X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=libavcodec%2Fwmaprodec.c;h=c9048a91d70dffec15f17ac91d9f4bcc42e19718;hb=20d33ee8b1d2cf5ba4b4d4a81db0ff7b4a125dd3;hp=742896d42e6690a057218b3844851c81fa77b07e;hpb=497d7991b5746307bb5d900bd87ba97463c55f04;p=ffmpeg diff --git a/libavcodec/wmaprodec.c b/libavcodec/wmaprodec.c index 742896d42e6..c9048a91d70 100644 --- a/libavcodec/wmaprodec.c +++ b/libavcodec/wmaprodec.c @@ -1,22 +1,22 @@ /* * Wmapro compatible decoder * Copyright (c) 2007 Baptiste Coudurier, Benjamin Larsson, Ulion - * Copyright (c) 2008 - 2009 Sascha Sommer, Benjamin Larsson + * Copyright (c) 2008 - 2011 Sascha Sommer, Benjamin Larsson * - * This file is part of FFmpeg. + * This file is part of Libav. * - * FFmpeg is free software; you can redistribute it and/or + * Libav 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, + * Libav 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 + * License along with Libav; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA */ @@ -92,6 +92,7 @@ #include "put_bits.h" #include "wmaprodata.h" #include "dsputil.h" +#include "sinewin.h" #include "wma.h" /** current decoder limitations */ @@ -100,9 +101,10 @@ #define MAX_BANDS 29 ///< max number of scale factor bands #define MAX_FRAMESIZE 32768 ///< maximum compressed frame size +#define WMAPRO_BLOCK_MIN_BITS 6 ///< log2 of min block size #define WMAPRO_BLOCK_MAX_BITS 12 ///< log2 of max block size #define WMAPRO_BLOCK_MAX_SIZE (1 << WMAPRO_BLOCK_MAX_BITS) ///< maximum block size -#define WMAPRO_BLOCK_SIZES (WMAPRO_BLOCK_MAX_BITS - BLOCK_MIN_BITS + 1) ///< possible block sizes +#define WMAPRO_BLOCK_SIZES (WMAPRO_BLOCK_MAX_BITS - WMAPRO_BLOCK_MIN_BITS + 1) ///< possible block sizes #define VLCBITS 9 @@ -142,6 +144,7 @@ typedef struct { int* scale_factors; ///< pointer to the scale factor values used for decoding uint8_t table_idx; ///< index in sf_offsets for the scale factor reference block float* coeffs; ///< pointer to the subframe decode buffer + uint16_t num_vec_coeffs; ///< number of vector coded coefficients DECLARE_ALIGNED(16, float, out)[WMAPRO_BLOCK_MAX_SIZE + WMAPRO_BLOCK_MAX_SIZE / 2]; ///< output buffer } WMAProChannelCtx; @@ -190,6 +193,7 @@ typedef struct WMAProDecodeCtx { /* packet decode state */ GetBitContext pgb; ///< bitstream reader context for the packet + int next_packet_start; ///< start offset of the next wma packet in the demuxer packet uint8_t packet_offset; ///< frame offset in the packet uint8_t packet_sequence_number; ///< current packet number int num_saved_bits; ///< saved number of bits @@ -213,6 +217,7 @@ typedef struct WMAProDecodeCtx { int8_t channels_for_cur_subframe; ///< number of channels that contain the subframe int8_t channel_indexes_for_cur_subframe[WMAPRO_MAX_CHANNELS]; int8_t num_bands; ///< number of scale factor bands + int8_t transmit_num_vec_coeffs; ///< number of vector coded coefficients is part of the bitstream int16_t* cur_sfb_offsets; ///< sfb offsets for the current block uint8_t table_idx; ///< index for the num_sfb, sfb_offsets, sf_offsets and subwoofer_cutoffs tables int8_t esc_len; ///< length of escaped coefficients @@ -276,7 +281,7 @@ static av_cold int decode_init(AVCodecContext *avctx) dsputil_init(&s->dsp, avctx); init_put_bits(&s->pb, s->frame_data, MAX_FRAMESIZE); - avctx->sample_fmt = SAMPLE_FMT_FLT; + avctx->sample_fmt = AV_SAMPLE_FMT_FLT; if (avctx->extradata_size >= 18) { s->decode_flags = AV_RL16(edata_ptr+14); @@ -284,8 +289,8 @@ static av_cold int decode_init(AVCodecContext *avctx) s->bits_per_sample = AV_RL16(edata_ptr); /** dump the extradata */ for (i = 0; i < avctx->extradata_size; i++) - dprintf(avctx, "[%x] ", avctx->extradata[i]); - dprintf(avctx, "\n"); + av_dlog(avctx, "[%x] ", avctx->extradata[i]); + av_dlog(avctx, "\n"); } else { av_log_ask_for_sample(avctx, "Unknown