BaseAudioContext: метод createChannelSplitter()

createChannelSplitter(numberOfOutputs)

const ac = new AudioContext();
ac.decodeAudioData(someStereoBuffer, (data) => {
  const source = ac.createBufferSource();
  source.buffer = data;
  const splitter = ac.createChannelSplitter(2);
  source.connect(splitter);
  const merger = ac.createChannelMerger(2);

  // Reduce the volume of the left channel only
  const gainNode = ac.createGain();
  gainNode.gain.setValueAtTime(0.5, ac.currentTime);
  splitter.connect(gainNode, 0);

  // Connect the splitter back to the second input of the merger: we
  // effectively swap the channels, here, reversing the stereo image.
  gainNode.connect(merger, 0, 1);
  splitter.connect(merger, 1, 0);

  const dest = ac.createMediaStreamDestination();

  // Because we have used a ChannelMergerNode, we now have a stereo
  // MediaStream we can use to pipe the Web Audio graph to WebRTC,
  // MediaRecorder, etc.
  merger.connect(dest);
});