* Copyright (c) 2008-2010 Paul Kendall <paul@kcbbs.gen.nz>
* Copyright (c) 2010 Janne Grunau <janne-libav@jannau.net>
*
- * This file is part of Libav.
+ * This file is part of FFmpeg.
*
- * Libav is free software; you can redistribute it and/or
+ * 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.
*
- * Libav is distributed in the hope that it will be useful,
+ * 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 Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
* N SinuSoidal Coding (Transient, Sinusoid, Noise)
* Y Parametric Stereo
* N Direct Stream Transfer
+ * Y Enhanced AAC Low Delay (ER AAC ELD)
*
* Note: - HE AAC v1 comprises LC AAC with Spectral Band Replication.
* - HE AAC v2 comprises LC AAC with Spectral Band Replication and
*/
#include "libavutil/float_dsp.h"
+#include "libavutil/opt.h"
#include "avcodec.h"
#include "internal.h"
#include "get_bits.h"
#if ARCH_ARM
# include "arm/aac.h"
+#elif ARCH_MIPS
+# include "mips/aacdec_mips.h"
#endif
static VLC vlc_scalefactors;
static VLC vlc_spectral[11];
-static const char overread_err[] = "Input buffer exhausted before END element found\n";
+static int output_configure(AACContext *ac,
+ uint8_t layout_map[MAX_ELEM_ID*4][3], int tags,
+ enum OCStatus oc_type, int get_new_frame);
+
+#define overread_err "Input buffer exhausted before END element found\n"
static int count_channels(uint8_t (*layout)[3], int tags)
{
/**
* Check for the channel element in the current channel position configuration.
* If it exists, make sure the appropriate element is allocated and map the
- * channel order to match the internal Libav channel layout.
+ * channel order to match the internal FFmpeg channel layout.
*
* @param che_pos current channel position configuration
* @param type channel element type
ff_aac_sbr_ctx_init(ac, &ac->che[type][id]->sbr);
}
if (type != TYPE_CCE) {
+ if (*channels >= MAX_CHANNELS - (type == TYPE_CPE || (type == TYPE_SCE && ac->oc[1].m4ac.ps == 1))) {
+ av_log(ac->avctx, AV_LOG_ERROR, "Too many channels\n");
+ return AVERROR_INVALIDDATA;
+ }
ac->output_element[(*channels)++] = &ac->che[type][id]->ch[0];
if (type == TYPE_CPE ||
(type == TYPE_SCE && ac->oc[1].m4ac.ps == 1)) {
/* get output buffer */
av_frame_unref(ac->frame);
ac->frame->nb_samples = 2048;
- if ((ret = ff_get_buffer(avctx, ac->frame, 0)) < 0) {
- av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
+ if ((ret = ff_get_buffer(avctx, ac->frame, 0)) < 0)
return ret;
- }
/* map output channel pointers to AVFrame data */
for (ch = 0; ch < avctx->channels; ch++) {
ac->oc[1] = ac->oc[0];
ac->avctx->channels = ac->oc[1].channels;
ac->avctx->channel_layout = ac->oc[1].channel_layout;
+ output_configure(ac, ac->oc[1].layout_map, ac->oc[1].layout_map_tags,
+ ac->oc[1].status, 0);
}
}
}
memcpy(ac->tag_che_map, ac->che, 4 * MAX_ELEM_ID * sizeof(ac->che[0][0]));
- avctx->channel_layout = ac->oc[1].channel_layout = layout;
+ if (layout) avctx->channel_layout = layout;
+ ac->oc[1].channel_layout = layout;
avctx->channels = ac->oc[1].channels = channels;
ac->oc[1].status = oc_type;
return 0;
}
+static void flush(AVCodecContext *avctx)
+{
+ AACContext *ac= avctx->priv_data;
+ int type, i, j;
+
+ for (type = 3; type >= 0; type--) {
+ for (i = 0; i < MAX_ELEM_ID; i++) {
+ ChannelElement *che = ac->che[type][i];
+ if (che) {
+ for (j = 0; j <= 1; j++) {
+ memset(che->ch[j].saved, 0, sizeof(che->ch[j].saved));
+ }
+ }
+ }
+ }
+}
+
/**
* Set up channel positions based on a default channel configuration
* as specified in table 1.17.
