stageResolvedAudioToSandbox function

Future<AudioMixPlan> stageResolvedAudioToSandbox({ required List<ResolvedAudioTrack> tracks, required RenderSandbox sandbox, required AudioByteLoader loadBytes, double masterVolume = 1, })

Stages every resolved audio tracks into sandbox and builds the encoder AudioMixPlan, without touching the file system.

Each track's bytes (fetched through loadBytes) are written under the bare name audio_<i>_<cacheKey> the encoder -is, then turned into one AudioTrackNode via AudioTrackNode.fromResolved (so delay, trim, gain, fade, and loop match every other backend). The cache key comes from audioSourceFromString, so an in-browser render and a desktop render name the same track identically and the encode args stay byte-identical. masterVolume scales the final mix. An empty list stages nothing and returns an empty plan, so the encoder's -an path is unchanged.

Implementation

Future<AudioMixPlan> stageResolvedAudioToSandbox({
  required List<ResolvedAudioTrack> tracks,
  required RenderSandbox sandbox,
  required AudioByteLoader loadBytes,
  double masterVolume = 1,
}) async {
  final nodes = <AudioTrackNode>[];
  for (var i = 0; i < tracks.length; i++) {
    final track = tracks[i];
    final name = 'audio_${i}_${audioSourceFromString(track.source).cacheKey}';
    final bytes = await loadBytes(track.source);
    await sandbox.writeBytes(name, bytes);
    nodes.add(AudioTrackNode.fromResolved(track, name: name));
  }
  return buildAudioMixPlan(nodes, masterVolume: masterVolume);
}