Recherche avancée

Sur d’autres sites (14576)

• libavutil : add an FFT & MDCT implementation

  2 mai 2019, par Lynne

  libavutil : add an FFT & MDCT implementation
  This commit adds a new API to libavutil to allow for arbitrary transformations
  
  on various types of data.
  
  This is a partly new implementation, with the power of two transforms taken
  
  from libavcodec/fft_template, the 5 and 15-point FFT taken from mdct15, while
  
  the 3-point FFT was written from scratch.
  
  The (i)mdct folding code is taken from mdct15 as well, as the mdct_template
  
  code was somewhat old, messy and not easy to separate.
  A notable feature of this implementation is that it allows for 3xM and 5xM
  
  based transforms, where M is a power of two, e.g. 384, 640, 768, 1280, etc.
  
  AC-4 uses 3xM transforms while Siren uses 5xM transforms, so the code will
  
  allow for decoding of such streams.
  
  A non-exaustive list of supported sizes :
  
  4, 8, 12, 16, 20, 24, 32, 40, 48, 60, 64, 80, 96, 120, 128, 160, 192, 240,
  
  256, 320, 384, 480, 512, 640, 768, 960, 1024, 1280, 1536, 1920, 2048, 2560...
  The API was designed such that it allows for not only 1D transforms but also
  
  2D transforms of certain block sizes. This was partly on accident as the stride
  
  argument is required for Opus MDCTs, but can be used in the context of a 2D
  
  transform as well.
  
  Also, various data types would be implemented eventually as well, such as
  
  "double" and "int32_t".
  Some performance comparisons with libfftw3f (SIMD disabled for both) :
  
  120 :
  
    22353 decicycles in     fftwf_execute,     1024 runs,      0 skips
  
    21836 decicycles in compound_fft_15x8,     1024 runs,      0 skips
  128 :
  
    22003 decicycles in       fftwf_execute,   1024 runs,      0 skips
  
    23132 decicycles in monolithic_fft_ptwo,   1024 runs,      0 skips
  384 :
  
    75939 decicycles in      fftwf_execute,    1024 runs,      0 skips
  
    73973 decicycles in compound_fft_3x128,    1024 runs,      0 skips
  640 :
  
   104354 decicycles in       fftwf_execute,   1024 runs,      0 skips
  
   149518 decicycles in compound_fft_5x128,    1024 runs,      0 skips
  768 :
  
   109323 decicycles in      fftwf_execute,    1024 runs,      0 skips
  
   164096 decicycles in compound_fft_3x256,    1024 runs,      0 skips
  960 :
  
   186210 decicycles in      fftwf_execute,    1024 runs,      0 skips
  
   215256 decicycles in compound_fft_15x64,    1024 runs,      0 skips
  1024 :
  
   163464 decicycles in       fftwf_execute,   1024 runs,      0 skips
  
   199686 decicycles in monolithic_fft_ptwo,   1024 runs,      0 skips
  With SIMD we should be faster than fftw for 15xM transforms as our fft15 SIMD
  
  is around 2x faster than theirs, even if our ptwo SIMD is slightly slower.
  The goal is to remove the libavcodec/mdct15 code and deprecate the
  
  libavcodec/avfft interface once aarch64 and x86 SIMD code has been ported.
  
  New code throughout the project should use this API.
  The implementation passes fate when used in Opus, AAC and Vorbis, and the output
  
  is identical with ATRAC9 as well.
  

  • [DH] libavutil/Makefile
  • [DH] libavutil/tx.c
  • [DH] libavutil/tx.h

• FFMPEG - Codec Copy transcoded audio from first output transport stream to other transport streams

  27 mars 2019, par Software Development Consultan

  I am currently trying to generate 3 output Transport streams from single input stream, for this I need to do Video encoding for each output as I wanted to have different bitrate for each stream. But for audio stream I have same audio bitrate and same params for all 3 output Transport stream.
  So I want to somehow transcode audio only first time (in first stream, let say from AC3 to AAC) and then just want to copy output transcoded audio from first output (aac) to other 2 outputs. so that will save processing power.

