Recherche avancée

Sur d’autres sites (11547)

• FFmpeg "faststart" flag with images to video

  22 août 2016, par Calvin W.

  I am using MoviePy to process my video and stream it.

  In MoviePy, it decode the input file to a frames * width * heigh * 3 (RGB values) NumPy array. And when it encode the video, it pass the NumPy array frame by frame to standard input of ffmpeg.

  In order to stream the video on-the-fly, I have to move the video header to the front.

  Is it posible to use faststart flag with such video encoding ?

• WASAPI resampling

  30 juin 2013, par magingax

  I am trying to play 24bit/48000 Hz audio on PC. my PC seems not support 32bit sample (event GetMixFormat return 32bit/44100Hz device)
  So I'm trying to convert 24bit/4800Hz audio into 16bit/44100Hz
  but It sound like noise.
  sample audio file 'titan.wav' is 24bit/48000Hz s32 format
  below is sample code

  #define WINVER 0x0600
  
  #define _WIN32_WINNT 0x0600
  
  
  
  #include 
  
  #include <iostream>
  
  #include "Mmdeviceapi.h"
  
  #include "Audioclient.h"
  
  #include "Endpointvolume.h"
  
  using namespace std;
  
  
  
  extern "C"
  
  {
  
  #include <libavformat></libavformat>avformat.h>
  
  #include <libavcodec></libavcodec>avcodec.h>
  
  #include <libswscale></libswscale>swscale.h>
  
  #include <libavutil></libavutil>opt.h>
  
  #include <libavutil></libavutil>channel_layout.h>
  
  #include <libavutil></libavutil>samplefmt.h>
  
  #include <libswresample></libswresample>swresample.h>
  
  }
  
  
  
  #define MAX_AUDIO_FRAME_SIZE    192000
  
  #define REFTIMES_PER_SEC      10000000  // 1 sec   100ns
  
  #define REFTIMES_PERMILLISEC     10000
  
  const CLSID CLSID_MMDeviceEnumerator = __uuidof(MMDeviceEnumerator);
  
  const IID   IID_IMMDeviceEnumerator  = __uuidof(IMMDeviceEnumerator);
  
  const IID   IID_IAudioClient         = __uuidof(IAudioClient);
  
  const IID   IID_IAudioRenderClient   = __uuidof(IAudioRenderClient);
  
  
  
  
  
  int alloc_samples_array_and_data(uint8_t*** data, int *linesize, int nb_channels,int nb_samples, enum AVSampleFormat sample_fmt, int align)
  
  {
  
      int nb_planes = av_sample_fmt_is_planar(sample_fmt) ? nb_channels : 1;
  
      *data = (uint8_t**)av_malloc(sizeof(*data) * nb_planes);
  
      return av_samples_alloc(*data, linesize, nb_channels,nb_samples, sample_fmt, align);
  
  }
  
  
  
  
  
  
  
  int main()
  
  {
  
      HRESULT hr;
  
      REFERENCE_TIME       buf_duration_request = REFTIMES_PER_SEC;
  
      REFERENCE_TIME       buf_duration_actual;
  
      IMMDeviceEnumerator  *pEnumerator      = NULL;
  
      IMMDevice            *pDevice          = NULL;
  
      IAudioClient         *pAudioClient     = NULL;
  
      IAudioRenderClient   *pRenderClient    = NULL;
  
      IAudioEndpointVolume *endpoint_vol     = NULL;
  
      WAVEFORMATEX         *fmt              = NULL;
  
      UINT32               frame_total       = 0;   // total frames in buffer
  
      UINT32               frame_avail       = 0;   // available frame number in buffer
  
      UINT32               frame_fill        = 0;   // filled frames
  
      BYTE                 *pData            = NULL;
  
      FILE                 *file_audio       = NULL;
  
      BYTE                 *buf              = NULL;
  
      DWORD                flags             = 0;
  
  
  
  
  
      CoInitializeEx(NULL,COINIT_MULTITHREADED);
  
      CoCreateInstance (CLSID_MMDeviceEnumerator,NULL,CLSCTX_ALL,IID_IMMDeviceEnumerator, (void**)&amp;pEnumerator);
  
      pEnumerator->GetDefaultAudioEndpoint(eRender, eConsole, &amp;pDevice);
  
      pDevice->Activate(IID_IAudioClient, CLSCTX_ALL, NULL, (void**)&amp;pAudioClient);
  
      pDevice->Activate(__uuidof(IAudioEndpointVolume),CLSCTX_ALL,NULL,(void**)&amp;endpoint_vol);
  
      pAudioClient->GetMixFormat((WAVEFORMATEX**)&amp;fmt);
  
  
  
  
  
      fmt->wFormatTag      = WAVE_FORMAT_PCM;
  
      fmt->nChannels       = 2 ; 
  
      fmt->nSamplesPerSec  = 44100;
  
      fmt->wBitsPerSample  = 16;
  
      fmt->nBlockAlign     = fmt->nChannels * (fmt->wBitsPerSample/8);
  
      fmt->nAvgBytesPerSec = fmt->nSamplesPerSec * fmt->nBlockAlign;  
  
      fmt->cbSize          = 0;
  
  
  
      WAVEFORMATEX* closest_format = NULL;
  
      hr = pAudioClient->IsFormatSupported(AUDCLNT_SHAREMODE_SHARED, fmt ,&amp;closest_format);
  
      hr = pAudioClient->Initialize(AUDCLNT_SHAREMODE_SHARED, NULL ,buf_duration_request,0, fmt ,NULL);
  
      hr = pAudioClient->GetBufferSize(&amp;frame_total);
  
      hr = pAudioClient->GetService(IID_IAudioRenderClient,(void**)&amp;pRenderClient);
  