int layout_map_tags;
push_output_configuration(ac);
+ av_log(ac->avctx, AV_LOG_DEBUG, "mono with CPE\n");
+
if (set_default_channel_config(ac->avctx, layout_map,
&layout_map_tags, 2) < 0)
return NULL;
int layout_map_tags;
push_output_configuration(ac);
+ av_log(ac->avctx, AV_LOG_DEBUG, "stereo with SCE\n");
+
if (set_default_channel_config(ac->avctx, layout_map,
&layout_map_tags, 1) < 0)
return NULL;
case AAC_CHANNEL_LFE:
syn_ele = TYPE_LFE;
break;
+ default:
+ av_assert0(0);
}
layout_map[0][0] = syn_ele;
layout_map[0][1] = get_bits(gb, 4);
if (get_bits1(gb))
skip_bits(gb, 3); // mixdown_coeff_index and pseudo_surround
+ if (get_bits_left(gb) < 4 * (num_front + num_side + num_back + num_lfe + num_assoc_data + num_cc)) {
+ av_log(avctx, AV_LOG_ERROR, "decode_pce: " overread_err);
+ return -1;
+ }
decode_channel_map(layout_map , AAC_CHANNEL_FRONT, gb, num_front);
tags = num_front;
decode_channel_map(layout_map + tags, AAC_CHANNEL_SIDE, gb, num_side);
/* comment field, first byte is length */
comment_len = get_bits(gb, 8) * 8;
if (get_bits_left(gb) < comment_len) {
- av_log(avctx, AV_LOG_ERROR, overread_err);
+ av_log(avctx, AV_LOG_ERROR, "decode_pce: " overread_err);
return AVERROR_INVALIDDATA;
}
skip_bits_long(gb, comment_len);
GetBitContext gb;
int i, ret;
- av_dlog(avctx, "extradata size %d\n", avctx->extradata_size);
- for (i = 0; i < avctx->extradata_size; i++)
- av_dlog(avctx, "%02x ", avctx->extradata[i]);
+ av_dlog(avctx, "audio specific config size %d\n", bit_size >> 3);
+ for (i = 0; i < bit_size >> 3; i++)
+ av_dlog(avctx, "%02x ", data[i]);
av_dlog(avctx, "\n");
if ((ret = init_get_bits(&gb, data, bit_size)) < 0)
*
* @return Returns a 32-bit pseudorandom integer
*/
-static av_always_inline int lcg_random(int previous_val)
+static av_always_inline int lcg_random(unsigned previous_val)
{
union { unsigned u; int s; } v = { previous_val * 1664525u + 1013904223 };
return v.s;
sizeof(ff_aac_spectral_codes[num][0]), \
size);
+static void aacdec_init(AACContext *ac);
+
static av_cold int aac_decode_init(AVCodecContext *avctx)
{
AACContext *ac = avctx->priv_data;
ac->avctx = avctx;
ac->oc[1].m4ac.sample_rate = avctx->sample_rate;
+ aacdec_init(ac);
+
avctx->sample_fmt = AV_SAMPLE_FMT_FLTP;
if (avctx->extradata_size > 0) {
}
}
+ if (avctx->channels > MAX_CHANNELS) {
+ av_log(avctx, AV_LOG_ERROR, "Too many channels\n");
+ return AVERROR_INVALIDDATA;
+ }
+
AAC_INIT_VLC_STATIC( 0, 304);
AAC_INIT_VLC_STATIC( 1, 270);
AAC_INIT_VLC_STATIC( 2, 550);
align_get_bits(gb);
if (get_bits_left(gb) < 8 * count) {
- av_log(ac->avctx, AV_LOG_ERROR, overread_err);
+ av_log(ac->avctx, AV_LOG_ERROR, "skip_data_stream_element: "overread_err);
return AVERROR_INVALIDDATA;
}
skip_bits_long(gb, 8 * count);
if (ics->predictor_present) {
if (aot == AOT_AAC_MAIN) {
if (decode_prediction(ac, ics, gb)) {
- return AVERROR_INVALIDDATA;
+ goto fail;
}
} else if (aot == AOT_AAC_LC ||
aot == AOT_ER_AAC_LC) {
av_log(ac->avctx, AV_LOG_ERROR,
"Prediction is not allowed in AAC-LC.\n");
- return AVERROR_INVALIDDATA;
+ goto fail;
} else {
if (aot == AOT_ER_AAC_LD) {
av_log(ac->avctx, AV_LOG_ERROR,
"Number of scalefactor bands in group (%d) "
"exceeds limit (%d).\n",
ics->max_sfb, ics->num_swb);
- return AVERROR_INVALIDDATA;
+ goto fail;
}
return 0;
+fail:
+ ics->max_sfb = 0;
+ return AVERROR_INVALIDDATA;
}
/**
sect_len_incr = get_bits(gb, bits);
sect_end += sect_len_incr;
if (get_bits_left(gb) < 0) {
- av_log(ac->avctx, AV_LOG_ERROR, overread_err);
+ av_log(ac->avctx, AV_LOG_ERROR, "decode_band_types: "overread_err);
return AVERROR_INVALIDDATA;
}
if (sect_end > ics->max_sfb) {
idx++)
cpe->ms_mask[idx] = get_bits1(gb);
} else if (ms_present == 2) {
- memset(cpe->ms_mask, 1, cpe->ch[0].ics.num_window_groups * cpe->ch[0].ics.max_sfb * sizeof(cpe->ms_mask[0]));
+ memset(cpe->ms_mask, 1, sizeof(cpe->ms_mask[0]) * cpe->ch[0].ics.num_window_groups * cpe->ch[0].ics.max_sfb);
}
}
return n;
}
+static int decode_fill(AACContext *ac, GetBitContext *gb, int len) {
+ uint8_t buf[256];
+ int i, major, minor;
+
+ if (len < 13+7*8)
+ goto unknown;
+
+ get_bits(gb, 13); len -= 13;
+
+ for(i=0; i+1<sizeof(buf) && len>=8; i++, len-=8)
+ buf[i] = get_bits(gb, 8);
+
+ buf[i] = 0;
+ if (ac->avctx->debug & FF_DEBUG_PICT_INFO)
+ av_log(ac->avctx, AV_LOG_DEBUG, "FILL:%s\n", buf);
+
+ if (sscanf(buf, "libfaac %d.%d", &major, &minor) == 2){
+ ac->avctx->internal->skip_samples = 1024;
+ }
+
+unknown:
+ skip_bits_long(gb, len);
+
+ return 0;
+}
+
/**
* Decode extension data (incomplete); reference: table 4.51.
