This episode covers another way to get dialogue audio (podcast or video) to appropriate loudness for a platform like YouTube (max -14 LUFS) without heavy-handed compression that starts sounding unnatural and "in your face." Previously we covered using a dialogue leveler; this time we look at a different plugin, Noiseworks' Dynassist ($150, no affiliation, still evaluating it myself).

Dynassist bundles a few useful features: a "vocal rider"-style leveler that targets a specific LUFS value directly (handy since that's exactly what platforms like YouTube specify); an AI-enabled noise gate/breath module that identifies and attenuates breaths between phrases by up to 12 dB so boosted breaths don't sound yucky; and a de-esser that can reduce sibilance by up to 6 dB.

Workflow demonstrated in DaVinci Resolve Fairlight: an RX mouth-declick plugin runs before Dynassist to clean up saliva/mouth-click noise, then Dynassist itself needs a "look-ahead" pass — you record/play through the audio once so it can analyze upcoming levels and write automation for level boosts, breath attenuation, and de-essing before actually processing in real time. Testing on a Sherlock Holmes passage targeting -19 LUFS, the plugin got very close to target, retained 3.3 dB of true-peak headroom (important since export compression/codecs eat into headroom — always keep at least 1.5-2 dB), handled most breaths well (it missed one, easily fixed by manually editing an automation point), and did a nice job managing sibilance.

Overall verdict: impressively convenient for hitting a loudness target while managing breaths and sibilance in one pass, worth considering the $150 price if this workflow fits your needs. Get out there, make some great sound, and we'll talk to you again soon.