  
  
      int file_size = 0;  
  
      int buf_size  = 44100 * 4;  
  
      int read_ret  = 0;
  
      int read_acc  = 0;
  
  
  
      fopen_s(&amp;file_audio,"E:\\MOVIE\\titan.wav","rb"); 
  
      fseek (file_audio , 0 , SEEK_END);
  
      file_size = ftell (file_audio);
  
      rewind (file_audio);
  
      buf  = (BYTE*) malloc(buf_size);
  
  
  
      pRenderClient->GetBuffer(frame_total,&amp;pData);
  
      read_ret  = fread(buf,1,buf_size,file_audio);
  
      read_acc += read_ret;
  
      memcpy(pData,buf,buf_size); 
  
      pRenderClient->ReleaseBuffer(frame_total,flags);
  
  
  
      buf_duration_actual = (double) REFTIMES_PER_SEC * frame_total / fmt->nSamplesPerSec;
  
      pAudioClient->Start();
  
  
  
  
  
      //-- resample
  
      uint8_t** src_data;
  
      uint8_t** dst_data;
  
      int       src_bufsize        = 0;
  
      int       dst_bufsize        = 0;
  
      int       src_linesize       = 0;
  
      int       dst_linesize       = 0;
  
      int       src_nb_channels    = 0;
  
      int       dst_nb_channels    = 0;
  
      int       src_nb_samples     = 1024;
  
      int       dst_nb_samples     = 0;
  
      int       max_dst_nb_samples = 0;
  
      int       ret = 0;
  
      SwrContext* swr_ctx = NULL;
  
      swr_ctx = swr_alloc();
  
      av_opt_set_int        (swr_ctx,"in_channel_layout" , AV_CH_LAYOUT_MONO,   0);
  
      av_opt_set_int        (swr_ctx,"in_sample_rate"    , 48000,0);
  
      av_opt_set_sample_fmt (swr_ctx,"in_sample_fmt"     , AV_SAMPLE_FMT_S32,   0);
  
      av_opt_set_int        (swr_ctx,"out_channel_layout", AV_CH_LAYOUT_STEREO, 0);
  
      av_opt_set_int        (swr_ctx,"out_sample_rate"   , 44100,0);
  
      av_opt_set_sample_fmt (swr_ctx,"out_sample_fmt"    , AV_SAMPLE_FMT_S16, 0);
  
      ret = swr_init(swr_ctx);
  
  
  
      src_nb_channels    = av_get_channel_layout_nb_channels(AV_CH_LAYOUT_MONO);
  
      src_nb_samples     = src_nb_channels * 48000;
  
      dst_nb_samples     = av_rescale_rnd(src_nb_samples, 44100,48000,AV_ROUND_UP);
  
      max_dst_nb_samples = dst_nb_samples;
  
      alloc_samples_array_and_data(&amp;src_data,&amp;src_linesize,src_nb_channels,src_nb_samples, AV_SAMPLE_FMT_S32,0);  
  
  
  
      dst_nb_channels    = av_get_channel_layout_nb_channels(AV_CH_LAYOUT_STEREO);
  
      alloc_samples_array_and_data(&amp;dst_data,&amp;dst_linesize,dst_nb_channels,dst_nb_samples, AV_SAMPLE_FMT_S16,0);
  
      //-- end resample
  
  
  
  
  
      while (read_acc &lt; file_size) 
  
      {
  
      Sleep(buf_duration_actual/REFTIMES_PERMILLISEC/2);          
  
          pAudioClient->GetCurrentPadding(&amp;frame_fill);
  
          frame_avail    = frame_total - frame_fill;
  
          dst_nb_samples = frame_avail;       
  
          src_nb_samples = av_rescale_rnd(dst_nb_samples,48000,44100,AV_ROUND_UP);
  
          src_bufsize    = av_samples_get_buffer_size(&amp;src_linesize,src_nb_channels,src_nb_samples,AV_SAMPLE_FMT_S32,1);
  
  
  
          cout&lt;&lt;"FILLED:"&lt;GetBuffer(frame_avail, &amp;pData );    
  
          read_ret    = fread(src_data[0],1,src_bufsize,file_audio);
  
          read_acc   += read_ret;
  
          ret         = swr_convert(swr_ctx,dst_data,dst_nb_samples,(const uint8_t**)src_data,src_nb_samples);
  
          dst_bufsize = av_samples_get_buffer_size(&amp;dst_linesize,dst_nb_channels,ret,AV_SAMPLE_FMT_S16,1);
  
          memcpy(pData, dst_data[0], dst_bufsize);        
  
          hr = pRenderClient->ReleaseBuffer(frame_avail,flags);                   
  
  
  
      }
  
  
  
  
  
      pAudioClient->Stop();
  
      CoTaskMemFree(fmt);
  
      pEnumerator->Release();
  
      pDevice->Release();
  
      pAudioClient->Release();
  
      pRenderClient->Release();
  
      CoUninitialize();
  
  
  
      return 0;
  
  }
  
  </iostream>

• Converting from wav alaw to unsigned 8bit PCM using ffmpeg

  27 juillet 2016, par Hanoun Fouad

  I am having some noise while i convert an ALAW wav file to u8, either to wav or pcm.

  I am using this command :

  ffmpeg -i input.wav -f alaw -acodec pcm_u8 output.pcm

  and

  ffmpeg -i input.wav -ar 8000 -acodec pcm_u8 -ac 1 output.wav

  I am having a very annoying noise which stay at low volume.