Keep the source's channels and sample rate when extracting audio

Send only video_url to POST /v1/audio-detach and the result keeps the source's channel layout and sample rate: channels defaults to source, and an omitted sample_rate is inherited (returned as null). The default format is sample-exact wav (pcm_s16le).

Adobe's Premiere 26.5 notes say Match Source now "includes your source's audio channels and sample rate, with automatic downmixing and warnings." This post covers what Sume does for one extraction, from the Audio detach docs, read 2026-09-30.

What are the knobs?

channels is source or mono. sample_rate is 16000, 44100 or 48000. format is wav or mp3 (128 kbps). range is an optional { start, end? }. The pairing 16000 plus mono is the speech-to-text shape.

curl -X POST https://api.sume.com/v1/audio-detach \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: detach-match-source-001" \
  -d '{
    "video_url": "https://media.sume.com/artifacts/artf_demo/talk.mp4"
  }'

How do I see what I got?

The result is kind: audio_detach with audio_url, duration_seconds, format, channels, sample_rate, source_duration_seconds and optional range and warnings[]. sample_rate is null when you did not set one, which means it was inherited.

Does Sume downmix with warnings like Premiere?

The docs describe mono as an explicit choice and do not describe automatic downmixing or a downmix warning. If you want a mono file, ask for it with channels: "mono". A silent source fails detach_source_has_no_audio.