*
} else if (ac->oc[1].m4ac.ps == -1 && ac->oc[1].status < OC_LOCKED && ac->avctx->channels == 1) {
ac->oc[1].m4ac.sbr = 1;
ac->oc[1].m4ac.ps = 1;
+ ac->avctx->profile = FF_PROFILE_AAC_HE_V2;
output_configure(ac, ac->oc[1].layout_map, ac->oc[1].layout_map_tags,
ac->oc[1].status, 1);
} else {
ac->oc[1].m4ac.sbr = 1;
+ ac->avctx->profile = FF_PROFILE_AAC_HE;
}
res = ff_decode_sbr_extension(ac, &che->sbr, gb, crc_flag, cnt, elem_type);
break;
res = decode_dynamic_range(&ac->che_drc, gb);
break;
case EXT_FILL:
+ decode_fill(ac, gb, 8 * cnt - 4);
+ break;
case EXT_FILL_DATA:
case EXT_DATA_ELEMENT:
default:
int w, filt, m, i;
int bottom, top, order, start, end, size, inc;
float lpc[TNS_MAX_ORDER];
- float tmp[TNS_MAX_ORDER + 1];
+ float tmp[TNS_MAX_ORDER+1];
for (w = 0; w < ics->num_windows; w++) {
bottom = ics->num_swb;
predTime[i] = sce->ltp_state[i + 2048 - ltp->lag] * ltp->coef;
memset(&predTime[i], 0, (2048 - i) * sizeof(float));
- windowing_and_mdct_ltp(ac, predFreq, predTime, &sce->ics);
+ ac->windowing_and_mdct_ltp(ac, predFreq, predTime, &sce->ics);
if (sce->tns.present)
- apply_tns(predFreq, &sce->tns, &sce->ics, 0);
+ ac->apply_tns(predFreq, &sce->tns, &sce->ics, 0);
for (sfb = 0; sfb < FFMIN(sce->ics.max_sfb, MAX_LTP_LONG_SFB); sfb++)
if (ltp->used[sfb])
// Inverse transform, mapped to the conventional IMDCT by
// Chivukula, R.K.; Reznik, Y.A.; Devarajan, V.,
// "Efficient algorithms for MPEG-4 AAC-ELD, AAC-LD and AAC-LC filterbanks,"
- // Audio, Language and Image Processing, 2008. ICALIP 2008. International Conference on
+ // International Conference on Audio, Language and Image Processing, ICALIP 2008.
// URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=4590245&isnumber=4589950
for (i = 0; i < n2; i+=2) {
float temp;
imdct_and_window = imdct_and_windowing_eld;
break;
default:
- imdct_and_window = imdct_and_windowing;
+ imdct_and_window = ac->imdct_and_windowing;
}
for (type = 3; type >= 0; type--) {
for (i = 0; i < MAX_ELEM_ID; i++) {
if (ac->oc[1].m4ac.object_type == AOT_AAC_LTP) {
if (che->ch[0].ics.predictor_present) {
if (che->ch[0].ics.ltp.present)
- apply_ltp(ac, &che->ch[0]);
+ ac->apply_ltp(ac, &che->ch[0]);
if (che->ch[1].ics.ltp.present && type == TYPE_CPE)
- apply_ltp(ac, &che->ch[1]);
+ ac->apply_ltp(ac, &che->ch[1]);
}
}
if (che->ch[0].tns.present)
- apply_tns(che->ch[0].coeffs, &che->ch[0].tns, &che->ch[0].ics, 1);
+ ac->apply_tns(che->ch[0].coeffs, &che->ch[0].tns, &che->ch[0].ics, 1);
if (che->ch[1].tns.present)
- apply_tns(che->ch[1].coeffs, &che->ch[1].tns, &che->ch[1].ics, 1);
+ ac->apply_tns(che->ch[1].coeffs, &che->ch[1].tns, &che->ch[1].ics, 1);
if (type <= TYPE_CPE)
apply_channel_coupling(ac, che, type, i, BETWEEN_TNS_AND_IMDCT, apply_dependent_coupling);
if (type != TYPE_CCE || che->coup.coupling_point == AFTER_IMDCT) {
imdct_and_window(ac, &che->ch[0]);
if (ac->oc[1].m4ac.object_type == AOT_AAC_LTP)
- update_ltp(ac, &che->ch[0]);
+ ac->update_ltp(ac, &che->ch[0]);
if (type == TYPE_CPE) {
imdct_and_window(ac, &che->ch[1]);
if (ac->oc[1].m4ac.object_type == AOT_AAC_LTP)
- update_ltp(ac, &che->ch[1]);
+ ac->update_ltp(ac, &che->ch[1]);
}
if (ac->oc[1].m4ac.sbr > 0) {
ff_sbr_apply(ac, &che->sbr, type, che->ch[0].ret, che->ch[1].ret);
size = avpriv_aac_parse_header(gb, &hdr_info);
if (size > 0) {
- if (hdr_info.num_aac_frames != 1) {
+ if (!ac->warned_num_aac_frames && hdr_info.num_aac_frames != 1) {
+ // This is 2 for "VLB " audio in NSV files.
