Recherche avancée

Sur d’autres sites (16689)

• Distorted audio output from microphone input

  25 août 2016, par Mohammad Abu Musa

  I am writing an ogg encoder with chrome native client, I managed to successfully export the files as the wave samples come in put I have two issues

  1. Timing in the file
  2. The voice is distorted, I hear noises instead of the voice collected from the microphone

  Here how the code works, first I collect the microphone samples and process it as the following

  const char* data = static_cast<const>(buffer.GetDataBuffer());
  
  audio_buffer_size = buffer.GetDataBufferSize();
  
  uint32_t channels = buffer.GetNumberOfChannels();
  
  uint32_t samples = buffer.GetNumberOfSamples() / channels;
  
  
  
  if (channel_count_ != channels || sample_count_ != samples) {
  
    channel_count_ = channels;
  
    sample_count_ = samples;
  
  
  
    samples_.resize(sample_count_ * channel_count_);
  
    // Try (+ 5) to ensure that we pick up a new set of samples between each
  
    // timer-generated repaint.
  
    timer_interval_ = (sample_count_ * 1000) / buffer.GetSampleRate() + 5;
  
    // Start the timer for the first buffer.
  
    if (first_buffer_) {
  
      first_buffer_ = false;
  
      ScheduleNextTimer();
  
    }
  
  }
  
  
  
  memcpy(samples_.data(), data,
  
      sample_count_ * channel_count_ * sizeof(char));
  
  </const>

  The result is a vector called samples_ of type char for the other information they are as follows

  1. Samples 480
  2. Channels 1
  3. Sample Rate 48000
  4. Time Interval is 15

  On the function ScheduleNextTimer I make the encoding and the storage of the packets to ogg file. I used the standard code provided in the encoder example but I do not think I am doing write since I am not importing from a wave file, I am rather collecting raw bytes so I do not think the loops I am using are correct.
  Attempt #1

  for (int i = 0; i < sampread; i++) {
  
      for (int j = 0; j < audioCh; j++) {
  
          vorbis_buffer[j][i] = ((samples_[count + 1] << 8)
  
                  | (0x00ff & (int) samples_[count])) / 32768.f;
  
          count += 2;
  
      }
  
  }

  Attempt #2

   for(int i=0;i4;i++){
  
            vorbis_buffer[0][i]=((samples_[i*4+1]<<8)|
  
                        (0x00ff&(int)samples_[i*4]))/32768.f;
  
            vorbis_buffer[1][i]=((samples_[i*4+3]<<8)|
  
                        (0x00ff&(int)samples_[i*4+2]))/32768.f;
  
        }

  Both of these attempts did not work. My question is should I keep using the vector or should I change it to something else. what are these values in the loop I do not get them or what they represent and how does the timing work.

  Here is what I get from ffprob

  Input #0, ogg, from 'file.ogg':
  
    Duration: 00:01:21.00, start: 0.000000, bitrate: 118 kb/s
  
      Stream #0:0: Audio: vorbis, 44100 Hz, stereo, fltp, 80 kb/s

• How to run the "ffmpeg -i input.mp4 output.avi" in Node.js ? [duplicate]

  3 septembre 2016, par Thales

  This question already has an answer here :

  • Execute a command line binary with Node.js
    
    6 answers
    

  I have a application on Electron that needs receive video and convert him another format, but I don’t know how to do. I’m thinking to do with node, but I need how to run the command "ffmpeg -i input.mp4 output.avi". I don’t know if this way is better. I thank first of all and I apologize for the ignorance.

• Use Output of FFMPEG as the Input of another FFMPEG command, in one execution

  1er juillet 2022, par Xavi Font

  I asked a question earlier but had too many steps and was too long maybe to receive an answer, so here is a simple one that will help me A LOT.

  


  I have a command, could be anything ffmpeg -loop 1 -i my_dog_playing.jpg -t 5 output.mp4 and want to take that output and add stuff to it, like a complex_filter (say... an image overlay on the corner), but I want to do so before that output.mp4 is encoded (rendered) so that no time is wasted (I want it to be done within one command).