extradata size\n"); @@ -300,11 +305,6 @@ static av_cold int decode_init(AVCodecContext *avctx) s->packet_loss = 1; s->len_prefix = (s->decode_flags & 0x40); - if (!s->len_prefix) { - av_log_ask_for_sample(avctx, "no length prefix\n"); - return AVERROR_INVALIDDATA; - } - /** get frame len */ s->samples_per_frame = 1 << ff_wma_get_frame_len_bits(avctx->sample_rate, 3, s->decode_flags); @@ -316,7 +316,7 @@ static av_cold int decode_init(AVCodecContext *avctx) /** subframe info */ log2_max_num_subframes = ((s->decode_flags & 0x38) >> 3); s->max_num_subframes = 1 << log2_max_num_subframes; - if (s->max_num_subframes == 16) + if (s->max_num_subframes == 16 || s->max_num_subframes == 4) s->max_subframe_len_bit = 1; s->subframe_len_bits = av_log2(log2_max_num_subframes) + 1; @@ -422,8 +422,8 @@ static av_cold int decode_init(AVCodecContext *avctx) /** init MDCT, FIXME: only init needed sizes */ for (i = 0; i < WMAPRO_BLOCK_SIZES; i++) - ff_mdct_init(&s->mdct_ctx[i], BLOCK_MIN_BITS+1+i, 1, - 1.0 / (1 << (BLOCK_MIN_BITS + i - 1)) + ff_mdct_init(&s->mdct_ctx[i], WMAPRO_BLOCK_MIN_BITS+1+i, 1, + 1.0 / (1 << (WMAPRO_BLOCK_MIN_BITS + i - 1)) / (1 << (s->bits_per_sample - 1))); /** init MDCT windows: simple sinus window */ @@ -583,7 +583,7 @@ static int decode_tilehdr(WMAProDecodeCtx *s) int i; int offset = 0; for (i = 0; i < s->channel[c].num_subframes; i++) { - dprintf(s->avctx, "frame[%i] channel[%i] subframe[%i]" + av_dlog(s->avctx, "frame[%i] channel[%i] subframe[%i]" " len %i\n", s->frame_num, c, i, s->channel[c].subframe_len[i]); s->channel[c].subframe_offset[i] = offset; @@ -782,7 +782,7 @@ static int decode_coeffs(WMAProDecodeCtx *s, int c) const uint16_t* run; const float* level; - dprintf(s->avctx, "decode coefficients for channel %i\n", c); + av_dlog(s->avctx, "decode coefficients for channel %i\n", c); vlctable = get_bits1(&s->gb); vlc = &coef_vlc[vlctable]; @@ -797,7 +797,8 @@ static int decode_coeffs(WMAProDecodeCtx *s, int c) /** decode vector coefficients (consumes up to 167 bits per iteration for 4 vector coded large values) */ - while (!rl_mode && cur_coeff + 3 < s->subframe_len) { + while ((s->transmit_num_vec_coeffs || !rl_mode) && + (cur_coeff + 3 < ci->num_vec_coeffs)) { int vals[4]; int i; unsigned int idx; @@ -846,7 +847,7 @@ static int decode_coeffs(WMAProDecodeCtx *s, int c) } /** decode run level coded coefficients */ - if (rl_mode) { + if (cur_coeff < s->subframe_len) { memset(&ci->coeffs[cur_coeff], 0, sizeof(*ci->coeffs) * (s->subframe_len - cur_coeff)); if (ff_wma_run_level_decode(s->avctx, &s->gb, vlc, @@ -1026,12 +1027,12 @@ static void wmapro_window(WMAProDecodeCtx *s) winlen = s->subframe_len; } - window = s->windows[av_log2(winlen) - BLOCK_MIN_BITS]; + window = s->windows[av_log2(winlen) - WMAPRO_BLOCK_MIN_BITS]; winlen >>= 1; s->dsp.vector_fmul_window(start, start, start + winlen, - window, 0, winlen); + window, winlen); s->channel[c].prev_block_len = s->subframe_len; } @@ -1066,7 +1067,7 @@ static int decode_subframe(WMAProDecodeCtx *s) } } - dprintf(s->avctx, + av_dlog(s->avctx, "processing subframe with offset %i len %i\n", offset, subframe_len); /** get a list of all channels that contain the estimated block */ @@ -1093,7 +1094,7 @@ static int decode_subframe(WMAProDecodeCtx *s) s->parsed_all_subframes = 1; - dprintf(s->avctx, "subframe is part of %i channels\n", + av_dlog(s->avctx, "subframe is part of %i channels\n", s->channels_for_cur_subframe); /** calculate number of scale factor bands and their offsets */ @@ -1151,10 +1152,19 @@ static int decode_subframe(WMAProDecodeCtx *s) if (transmit_coeffs) { int step; int quant_step = 90 * s->bits_per_sample >> 4; - if ((get_bits1(&s->gb))) { - /** FIXME: might change run level mode decision */ - av_log_ask_for_sample(s->avctx, "unsupported quant step coding\n"); - return AVERROR_INVALIDDATA; + + /** decode