  I have setup overall command for this, but for audio copy I am not able to set so that I can get output audio from first Transport stream output

  Please refer following command :

  ffmpeg -loglevel 40 -analyzeduration 10M -probesize 10M -thread_queue_size 1024 -c:v h264_cuvid -gpu 2 -i "udp ://@239.1.1.111:1234 ?fifo_size=1000000&buffer_size=100000&overrun_nonfatal=1" -filter_complex "[0:v]split=3[cl1][cl2][temp1] ;[temp1]scale=-1:480[cl3]" -map [cl1] -map 0:a:0 -c:v h264_nvenc -gpu 2 -preset hq -strict experimental -b:v 4500k -bufsize 5000k -bf 3 -profile:v high -level 4 -coder 1 -sc_threshold 0 -minrate 4500k -maxrate 4500k -g 15 -acodec libfdk_aac -b:a 192k -ar 48000 -ac 2 -flush_packets 0 -f mpegts -mpegts_flags +resend_headers+pat_pmt_at_frames -metadata service_provider=Test_OP1 -metadata service_name=MyService1 -max_muxing_queue_size 1024 -pcr_period 20 -y "udp ://239.1.1.222:1111 ?pkt_size=1316&reuse=1&buffer_size=1000&fifo_size=1000000&burst_bits=50000" -map [cl2] -map 0:a:0 -c:v h264_nvenc -gpu 2 -preset hq -strict experimental -b:v 3500k -bufsize 4000k -bf 3 -profile:v high -level 4 -coder 1 -sc_threshold 0 -minrate 3500k -maxrate 3500k -g 15 -acodec copy -flush_packets 0 -f mpegts -mpegts_flags +resend_headers+pat_pmt_at_frames -metadata service_provider=Test_OP2 -metadata service_name=MyService2 -max_muxing_queue_size 1024 -pcr_period 20 -y "udp ://239.1.1.222:2222 ?pkt_size=1316&reuse=1&buffer_size=1000&fifo_size=1000000&burst_bits=50000" -map [cl3] -map 0:a:0 -c:v h264_nvenc -gpu 2 -preset hq -strict experimental -b:v 2500k -bufsize 3000k -bf 3 -profile:v high -level 4 -minrate 2500k -maxrate 2500k -g 15 -coder 1 -sc_threshold 0 -acodec copy -flush_packets 0 -f mpegts -mpegts_flags +resend_headers+pat_pmt_at_frames -metadata service_provider=Test_OP3 -metadata service_name=MyService3 -max_muxing_queue_size 1024 -pcr_period 20 -y "udp ://239.1.1.222:3333 ?pkt_size=1316&reuse=1&buffer_size=1000&fifo_size=1000000&burst_bits=50000"

  I want to have output stream "239.1.1.222:2222" and "239.1.1.222:3333" should have copied audio stream (AAC) from first output stream "239.1.1.222:1111" and Not from main input stream "239.1.1.111:1234"

• ffmpeg record timelapse camera v4l2

  21 mars 2019, par Paul G.

  How can I instruct ffmpeg (v. : 3.4.5) to record only one frame per minute from my video camera at /dev/video0 and copy it to a new file with 30 frames per second ? Goal is timelapse video of some months and record only one frame per minute into the out video to save space and processing power instead of just capturing the camera at 30fps and then preprocess the huge video file for speed up...

  The camera delivers a 1920x1080 30fps stream with "mjpeg" selected in commandline. I would like to keep the raw frames and just copy them into the out file. I tried this command to get 1fps input to 30fps output but this seems not to work :

  ffmpeg -framerate 1 -input_format mjpeg -i /dev/video0 -r 30 -c:v copy -an out.mp4
  
  
  
  [video4linux2,v4l2 @ 0x55ece63ac360] The driver changed the time per frame from 1/1 to 1/30
  
  Input #0, video4linux2,v4l2, from '/dev/video0':
  
    Duration: N/A, start: 37064.594605, bitrate: N/A
  
      Stream #0:0: Video: mjpeg, yuvj422p(pc, bt470bg/unknown/unknown), 1920x1080, 30 fps, 30 tbr, 1000k tbn, 1000k tbc

  But this doesnt work.