+ // See samples/nsv/vlb_audio.
avpriv_report_missing_feature(ac->avctx,
"More than one AAC RDB per ADTS frame");
- return AVERROR_PATCHWELCOME;
+ ac->warned_num_aac_frames = 1;
}
push_output_configuration(ac);
if (hdr_info.chan_config) {
return ret;
} else {
ac->oc[1].m4ac.chan_config = 0;
+ /**
+ * dual mono frames in Japanese DTV can have chan_config 0
+ * WITHOUT specifying PCE.
+ * thus, set dual mono as default.
+ */
+ if (ac->dmono_mode && ac->oc[0].status == OC_NONE) {
+ layout_map_tags = 2;
+ layout_map[0][0] = layout_map[1][0] = TYPE_SCE;
+ layout_map[0][2] = layout_map[1][2] = AAC_CHANNEL_FRONT;
+ layout_map[0][1] = 0;
+ layout_map[1][1] = 1;
+ if (output_configure(ac, layout_map, layout_map_tags,
+ OC_TRIAL_FRAME, 0))
+ return -7;
+ }
}
ac->oc[1].m4ac.sample_rate = hdr_info.sample_rate;
ac->oc[1].m4ac.sampling_index = hdr_info.sampling_index;
if ((err = frame_configure_elements(avctx)) < 0)
return err;
+ // The FF_PROFILE_AAC_* defines are all object_type - 1
+ // This may lead to an undefined profile being signaled
+ ac->avctx->profile = ac->oc[1].m4ac.object_type - 1;
+
ac->tags_mapped = 0;
if (chan_config < 0 || chan_config >= 8) {
}
static int aac_decode_frame_int(AVCodecContext *avctx, void *data,
- int *got_frame_ptr, GetBitContext *gb)
+ int *got_frame_ptr, GetBitContext *gb, AVPacket *avpkt)
{
AACContext *ac = avctx->priv_data;
ChannelElement *che = NULL, *che_prev = NULL;
enum RawDataBlockType elem_type, elem_type_prev = TYPE_END;
int err, elem_id;
int samples = 0, multiplier, audio_found = 0, pce_found = 0;
+ int is_dmono, sce_count = 0;
ac->frame = data;
if ((err = frame_configure_elements(avctx)) < 0)
goto fail;
+ // The FF_PROFILE_AAC_* defines are all object_type - 1
+ // This may lead to an undefined profile being signaled
+ ac->avctx->profile = ac->oc[1].m4ac.object_type - 1;
+
ac->tags_mapped = 0;
// parse
while ((elem_type = get_bits(gb, 3)) != TYPE_END) {
case TYPE_SCE:
err = decode_ics(ac, &che->ch[0], gb, 0, 0);
audio_found = 1;
+ sce_count++;
break;
case TYPE_CPE:
if (pce_found) {
av_log(avctx, AV_LOG_ERROR,
"Not evaluating a further program_config_element as this construct is dubious at best.\n");
- pop_output_configuration(ac);
} else {
err = output_configure(ac, layout_map, tags, OC_TRIAL_PCE, 1);
+ if (!err)
+ ac->oc[1].m4ac.chan_config = 0;
pce_found = 1;
}
break;
if (elem_id == 15)
elem_id += get_bits(gb, 8) - 1;
if (get_bits_left(gb) < 8 * elem_id) {
- av_log(avctx, AV_LOG_ERROR, overread_err);
+ av_log(avctx, AV_LOG_ERROR, "TYPE_FIL: "overread_err);
err = AVERROR_INVALIDDATA;
goto fail;
}
multiplier = (ac->oc[1].m4ac.sbr == 1) ? ac->oc[1].m4ac.ext_sample_rate > ac->oc[1].m4ac.sample_rate : 0;
samples <<= multiplier;
+ /* for dual-mono audio (SCE + SCE) */
+ is_dmono = ac->dmono_mode && sce_count == 2 &&
+ ac->oc[1].channel_layout == (AV_CH_FRONT_LEFT | AV_CH_FRONT_RIGHT);
if (samples)
ac->frame->nb_samples = samples;
+ else
+ av_frame_unref(ac->frame);
*got_frame_ptr = !!samples;
+ if (is_dmono) {
+ if (ac->dmono_mode == 1)
+ ((AVFrame *)data)->data[1] =((AVFrame *)data)->data[0];
+ else if (ac->dmono_mode == 2)
+ ((AVFrame *)data)->data[0] =((AVFrame *)data)->data[1];
+ }
+
if (ac->oc[1].status && audio_found) {
avctx->sample_rate = ac->oc[1].m4ac.sample_rate << multiplier;
