Loudness Optimize

When should you choose this tool?

In a conversation, a distant speaker may be barely audible while the nearby voice is already loud. Uniform gain raises both. Loudness Optimize changes the balance of quieter passages, using measurements to guide settings.

How it changes audio

It uses upward compression on quieter audio and limits peaks. Full-clip integrated LUFS and its measurement gate are distinct from the processing threshold in dBFS. The graph shows expected output for its labelled scope.

What do the parameters control?

Quiet boost; Threshold

Boost spans 0–40 dB and threshold −50–−3 dBFS. The threshold governs processing; it cannot reliably classify speech versus noise.

Ratio; Speed

Ratio spans 1:1–20:1 and speed 10–1000 ms. These control the strength and tracking of level changes. Fast, strong processing can make the background pump.

Noise-floor protection; Peak ceiling

Advanced settings shape boosting of very quiet audio and sample-peak limiting. Noise-floor protection is not a speech-detecting gate.

Load measured suggestion

Loads a starting point based on full-clip measurements. Review it carefully with high noise or speakers at different distances.

Try a short passage

1. Wait for full-clip analysis and select a passage containing both quiet and loud speech.
2. Load the measured suggestion or begin with moderate boost. During Preview, check quiet sentences and noise in pauses together.
3. Refine threshold and timing. Ease the boost if noise breathes or pumps, then review the whole conversation after applying.

What to listen for

Quiet speech should move closer to loud speech without disproportionately raising noisy pauses. Speaker changes should not produce abrupt level jumps.

Limits of the method

Measurements describe signal level, not word intelligibility. No universal setting lifts every distant speaker without also raising background.

Preview covers a short passage. After applying, review the full treated section; you can undo the step from history.