Podcast loudness: Apple -16 LKFS, Spotify -14 LUFS, and Sume

Apple recommends podcast loudness around -16 dB LKFS with a tolerance of 1 dB and a true peak no higher than -1 dB FS (read 2026-10-02). Spotify's page describes -14 LUFS as its Normal playback level, but it is written for music, so it does not set a podcast target. Sume does not measure or normalize loudness: its docs offer a static gain and music ducking, not a loudness target, so hit the level in an audio editor and bring the file in.

The two numbers, side by side

The pages answer different questions. Apple states a recommendation for podcast audio. Spotify states how its player normalizes, plus mastering advice for tracks.

What Sume has for audio level

In the docs I checked, the level controls are on Timeline 1.0. They are fixed settings, not measurements.

• audio.gain_db is a static gain from -60 to 12 dB on the spine.
• soundtrack.duck_db (0 to 20) lowers a music bed under a real voice spine.
• There is no integrated-loudness meter, no true-peak limiter and no target-LUFS field in these docs.

A practical flow

Static gain cannot reach a target on its own, because you must first know the current level. Measure and master the episode in your audio tool, export it, and import the file into your workspace. After that, use Sume for the parts it does: audio detach to pull the track from a video, or a Timeline render to join it with video.

If you render a timeline with a music bed, keep the voice as the spine and set duck_db, then re-check the final mix in your meter, because ducking changes the overall level too. Whether platforms re-normalize your upload is theirs to decide; Spotify's page says normalization is applied during playback only.

What to take from this

Choose the target from the destination: Apple's recommendation for Apple Podcasts, and your own judgment for Spotify given its page does not cover podcasts. Sume will not tell you whether a file hit the target, so keep the measurement in your editor.