avctx->frame_size = samples;
ac->oc[1].status = OC_LOCKED;
}
+ if (multiplier) {
+ int side_size;
+ const uint8_t *side = av_packet_get_side_data(avpkt, AV_PKT_DATA_SKIP_SAMPLES, &side_size);
+ if (side && side_size>=4)
+ AV_WL32(side, 2*AV_RL32(side));
+ }
return 0;
fail:
pop_output_configuration(ac);
const uint8_t *new_extradata = av_packet_get_side_data(avpkt,
AV_PKT_DATA_NEW_EXTRADATA,
&new_extradata_size);
+ int jp_dualmono_size;
+ const uint8_t *jp_dualmono = av_packet_get_side_data(avpkt,
+ AV_PKT_DATA_JP_DUALMONO,
+ &jp_dualmono_size);
- if (new_extradata) {
+ if (new_extradata && 0) {
av_free(avctx->extradata);
avctx->extradata = av_mallocz(new_extradata_size +
FF_INPUT_BUFFER_PADDING_SIZE);
}
}
+ ac->dmono_mode = 0;
+ if (jp_dualmono && jp_dualmono_size > 0)
+ ac->dmono_mode = 1 + *jp_dualmono;
+ if (ac->force_dmono_mode >= 0)
+ ac->dmono_mode = ac->force_dmono_mode;
+
+ if (INT_MAX / 8 <= buf_size)
+ return AVERROR_INVALIDDATA;
+
if ((err = init_get_bits(&gb, buf, buf_size * 8)) < 0)
return err;
err = aac_decode_er_frame(avctx, data, got_frame_ptr, &gb);
break;
default:
- err = aac_decode_frame_int(avctx, data, got_frame_ptr, &gb);
+ err = aac_decode_frame_int(avctx, data, got_frame_ptr, &gb, avpkt);
}
if (err < 0)
return err;
struct LATMContext {
AACContext aac_ctx; ///< containing AACContext
- int initialized; ///< initilized after a valid extradata was seen
+ int initialized; ///< initialized after a valid extradata was seen
// parser data
int audio_mux_version_A; ///< LATM syntax version
if (bits_consumed < 0)
return AVERROR_INVALIDDATA;
- if (ac->oc[1].m4ac.sample_rate != m4ac.sample_rate ||
+ if (!latmctx->initialized ||
+ ac->oc[1].m4ac.sample_rate != m4ac.sample_rate ||
ac->oc[1].m4ac.chan_config != m4ac.chan_config) {
- av_log(avctx, AV_LOG_INFO, "audio config changed\n");
+ if(latmctx->initialized) {
+ av_log(avctx, AV_LOG_INFO, "audio config changed\n");
+ } else {
+ av_log(avctx, AV_LOG_DEBUG, "initializing latmctx\n");
+ }
latmctx->initialized = 0;
esize = (bits_consumed+7) / 8;
return AVERROR_PATCHWELCOME;
}
- // for each program (which there is only on in DVB)
+ // for each program (which there is only one in DVB)
- // for each layer (which there is only on in DVB)
+ // for each layer (which there is only one in DVB)
if (get_bits(gb, 3)) { // numLayer
avpriv_request_sample(latmctx->aac_ctx.avctx, "Multiple layers");
return AVERROR_PATCHWELCOME;
int muxlength, err;
GetBitContext gb;
- if ((err = init_get_bits(&gb, avpkt->data, avpkt->size * 8)) < 0)
+ if ((err = init_get_bits8(&gb, avpkt->data, avpkt->size)) < 0)
return err;
// check for LOAS sync word
return AVERROR_INVALIDDATA;
muxlength = get_bits(&gb, 13) + 3;
- // not enough data, the parser should have sorted this
+ // not enough data, the parser should have sorted this out
if (muxlength > avpkt->size)
return AVERROR_INVALIDDATA;
return AVERROR_INVALIDDATA;
}
- if ((err = aac_decode_frame_int(avctx, out, got_frame_ptr, &gb)) < 0)
+ if ((err = aac_decode_frame_int(avctx, out, got_frame_ptr, &gb, avpkt)) < 0)
return err;
return muxlength;
return ret;
}
+static void aacdec_init(AACContext *c)
+{
+ c->imdct_and_windowing = imdct_and_windowing;
+ c->apply_ltp = apply_ltp;
+ c->apply_tns = apply_tns;
