Recherche avancée

Sur d’autres sites (5683)

• FFMPEG with Neon Optimization

  11 juin 2015, par madadi

  I am decoding an h.264 video file on android using ffmpeg. The performace is very low. I would like to enable neon optimzation in ffmpeg to improve the performace. Inspite of adding the neon related commands in the config file, I don’t see performance gain. Can someone tell me if I am going wrong anywhere ? My config file is as given below.

      function build_one_r6
  
  {
  
  make clean
  
  ./configure \
  
      --prefix=$PREFIX \
  
      --disable-shared \
  
      --enable-static \
  
      --enable-version3 \
  
      --disable-doc \
  
      --disable-ffmpeg \
  
      --disable-ffplay \
  
      --disable-ffprobe \
  
      --disable-ffserver \
  
      --disable-avdevice \
  
      --disable-avfilter \
  
      --disable-postproc \
  
      --enable-small \
  
      --cross-prefix=$PREBUILT/bin/arm-linux-androideabi- \
  
      --enable-cross-compile \
  
      --target-os=linux \
  
      --extra-cflags="-I$PLATFORM/usr/include -fpic -DANDROID -DHAVE_SYS_UIO_H=1 -fasm -Wno-psabi -fno-short-enums -fno-strict-aliasing $OPTIMIZE_CFLAGS" \
  
      --extra-ldflags="-Wl,-rpath-link=$PLATFORM/usr/lib -L$PLATFORM/usr/lib -nostdlib -lc -lm -ldl -llog" \
  
      --arch=arm \
  
      --disable-symver \
  
      --disable-debug \
  
      --disable-indevs \
  
      --disable-encoders \
  
      --disable-bsfs \
  
      --disable-filters \
  
  $ADDITIONAL_CONFIGURE_FLAG || exit 1;
  
  sed -i 's/HAVE_LRINT 0/HAVE_LRINT 1/g' config.h
  
  sed -i 's/HAVE_LRINTF 0/HAVE_LRINTF 1/g' config.h
  
  sed -i 's/HAVE_ROUND 0/HAVE_ROUND 1/g' config.h
  
  sed -i 's/HAVE_ROUNDF 0/HAVE_ROUNDF 1/g' config.h
  
  sed -i 's/HAVE_TRUNC 0/HAVE_TRUNC 1/g' config.h
  
  sed -i 's/HAVE_TRUNCF 0/HAVE_TRUNCF 1/g' config.h
  
  make  -j4 install || exit 1;
  
  $PREBUILT/bin/arm-linux-androideabi-ar d libavcodec/libavcodec.a inverse.o || exit 1;
  
  $PREBUILT/bin/arm-linux-androideabi-ld -rpath-link=$PLATFORM/usr/lib -L$PLATFORM/usr/lib  -soname libffmpeg.so -shared -nostdlib  -z,noexecstack -Bsymbolic --whole-archive --no-undefined -o $PREFIX/lib/libffmpeg.so libavcodec/libavcodec.a libavformat/libavformat.a libavutil/libavutil.a libswscale/libswscale.a -lc -lm -lz -ldl -llog  --warn-once  --dynamic-linker=/system/bin/linker $PREBUILT/lib/gcc/arm-linux-androideabi/4.4.3/libgcc.a || exit 1;
  
  
  
  }
  
  
  
  
  
  #armv5te
  
  CPU=armv5te
  
  OPTIMIZE_CFLAGS="-marm -march=$CPU "
  
  PREFIX=../../build/android/armeabi
  
  ADDITIONAL_CONFIGURE_FLAG=
  
  build_one_r6
  
  
  
  
  
  #arm v7n
  
  CPU=armv7-a
  
  OPTIMIZE_CFLAGS="-mfloat-abi=softfp -mfpu=neon -marm -march=$CPU -mtune=cortex-a8"
  
  PREFIX=../../build/android/armeabi-v7a
  
  ADDITIONAL_CONFIGURE_FLAG="--enable-neon --disable-armv5te --disable-armv6 --disable-armv6t2"
  
  build_one_r6enter code here

• Unable to find audio stream

  18 novembre 2011, par Kurt

  I'm trying to adapt a code in order to add a sound capture feature (on a live stream), with the help of ffmpeg and directshow.

  When i try to play with ffplay the AVIsynth file, everything works perfectly i've got the audio and video. But when i open this input file by code i only find the video stream.