number of vector coded coefficients */ + if ((s->transmit_num_vec_coeffs = get_bits1(&s->gb))) { + int num_bits = av_log2((s->subframe_len + 3)/4) + 1; + for (i = 0; i < s->channels_for_cur_subframe; i++) { + int c = s->channel_indexes_for_cur_subframe[i]; + s->channel[c].num_vec_coeffs = get_bits(&s->gb, num_bits) << 2; + } + } else { + for (i = 0; i < s->channels_for_cur_subframe; i++) { + int c = s->channel_indexes_for_cur_subframe[i]; + s->channel[c].num_vec_coeffs = s->subframe_len; + } } /** decode quantization step */ step = get_sbits(&s->gb, 6); @@ -1195,7 +1205,7 @@ static int decode_subframe(WMAProDecodeCtx *s) return AVERROR_INVALIDDATA; } - dprintf(s->avctx, "BITSTREAM: subframe header length was %i\n", + av_dlog(s->avctx, "BITSTREAM: subframe header length was %i\n", get_bits_count(&s->gb) - s->subframe_offset); /** parse coefficients */ @@ -1209,10 +1219,11 @@ static int decode_subframe(WMAProDecodeCtx *s) sizeof(*s->channel[c].coeffs) * subframe_len); } - dprintf(s->avctx, "BITSTREAM: subframe length was %i\n", + av_dlog(s->avctx, "BITSTREAM: subframe length was %i\n", get_bits_count(&s->gb) - s->subframe_offset); if (transmit_coeffs) { + FFTContext *mdct = &s->mdct_ctx[av_log2(subframe_len) - WMAPRO_BLOCK_MIN_BITS]; /** reconstruct the per channel data */ inverse_channel_transform(s); for (i = 0; i < s->channels_for_cur_subframe; i++) { @@ -1237,9 +1248,8 @@ static int decode_subframe(WMAProDecodeCtx *s) quant, end - start); } - /** apply imdct (ff_imdct_half == DCTIV with reverse) */ - ff_imdct_half(&s->mdct_ctx[av_log2(subframe_len) - BLOCK_MIN_BITS], - s->channel[c].coeffs, s->tmp); + /** apply imdct (imdct_half == DCTIV with reverse) */ + mdct->imdct_half(mdct, s->channel[c].coeffs, s->tmp); } } @@ -1285,7 +1295,7 @@ static int decode_frame(WMAProDecodeCtx *s) if (s->len_prefix) len = get_bits(gb, s->log2_frame_size); - dprintf(s->avctx, "decoding frame with length %x\n", len); + av_dlog(s->avctx, "decoding frame with length %x\n", len); /** decode tile information */ if (decode_tilehdr(s)) { @@ -1295,15 +1305,16 @@ static int decode_frame(WMAProDecodeCtx *s) /** read postproc transform */ if (s->num_channels > 1 && get_bits1(gb)) { - av_log_ask_for_sample(s->avctx, "Unsupported postproc transform found\n"); - s->packet_loss = 1; - return 0; + if (get_bits1(gb)) { + for (i = 0; i < s->num_channels * s->num_channels; i++) + skip_bits(gb, 4); + } } /** read drc info */ if (s->dynamic_range_compression) { s->drc_gain = get_bits(gb, 8); - dprintf(s->avctx, "drc_gain %i\n", s->drc_gain); + av_dlog(s->avctx, "drc_gain %i\n", s->drc_gain); } /** no idea what these are for, might be the number of samples @@ -1314,18 +1325,18 @@ static int decode_frame(WMAProDecodeCtx *s) /** usually true for the first frame */ if (get_bits1(gb)) { skip = get_bits(gb, av_log2(s->samples_per_frame * 2)); - dprintf(s->avctx, "start skip: %i\n", skip); + av_dlog(s->avctx, "start skip: %i\n", skip); } /** sometimes true for the last frame */ if (get_bits1(gb)) { skip = get_bits(gb, av_log2(s->samples_per_frame * 2)); - dprintf(s->avctx, "end skip: %i\n", skip); + av_dlog(s->avctx, "end skip: %i\n", skip); } } - dprintf(s->avctx, "BITSTREAM: frame header length was %i\n", + av_dlog(s->avctx, "BITSTREAM: frame header length was %i\n", get_bits_count(gb) - s->frame_offset); /** reset subframe states */ @@ -1368,16 +1379,22 @@ static int decode_frame(WMAProDecodeCtx *s) } else s->samples += s->num_channels * s->samples_per_frame; - if (len != (get_bits_count(gb) - s->frame_offset) + 2) { - /** FIXME: not sure if this is always an error */ - av_log(s->avctx, AV_LOG_ERROR, "frame[%i] would have to skip %i bits\n", - s->frame_num, len - (get_bits_count(gb) - s->frame_offset) - 1); - s->packet_loss = 1; - return 0; - } + if (s->len_prefix) { + if (len != (get_bits_count(gb) - s->frame_offset) + 2) { + /** FIXME: not sure if this is always an error */ + av_log(s->avctx, AV_LOG_ERROR, + "frame[%i] would have to skip %i bits\n", s->frame_num, + len - (get_bits_count(gb) - s->frame_offset) - 1); + s->packet_loss = 1; + return 0; + } - /** skip the rest of the frame data */ - skip_bits_long(gb, len - (get_bits_count(gb) - s->frame_offset) - 1); + /** skip the rest of the frame data */ + skip_bits_long(gb, len - (get_bits_count(gb) - s->frame_offset) - 1); + } else { + while (get_bits_count(gb) < s->num_saved_bits && get_bits1(gb) == 0) { + } + } /** decode trailer bit */ more_frames = get_bits1(gb); @@ -1473,13 +1490,14 @@ static int decode_packet(AVCodecContext *avctx, if (s->packet_done || s->packet_loss) { s->packet_done = 0; - s->buf_bit_size = buf_size << 3; /** sanity check for the buffer length */ if (buf_size < avctx->block_align) return 0; + s->next_packet_start = buf_size - avctx->block_align; buf_size = avctx->block_align; + s->buf_bit_size = buf_size << 3; /** parse packet header */ init_get_bits(gb, buf, s->buf_bit_size); @@ -1488,7 +1506,7 @@ static int decode_packet(AVCodecContext *avctx, /** get number of bits that need to be added to the previous frame */ num_bits_prev_frame = get_bits(gb, s->log2_frame_size); - dprintf(avctx, "packet[%d]: nbpf %x\n", avctx->frame_number, + av_dlog(avctx, "packet[%d]: nbpf %x\n", avctx->frame_number, num_bits_prev_frame); /** check for packet loss */ @@ -1501,32 +1519,54 @@ static int decode_packet(AVCodecContext *avctx, s->packet_sequence_number = packet_sequence_number; if (num_bits_prev_frame > 0) { + int remaining_packet_bits = s->buf_bit_size - get_bits_count(gb); + if (num_bits_prev_frame >= remaining_packet_bits) { + num_bits_prev_frame = remaining_packet_bits; + s->packet_done = 1; + } + /** append the previous frame data to the remaining data from the previous packet to create a full frame */ save_bits(s, gb, num_bits_prev_frame, 1); - dprintf(avctx, "accumulated %x bits of frame data\n", + av_dlog(avctx, "accumulated %x bits of frame data\n", s->num_saved_bits - s->frame_offset); /** decode the cross packet frame if it is valid */ if (!s->packet_loss) decode_frame(s); } else if (s->num_saved_bits - s->frame_offset) { - dprintf(avctx, "ignoring %x previously saved bits\n", + av_dlog(avctx, "ignoring %x previously saved bits\n", s->num_saved_bits - s->frame_offset); } - s->packet_loss = 0; + if (s->packet_loss) { + /** reset number of saved bits so that the decoder + does not start to decode incomplete frames in the + s->len_prefix == 0 case */ + s->num_saved_bits = 0; + s->packet_loss = 0; + } } else { int frame_size; - s->buf_bit_size = avpkt->size << 3; + s->buf_bit_size = (avpkt->size - s->next_packet_start) << 3; init_get_bits(gb, avpkt->data, s->buf_bit_size); skip_bits(gb, s->packet_offset); - if (remaining_bits(s, gb) > s->log2_frame_size && + if (s->len_prefix && remaining_bits(s, gb) > s->log2_frame_size && (frame_size = show_bits(gb, s->log2_frame_size)) && frame_size <= remaining_bits(s, gb)) { save_bits(s, gb, frame_size, 0); s->packet_done = !decode_frame(s); + } else if (!s->len_prefix + && s->num_saved_bits > get_bits_count(&s->gb)) { + /** when the frames do not have a length prefix, we don't know + the compressed length of the individual frames + however, we know what part of a new packet belongs to the + previous frame + therefore we save the incoming packet first, then we append + the "previous frame" data from the next packet so that + we get a buffer that only contains full frames */ + s->packet_done = !decode_frame(s); } else s->packet_done = 1; } @@ -1564,7 +1604,7 @@ static void flush(AVCodecContext *avctx) /** *@brief wmapro decoder */ -AVCodec wmapro_decoder = { +AVCodec ff_wmapro_decoder = { "wmapro", AVMEDIA_TYPE_AUDIO, CODEC_ID_WMAPRO,