+ c->windowing_and_mdct_ltp = windowing_and_mdct_ltp;
+ c->update_ltp = update_ltp;
+
+ if(ARCH_MIPS)
+ ff_aacdec_init_mips(c);
+}
+/**
+ * AVOptions for Japanese DTV specific extensions (ADTS only)
+ */
+#define AACDEC_FLAGS AV_OPT_FLAG_DECODING_PARAM | AV_OPT_FLAG_AUDIO_PARAM
+static const AVOption options[] = {
+ {"dual_mono_mode", "Select the channel to decode for dual mono",
+ offsetof(AACContext, force_dmono_mode), AV_OPT_TYPE_INT, {.i64=-1}, -1, 2,
+ AACDEC_FLAGS, "dual_mono_mode"},
+
+ {"auto", "autoselection", 0, AV_OPT_TYPE_CONST, {.i64=-1}, INT_MIN, INT_MAX, AACDEC_FLAGS, "dual_mono_mode"},
+ {"main", "Select Main/Left channel", 0, AV_OPT_TYPE_CONST, {.i64= 1}, INT_MIN, INT_MAX, AACDEC_FLAGS, "dual_mono_mode"},
+ {"sub" , "Select Sub/Right channel", 0, AV_OPT_TYPE_CONST, {.i64= 2}, INT_MIN, INT_MAX, AACDEC_FLAGS, "dual_mono_mode"},
+ {"both", "Select both channels", 0, AV_OPT_TYPE_CONST, {.i64= 0}, INT_MIN, INT_MAX, AACDEC_FLAGS, "dual_mono_mode"},
+
+ {NULL},
+};
+
+static const AVClass aac_decoder_class = {
+ .class_name = "AAC decoder",
+ .item_name = av_default_item_name,
+ .option = options,
+ .version = LIBAVUTIL_VERSION_INT,
+};
AVCodec ff_aac_decoder = {
.name = "aac",
},
.capabilities = CODEC_CAP_CHANNEL_CONF | CODEC_CAP_DR1,
.channel_layouts = aac_channel_layout,
+ .flush = flush,
+ .priv_class = &aac_decoder_class,
};
/*
},
.capabilities = CODEC_CAP_CHANNEL_CONF | CODEC_CAP_DR1,
.channel_layouts = aac_channel_layout,
+ .flush = flush,
};
* Copyright (c) 2008-2009 Robert Swain ( rob opendot cl )
* Copyright (c) 2009-2010 Alex Converse <alex.converse@gmail.com>
*
- * This file is part of Libav.
+ * This file is part of FFmpeg.
*
- * Libav is free software; you can redistribute it and/or
+ * 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.
*
- * Libav is distributed in the hope that it will be useful,
+ * 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 Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "sbrdsp.h"
#include "libavutil/internal.h"
#include "libavutil/libm.h"
+#include "libavutil/avassert.h"
#include <stdint.h>
#include <float.h>
+#include <math.h>
#define ENVELOPE_ADJUSTMENT_OFFSET 2
#define NOISE_FLOOR_OFFSET 6.0f
+#if ARCH_MIPS
+#include "mips/aacsbr_mips.h"
+#endif /* ARCH_MIPS */
+
/**
* SBR VLC tables
*/
#define SBR_VLC_ROW(name) \
{ name ## _codes, name ## _bits, sizeof(name ## _codes), sizeof(name ## _codes[0]) }
+static void aacsbr_func_ptr_init(AACSBRContext *c);
+
av_cold void ff_aac_sbr_init(void)
{
int n;
av_cold void ff_aac_sbr_ctx_init(AACContext *ac, SpectralBandReplication *sbr)
{
+ if(sbr->mdct.mdct_bits)
+ return;
sbr->kx[0] = sbr->kx[1];
sbr_turnoff(sbr);
sbr->data[0].synthesis_filterbank_samples_offset = SBR_SYNTHESIS_BUF_SIZE - (1280 - 128);
ff_mdct_init(&sbr->mdct_ana, 7, 1, -2.0 * 32768.0);
ff_ps_ctx_init(&sbr->ps);
ff_sbrdsp_init(&sbr->dsp);
+ aacsbr_func_ptr_init(&sbr->c);
}
av_cold void ff_aac_sbr_ctx_close(SpectralBandReplication *sbr)
} else
temp = 5000;
- start_min = ((temp << 7) + (sbr->sample_rate >> 1)) / sbr->sample_rate;
- stop_min = ((temp << 8) + (sbr->sample_rate >> 1)) / sbr->sample_rate;
-
switch (sbr->sample_rate) {
case 16000:
sbr_offset_ptr = sbr_offset[0];
return -1;
}