  The Avs file :

  V = DirectShowSource("Decklink_HDMI.grf", fps=10, framecount=1000000000, seek=false, audio=false)
  
  A = DirectShowSource("Decklink_Audio.grf", fps=1, framecount=1000000000, video=false)
  
  AudioDub(V, A)

  The opening code :

  ffmpeg::AVInputFormat * ifmt;
  
  
  
  ifmt = ffmpeg::av_find_input_format("avs");
  
  
  
  // Open input file
  
   if(ffmpeg::avformat_open_input(&pFormatCtx, filename.toStdString().c_str(), ifmt, NULL) != 0)

  When i make a variable lookout on gdb just after the opening.

  i'm looking at nb_streams in pFormatCtx->nb_streams and it's at 1
  The only stream i can find in pFormatCtx->stream is a video one. And that's why i'm not able to capture the sound.

  So i'm asking if i'm doing something wrong with my opening or if i miss something.

  Ps : I know the code shown is short but the problem appears at this very step !

  Thank you

  Kurt

  — EDIT —

  I've also noticed that when I dump the AVFormatContext the video stream got a 456x32 size.
  And i can find this very same size of the window displayed when i try to launch a corrupted script with ffplay.

  The original video format when i play the correct script with ffplay is of 1920x1080

  I think my problem is maybe deeper than the simple fact of not being able to get the audio stream.

  I'm trying to find out how to know the error message that is displayed on this 456x32 windows

  — EDIT2 —

  I Find out what is written on this image and my problem is solved, badly placed avs script an old one was is place.

  I'm ashamed.

• WMV Converter Fails FFMpeg

  15 novembre 2011, par zomerc

  I just cann't seem to get the FFMpeg working with using the library. Everytime I try to convert asf file to wmv. I get the following issue on run time :

  [wmv1 @ 0x81ee000]error, slice code was 2
  [wmv1 @ 0x81ee000]header damaged

  This my code :

  static void audio_decode_example(const char *outfilename, const char *filename)
  
  {
  
      AVCodec *codec;
  
      AVCodecContext *c= NULL;
  
      int out_size, len, in_size;
  
      FILE *f, *outfile;
  
      uint8_t *outbuf;
  
      uint8_t inbuf[AUDIO_INBUF_SIZE + FF_INPUT_BUFFER_PADDING_SIZE];
  
      AVPacket avpkt;
  
  
  
      av_init_packet(&avpkt);
  
  
  
      printf("Audio decoding\n");
  
  
  
      /* find the mpeg audio decoder */
  
      codec = avcodec_find_decoder(CODEC_ID_WMV1);
  
      if (!codec) {
  
          fprintf(stderr, "codec not found\n");
  
          return;
  
      }
  
  
  
      c= avcodec_alloc_context2(CODEC_TYPE_AUDIO);
  
  
  
      /* open it */
  
      if (avcodec_open(c, codec) < 0) {
  
          fprintf(stderr, "could not open codec\n");
  
          return;
  
      }
  
  
  
      outbuf = malloc(AVCODEC_MAX_AUDIO_FRAME_SIZE);
  
      f = fopen(filename, "rb");
  
      if (!f) {
  
          fprintf(stderr, "could not open %s\n", filename);
  
          return;
  
      }
  
      outfile = fopen(outfilename, "wb");
  
      if (!outfile) {
  
          av_free(c);
  
          return;
  
      }
  
  
  
      /* decode until eof */
  
      avpkt.data = inbuf;
  
      len = avpkt.size = fread(inbuf, 1, INBUF_SIZE, f);
  
      NSLog(@"%d", avpkt.size); 
  
  
  
      while (avpkt.size > 0) {
  
          out_size = AVCODEC_MAX_AUDIO_FRAME_SIZE;
  
      len = avcodec_decode_audio2(c, (short *)outbuf, &out_size, inbuf,len);// avpkt.size);
  
      NSLog(@"%d", len);
  
      if (len < 0) {
  
          fprintf(stderr, "Error while decoding\n");
  
          fclose(outfile);
  
          return;
  
      }
  
      if (out_size > 0) {
  
          /* if a frame has been decoded, output it */
  
          fwrite(outbuf, 1, out_size, outfile);
  
      }
  
      avpkt.size -= len;
  
      avpkt.data += len;
  
      if (avpkt.size < AUDIO_REFILL_THRESH) {
  
          /* Refill the input buffer, to avoid trying to decode
  
           * incomplete frames. Instead of this, one could also use
  
           * a parser, or use a proper container format through
  
           * libavformat. */
  
          memmove(inbuf, avpkt.data, avpkt.size);
  
          avpkt.data = inbuf;
  
          len = fread(avpkt.data + avpkt.size, 1,
  
                      INBUF_SIZE - avpkt.size, f);
  
          if (len > 0)
  
              avpkt.size += len;
  
      }
  
  }
  
  
  
  fclose(outfile);
  
  fclose(f);
  
  free(outbuf);
  
  avcodec_close(c);
  
  av_free(c);
  
  }

  I have try the command line utilities and it successfully convert the file. Any help would be helpfully. thanks