Audio stream was preserved as audio.wav, but no local Whisper/faster-whisper command was available in this environment during capture; article uses the Instagram caption, visible on-screen text, frame analysis and source records rather than a speech transcript.