+ start_min = ((temp << 7) + (sbr->sample_rate >> 1)) / sbr->sample_rate;
+ stop_min = ((temp << 8) + (sbr->sample_rate >> 1)) / sbr->sample_rate;
+
sbr->k[0] = start_min + sbr_offset_ptr[spectrum->bs_start_freq];
if (spectrum->bs_stop_freq < 14) {
max_qmf_subbands = 35;
} else if (sbr->sample_rate >= 48000)
max_qmf_subbands = 32;
+ else
+ av_assert0(0);
if (sbr->k[2] - sbr->k[0] > max_qmf_subbands) {
av_log(ac->avctx, AV_LOG_ERROR,
k = sbr->n_master;
} while (sb != sbr->kx[1] + sbr->m[1]);
- if (sbr->patch_num_subbands[sbr->num_patches-1] < 3 && sbr->num_patches > 1)
+ if (sbr->num_patches > 1 && sbr->patch_num_subbands[sbr->num_patches-1] < 3)
sbr->num_patches--;
return 0;
if (ch_data->bs_frame_class == FIXFIX) {
idx = ch_data->bs_num_env >> 1;
} else if (ch_data->bs_frame_class & 1) { // FIXVAR or VARVAR
- idx = ch_data->bs_num_env - FFMAX(bs_pointer - 1, 1);
+ idx = ch_data->bs_num_env - FFMAX((int)bs_pointer - 1, 1);
} else { // VARFIX
if (!bs_pointer)
idx = 1;
} else {
#if 1
*num_bits_left -= ff_ps_read_data(ac->avctx, gb, &sbr->ps, *num_bits_left);
+ ac->avctx->profile = FF_PROFILE_AAC_HE_V2;
#else
avpriv_report_missing_feature(ac->avctx, "Parametric Stereo");
skip_bits_long(gb, *num_bits_left); // bs_fill_bits
}
break;
default:
- avpriv_request_sample(ac->avctx, "Reserved SBR extensions");
+ // some files contain 0-padding
+ if (bs_extension_id || *num_bits_left > 16 || show_bits(gb, *num_bits_left))
+ avpriv_request_sample(ac->avctx, "Reserved SBR extensions");
skip_bits_long(gb, *num_bits_left); // bs_fill_bits
*num_bits_left = 0;
break;
for (k = 0; k < sbr->n[sbr->data[0].bs_freq_res[e]]; k++) {
float temp1 = exp2f(sbr->data[0].env_facs[e][k] * alpha + 7.0f);
float temp2 = exp2f((pan_offset - sbr->data[1].env_facs[e][k]) * alpha);
- float fac = temp1 / (1.0f + temp2);
+ float fac;
+ if (temp1 > 1E20) {
+ av_log(NULL, AV_LOG_ERROR, "envelope scalefactor overflow in dequant\n");
+ temp1 = 1;
+ }
+ fac = temp1 / (1.0f + temp2);
sbr->data[0].env_facs[e][k] = fac;
sbr->data[1].env_facs[e][k] = fac * temp2;
}
for (k = 0; k < sbr->n_q; k++) {
float temp1 = exp2f(NOISE_FLOOR_OFFSET - sbr->data[0].noise_facs[e][k] + 1);
float temp2 = exp2f(12 - sbr->data[1].noise_facs[e][k]);
- float fac = temp1 / (1.0f + temp2);
+ float fac;
+ if (temp1 > 1E20) {
+ av_log(NULL, AV_LOG_ERROR, "envelope scalefactor overflow in dequant\n");
+ temp1 = 1;
+ }
+ fac = temp1 / (1.0f + temp2);
sbr->data[0].noise_facs[e][k] = fac;
sbr->data[1].noise_facs[e][k] = fac * temp2;
}
for (ch = 0; ch < (id_aac == TYPE_CPE) + 1; ch++) {
float alpha = sbr->data[ch].bs_amp_res ? 1.0f : 0.5f;
for (e = 1; e <= sbr->data[ch].bs_num_env; e++)
- for (k = 0; k < sbr->n[sbr->data[ch].bs_freq_res[e]]; k++)
+ for (k = 0; k < sbr->n[sbr->data[ch].bs_freq_res[e]]; k++){
sbr->data[ch].env_facs[e][k] =
exp2f(alpha * sbr->data[ch].env_facs[e][k] + 6.0f);
+ if (sbr->data[ch].env_facs[e][k] > 1E20) {
+ av_log(NULL, AV_LOG_ERROR, "envelope scalefactor overflow in dequant\n");
+ sbr->data[ch].env_facs[e][k] = 1;
+ }
+ }
+
for (e = 1; e <= sbr->data[ch].bs_num_noise; e++)
for (k = 0; k < sbr->n_q; k++)
sbr->data[ch].noise_facs[e][k] =
* @param x pointer to the beginning of the first sample window
* @param W array of complex-valued samples split into subbands
*/
