; |
|
|
|
|
s->reset_flag_old=1; |
|
|
|
|
|
|
|
|
|
if(Speech_Decode_Frame_init(&s->speech_decoder_state, "Decoder")) |
|
|
|
|
{ |
|
|
|
|
av_log(avctx, AV_LOG_ERROR, "Speech_Decode_Frame_init error\n"); |
|
|
|
|
return -1; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
amr_decode_fix_avctx(avctx); |
|
|
|
|
|
|
|
|
|
if(avctx->channels > 1) |
|
|
|
|
{ |
|
|
|
|
av_log(avctx, AV_LOG_ERROR, "amr_nb: multichannel decoding not supported\n"); |
|
|
|
|
return -1; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
return 0; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static av_cold int amr_nb_encode_init(AVCodecContext * avctx) |
|
|
|
|
{ |
|
|
|
|
AMRContext *s = avctx->priv_data; |
|
|
|
|
|
|
|
|
|
s->frameCount=0; |
|
|
|
|
s->speech_decoder_state=NULL; |
|
|
|
|
s->rx_type = (enum RXFrameType)0; |
|
|
|
|
s->mode= (enum Mode)0; |
|
|
|
|
s->reset_flag=0; |
|
|
|
|
s->reset_flag_old=1; |
|
|
|
|
|
|
|
|
|
if(avctx->sample_rate!=8000) |
|
|
|
|
{ |
|
|
|
|
av_log(avctx, AV_LOG_ERROR, "Only 8000Hz sample rate supported\n"); |
|
|
|
|
return -1; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
if(avctx->channels!=1) |
|
|
|
|
{ |
|
|
|
|
av_log(avctx, AV_LOG_ERROR, "Only mono supported\n"); |
|
|
|
|
return -1; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
avctx->frame_size=160; |
|
|
|
|
avctx->coded_frame= avcodec_alloc_frame(); |
|
|
|
|
|
|
|
|
|
if(Speech_Encode_Frame_init(&s->enstate, 0, "encoder") || sid_sync_init (&s->sidstate)) |
|
|
|
|
{ |
|
|
|
|
av_log(avctx, AV_LOG_ERROR, "Speech_Encode_Frame_init error\n"); |
|
|
|
|
return -1; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
if((s->enc_bitrate=getBitrateMode(avctx->bit_rate))<0) |
|
|
|
|
{ |
|
|
|
|
av_log(avctx, AV_LOG_ERROR, nb_bitrate_unsupported); |
|
|
|
|
return -1; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
return 0; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static av_cold int amr_nb_encode_close(AVCodecContext * avctx) |
|
|
|
|
{ |
|
|
|
|
AMRContext *s = avctx->priv_data; |
|
|
|
|
|
|
|
|
|
Speech_Encode_Frame_exit(&s->enstate); |
|
|
|
|
sid_sync_exit (&s->sidstate); |
|
|
|
|
av_freep(&avctx->coded_frame); |
|
|
|
|
return 0; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static av_cold int amr_nb_decode_close(AVCodecContext * avctx) |
|
|
|
|
{ |
|
|
|
|
AMRContext *s = avctx->priv_data; |
|
|
|
|
|
|
|
|
|
Speech_Decode_Frame_exit(&s->speech_decoder_state); |
|
|
|
|
return 0; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static int amr_nb_decode_frame(AVCodecContext * avctx, |
|
|
|
|
void *data, int *data_size, |
|
|
|
|
AVPacket *avpkt) |
|
|
|
|
{ |
|
|
|
|
const uint8_t *buf = avpkt->data; |
|
|
|
|
int buf_size = avpkt->size; |
|
|
|
|
AMRContext *s = avctx->priv_data; |
|
|
|
|
const uint8_t*amrData=buf; |
|
|
|
|
int offset=0; |
|
|
|
|
UWord8 toc, q, ft; |
|
|
|
|
Word16 serial[SERIAL_FRAMESIZE]; /* coded bits */ |
|
|
|
|
Word16 *synth; |
|
|
|
|
UWord8 *packed_bits; |
|
|
|
|
static Word16 packed_size[16] = {12, 13, 15, 17, 19, 20, 26, 31, 5, 0, 0, 0, 0, 0, 0, 0}; |
|
|
|
|
int i; |
|
|
|
|
|
|
|
|
|
//printf("amr_decode_frame data_size=%i buf=0x%X buf_size=%d frameCount=%d!!\n",*data_size,buf,buf_size,s->frameCount);
|
|
|
|
|
|
|
|
|
|
synth=data; |
|
|
|
|
|
|
|
|
|
toc=amrData[offset]; |
|
|
|
|
/* read rest of the frame based on ToC byte */ |
|
|
|
|
q = (toc >> 2) & 0x01; |
|
|
|
|
ft = (toc >> 3) & 0x0F; |
|
|
