Issue on docs: Voice Isolation Response

[overfit]

Path: /api-reference/audio-isolation

Hi, can someone clarify how to extract audio from the audio-isolation response? I first tried to load it to JSON given the response example in the documentation, i.e.

{
  "audio": {
    "audio_isolation_id": "<string>",
    "created_at_unix": 123
  },
  "waveform_base_64": "<string>"
}

But it looks like that the response is plain binary. Then I tried to decode the binary response and it didn't work out.

Generally, more detailed instructions would be really helpful.