+#ifndef sbr_qmf_analysis
static void sbr_qmf_analysis(AVFloatDSPContext *dsp, FFTContext *mdct,
SBRDSPContext *sbrdsp, const float *in, float *x,
float z[320], float W[2][32][32][2], int buf_idx)
x += 32;
}
}
+#endif
/**
* Synthesis QMF Bank (14496-3 sp04 p206) and Downsampled Synthesis QMF Bank
* (14496-3 sp04 p206)
*/
+#ifndef sbr_qmf_synthesis
static void sbr_qmf_synthesis(FFTContext *mdct,
SBRDSPContext *sbrdsp, AVFloatDSPContext *dsp,
float *out, float X[2][38][64],
out += 64 >> div;
}
}
+#endif
/** High Frequency Generation (14496-3 sp04 p214+) and Inverse Filtering
* (14496-3 sp04 p214)
0.11516383427084,
0.03183050093751,
};
- static const int8_t phi[2][4] = {
- { 1, 0, -1, 0}, // real
- { 0, 1, 0, -1}, // imaginary
- };
float (*g_temp)[48] = ch_data->g_temp, (*q_temp)[48] = ch_data->q_temp;
int indexnoise = ch_data->f_indexnoise;
int indexsine = ch_data->f_indexsine;
for (e = 0; e < ch_data->bs_num_env; e++) {
for (i = 2 * ch_data->t_env[e]; i < 2 * ch_data->t_env[e + 1]; i++) {
- int phi_sign = (1 - 2*(kx & 1));
LOCAL_ALIGNED_16(float, g_filt_tab, [48]);
LOCAL_ALIGNED_16(float, q_filt_tab, [48]);
float *g_filt, *q_filt;
q_filt, indexnoise,
kx, m_max);
} else {
- for (m = 0; m < m_max; m++) {
- Y1[i][m + kx][0] +=
- sbr->s_m[e][m] * phi[0][indexsine];
- Y1[i][m + kx][1] +=
- sbr->s_m[e][m] * (phi[1][indexsine] * phi_sign);
- phi_sign = -phi_sign;
+ int idx = indexsine&1;
+ int A = (1-((indexsine+(kx & 1))&2));
+ int B = (A^(-idx)) + idx;
+ float *out = &Y1[i][kx][idx];
+ float *in = sbr->s_m[e];
+ for (m = 0; m+1 < m_max; m+=2) {
+ out[2*m ] += in[m ] * A;
+ out[2*m+2] += in[m+1] * B;
}
+ if(m_max&1)
+ out[2*m ] += in[m ] * A;
}
indexnoise = (indexnoise + m_max) & 0x1ff;
indexsine = (indexsine + 1) & 3;
sbr_qmf_analysis(&ac->fdsp, &sbr->mdct_ana, &sbr->dsp, ch ? R : L, sbr->data[ch].analysis_filterbank_samples,
(float*)sbr->qmf_filter_scratch,
sbr->data[ch].W, sbr->data[ch].Ypos);
- sbr_lf_gen(ac, sbr, sbr->X_low, sbr->data[ch].W, sbr->data[ch].Ypos);
+ sbr->c.sbr_lf_gen(ac, sbr, sbr->X_low, sbr->data[ch].W, sbr->data[ch].Ypos);
sbr->data[ch].Ypos ^= 1;
if (sbr->start) {
- sbr_hf_inverse_filter(&sbr->dsp, sbr->alpha0, sbr->alpha1, sbr->X_low, sbr->k[0]);
+ sbr->c.sbr_hf_inverse_filter(&sbr->dsp, sbr->alpha0, sbr->alpha1, sbr->X_low, sbr->k[0]);
sbr_chirp(sbr, &sbr->data[ch]);
sbr_hf_gen(ac, sbr, sbr->X_high, sbr->X_low, sbr->alpha0, sbr->alpha1,
sbr->data[ch].bw_array, sbr->data[ch].t_env,
if (!err) {
sbr_env_estimate(sbr->e_curr, sbr->X_high, sbr, &sbr->data[ch]);
sbr_gain_calc(ac, sbr, &sbr->data[ch], sbr->data[ch].e_a);
- sbr_hf_assemble(sbr->data[ch].Y[sbr->data[ch].Ypos],
+ sbr->c.sbr_hf_assemble(sbr->data[ch].Y[sbr->data[ch].Ypos],
sbr->X_high, sbr, &sbr->data[ch],
sbr->data[ch].e_a);
}
}
/* synthesis */
- sbr_x_gen(sbr, sbr->X[ch],
+ sbr->c.sbr_x_gen(sbr, sbr->X[ch],
sbr->data[ch].Y[1-sbr->data[ch].Ypos],
sbr->data[ch].Y[ sbr->data[ch].Ypos],
sbr->X_low, ch);
&sbr->data[1].synthesis_filterbank_samples_offset,
downsampled);
}
+
+static void aacsbr_func_ptr_init(AACSBRContext *c)
+{
+ c->sbr_lf_gen = sbr_lf_gen;
+ c->sbr_hf_assemble = sbr_hf_assemble;
+ c->sbr_x_gen = sbr_x_gen;
+ c->sbr_hf_inverse_filter = sbr_hf_inverse_filter;
+
+ if(ARCH_MIPS)
+ ff_aacsbr_func_ptr_init_mips(c);
+}