|
|
|
|
|
|
|
//printf("offset=%d, packet_size=%d amrData= 0x%X %X %X %X\n",offset,packed_size[ft],amrData[offset],amrData[offset+1],amrData[offset+2],amrData[offset+3]);
|
|
|
|
|
|
|
|
|
|
offset++; |
|
|
|
|
|
|
|
|
|
packed_bits=amrData+offset; |
|
|
|
|
|
|
|
|
|
offset+=packed_size[ft]; |
|
|
|
|
|
|
|
|
|
//Unsort and unpack bits
|
|
|
|
|
s->rx_type = UnpackBits(q, ft, packed_bits, &s->mode, &serial[1]); |
|
|
|
|
|
|
|
|
|
//We have a new frame
|
|
|
|
|
s->frameCount++; |
|
|
|
|
|
|
|
|
|
if (s->rx_type == RX_NO_DATA) |
|
|
|
|
{ |
|
|
|
|
s->mode = s->speech_decoder_state->prev_mode; |
|
|
|
|
} |
|
|
|
|
else { |
|
|
|
|
s->speech_decoder_state->prev_mode = s->mode; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
/* if homed: check if this frame is another homing frame */ |
|
|
|
|
if (s->reset_flag_old == 1) |
|
|
|
|
{ |
|
|
|
|
/* only check until end of first subframe */ |
|
|
|
|
s->reset_flag = decoder_homing_frame_test_first(&serial[1], s->mode); |
|
|
|
|
} |
|
|
|
|
/* produce encoder homing frame if homed & input=decoder homing frame */ |
|
|
|
|
if ((s->reset_flag != 0) && (s->reset_flag_old != 0)) |
|
|
|
|
{ |
|
|
|
|
for (i = 0; i < L_FRAME; i++) |
|
|
|
|
{ |
|
|
|
|
synth[i] = EHF_MASK; |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
else |
|
|
|
|
{ |
|
|
|
|
/* decode frame */ |
|
|
|
|
Speech_Decode_Frame(s->speech_decoder_state, s->mode, &serial[1], s->rx_type, synth); |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
//Each AMR-frame results in 160 16-bit samples
|
|
|
|
|
*data_size=160*2; |
|
|
|
|
|
|
|
|
|
/* if not homed: check whether current frame is a homing frame */ |
|
|
|
|
if (s->reset_flag_old == 0) |
|
|
|
|
{ |
|
|
|
|
/* check whole frame */ |
|
|
|
|
s->reset_flag = decoder_homing_frame_test(&serial[1], s->mode); |
|
|
|
|
} |
|
|
|
|
/* reset decoder if current frame is a homing frame */ |
|
|
|
|
if (s->reset_flag != 0) |
|
|
|
|
{ |
|
|
|
|
Speech_Decode_Frame_reset(s->speech_decoder_state); |
|
|
|
|
} |
|
|
|
|
s->reset_flag_old = s->reset_flag; |
|
|
|
|
|
|
|
|
|
return offset; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
static int amr_nb_encode_frame(AVCodecContext *avctx, |
|
|
|
|
unsigned char *frame/*out*/, int buf_size, void *data/*in*/) |
|
|
|
|
{ |
|
|
|
|
short serial_data[250] = {0}; |
|
|
|
|
AMRContext *s = avctx->priv_data; |
|
|
|
|
int written; |
|
|
|
|
|
|
|
|
|
s->reset_flag = encoder_homing_frame_test(data); |
|
|
|
|
|
|
|
|
|
Speech_Encode_Frame(s->enstate, s->enc_bitrate, data, &serial_data[1], &s->mode); |
|
|
|
|
|
|
|
|
|
/* add frame type and mode */ |
|
|
|
|
sid_sync (s->sidstate, s->mode, &s->tx_frametype); |
|
|
|
|
|
|
|
|
|
written = PackBits(s->mode, s->enc_bitrate, s->tx_frametype, &serial_data[1], frame); |
|
|
|
|
|
|
|
|
|
if (s->reset_flag != 0) |
|
|
|
|
{ |
|
|
|
|
Speech_Encode_Frame_reset(s->enstate); |
|
|
|
|
sid_sync_reset(s->sidstate); |
|
|
|
|
} |
|
|
|
|
return written; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
#elif CONFIG_LIBAMR_NB /* Float point version*/ |
|
|
|
|
#if CONFIG_LIBAMR_NB |
|
|
|
|
|
|
|
|
|
typedef struct AMRContext { |
|
|
|
|
int frameCount; |
|
|
|
@ -496,7 +257,7 @@ static int amr_nb_encode_frame(AVCodecContext *avctx, |
|
|
|
|
|
|
|
|
|
#endif |
|
|
|
|
|
|
|
|
|
#if CONFIG_LIBAMR_NB || CONFIG_LIBAMR_NB_FIXED |
|
|
|
|
#if CONFIG_LIBAMR_NB |
|
|
|
|
|
|
|
|
|
AVCodec libamr_nb_decoder = |
|
|
|
|
{ |
|
|
|
|