Mercurial > libavcodec.hg
view libvorbis.c @ 10874:bcfe2acbf190 libavcodec
AAC: Compress codebook tables and optimise sign bit handling
The codebooks each consist of small number of values repeated in
groups of 2 or 4. Storing the codebooks as a packed list of 2- or
4-bit indexes into a table reduces their size substantially (from 7.5k
to 1.5k), resulting in less cache pressure.
For the band types with sign bits in the bitstream, storing the number
and position of non-zero codebook values using a few bits avoids
multiple get_bits() calls and floating-point comparisons which gcc
handles miserably.
Some float/int type punning also avoids gcc brain damage.
Overall speedup 20-35% on Cortex-A8, 20% on Core i7.
| author | mru |
|---|---|
| date | Wed, 13 Jan 2010 16:46:28 +0000 |
| parents | 7955db355703 |
| children | 8a4984c5cacc |
line wrap: on
line source
/* * copyright (c) 2002 Mark Hills <mark@pogo.org.uk> * * 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 libavcodec/libvorbis.c * Ogg Vorbis codec support via libvorbisenc. * @author Mark Hills <mark@pogo.org.uk> */ #include <vorbis/vorbisenc.h> #include "avcodec.h" #include "bytestream.h" #undef NDEBUG #include <assert.h> #define OGGVORBIS_FRAME_SIZE 64 #define BUFFER_SIZE (1024*64) typedef struct OggVorbisContext { vorbis_info vi ; vorbis_dsp_state vd ; vorbis_block vb ; uint8_t buffer[BUFFER_SIZE]; int buffer_index; int eof; /* decoder */ vorbis_comment vc ; ogg_packet op; } OggVorbisContext ; static av_cold int oggvorbis_init_encoder(vorbis_info *vi, AVCodecContext *avccontext) { double cfreq; if(avccontext->flags & CODEC_FLAG_QSCALE) { /* variable bitrate */ if(vorbis_encode_setup_vbr(vi, avccontext->channels, avccontext->sample_rate, avccontext->global_quality / (float)FF_QP2LAMBDA / 10.0)) return -1; } else { /* constant bitrate */ if(vorbis_encode_setup_managed(vi, avccontext->channels, avccontext->sample_rate, -1, avccontext->bit_rate, -1)) return -1; #ifdef OGGVORBIS_VBR_BY_ESTIMATE /* variable bitrate by estimate */ if(vorbis_encode_ctl(vi, OV_ECTL_RATEMANAGE_AVG, NULL)) return -1; #endif } /* cutoff frequency */ if(avccontext->cutoff > 0) { cfreq = avccontext->cutoff / 1000.0; if(vorbis_encode_ctl(vi, OV_ECTL_LOWPASS_SET, &cfreq)) return -1; } return vorbis_encode_setup_init(vi); } static av_cold int oggvorbis_encode_init(AVCodecContext *avccontext) { OggVorbisContext *context = avccontext->priv_data ; ogg_packet header, header_comm, header_code; uint8_t *p; unsigned int offset, len; vorbis_info_init(&context->vi) ; if(oggvorbis_init_encoder(&context->vi, avccontext) < 0) { av_log(avccontext, AV_LOG_ERROR, "oggvorbis_encode_init: init_encoder failed\n") ; return -1 ; } vorbis_analysis_init(&context->vd, &context->vi) ; vorbis_block_init(&context->vd, &context->vb) ; vorbis_comment_init(&context->vc); vorbis_comment_add_tag(&context->vc, "encoder", LIBAVCODEC_IDENT) ; vorbis_analysis_headerout(&context->vd, &context->vc, &header, &header_comm, &header_code); len = header.bytes + header_comm.bytes + header_code.bytes; avccontext->extradata_size= 64 + len + len/255; p = avccontext->extradata= av_mallocz(avccontext->extradata_size); p[0] = 2; offset = 1; offset += av_xiphlacing(&p[offset], header.bytes); offset += av_xiphlacing(&p[offset], header_comm.bytes); memcpy(&p[offset], header.packet, header.bytes); offset += header.bytes; memcpy(&p[offset], header_comm.packet, header_comm.bytes); offset += header_comm.bytes; memcpy(&p[offset], header_code.packet, header_code.bytes); offset += header_code.bytes; avccontext->extradata_size = offset; avccontext->extradata= av_realloc(avccontext->extradata, avccontext->extradata_size); /* vorbis_block_clear(&context->vb); vorbis_dsp_clear(&context->vd); vorbis_info_clear(&context->vi);*/ vorbis_comment_clear(&context->vc); avccontext->frame_size = OGGVORBIS_FRAME_SIZE ; avccontext->coded_frame= avcodec_alloc_frame(); avccontext->coded_frame->key_frame= 1; return 0 ; } static int oggvorbis_encode_frame(AVCodecContext *avccontext, unsigned char *packets, int buf_size, void *data) { OggVorbisContext *context = avccontext->priv_data ; ogg_packet op ; signed short *audio = data ; int l; if(data) { int samples = OGGVORBIS_FRAME_SIZE; float **buffer ; buffer = vorbis_analysis_buffer(&context->vd, samples) ; if(context->vi.channels == 1) { for(l = 0 ; l < samples ; l++) buffer[0][l]=audio[l]/32768.f; } else { for(l = 0 ; l < samples ; l++){ buffer[0][l]=audio[l*2]/32768.f; buffer[1][l]=audio[l*2+1]/32768.f; } } vorbis_analysis_wrote(&context->vd, samples) ; } else { if(!context->eof) vorbis_analysis_wrote(&context->vd, 0) ; context->eof = 1; } while(vorbis_analysis_blockout(&context->vd, &context->vb) == 1) { vorbis_analysis(&context->vb, NULL); vorbis_bitrate_addblock(&context->vb) ; while(vorbis_bitrate_flushpacket(&context->vd, &op)) { /* i'd love to say the following line is a hack, but sadly it's * not, apparently the end of stream decision is in libogg. */ if(op.bytes==1) continue; memcpy(context->buffer + context->buffer_index, &op, sizeof(ogg_packet)); context->buffer_index += sizeof(ogg_packet); memcpy(context->buffer + context->buffer_index, op.packet, op.bytes); context->buffer_index += op.bytes; // av_log(avccontext, AV_LOG_DEBUG, "e%d / %d\n", context->buffer_index, op.bytes); } } l=0; if(context->buffer_index){ ogg_packet *op2= (ogg_packet*)context->buffer; op2->packet = context->buffer + sizeof(ogg_packet); l= op2->bytes; avccontext->coded_frame->pts= av_rescale_q(op2->granulepos, (AVRational){1, avccontext->sample_rate}, avccontext->time_base); //FIXME we should reorder the user supplied pts and not assume that they are spaced by 1/sample_rate memcpy(packets, op2->packet, l); context->buffer_index -= l + sizeof(ogg_packet); memcpy(context->buffer, context->buffer + l + sizeof(ogg_packet), context->buffer_index); // av_log(avccontext, AV_LOG_DEBUG, "E%d\n", l); } return l; } static av_cold int oggvorbis_encode_close(AVCodecContext *avccontext) { OggVorbisContext *context = avccontext->priv_data ; /* ogg_packet op ; */ vorbis_analysis_wrote(&context->vd, 0) ; /* notify vorbisenc this is EOF */ vorbis_block_clear(&context->vb); vorbis_dsp_clear(&context->vd); vorbis_info_clear(&context->vi); av_freep(&avccontext->coded_frame); av_freep(&avccontext->extradata); return 0 ; } AVCodec libvorbis_encoder = { "libvorbis", CODEC_TYPE_AUDIO, CODEC_ID_VORBIS, sizeof(OggVorbisContext), oggvorbis_encode_init, oggvorbis_encode_frame, oggvorbis_encode_close, .capabilities= CODEC_CAP_DELAY, .sample_fmts = (const enum SampleFormat[]){SAMPLE_FMT_S16,SAMPLE_FMT_NONE}, .long_name= NULL_IF_CONFIG_SMALL("libvorbis Vorbis"), } ;
