Dictation portals are quietly upgrading their audio capture

Seeing more EHR dictation portals quietly switch to the Opus codec lately, and it’s having real effects on transcripts. In a quick comparison last week in a noisy clinic hallway with a small gooseneck mic, consonants came through cleaner and low-voice clinicians were easier to parse without cranking gain.

If your audio suddenly turns phasey or watery after an update, disable aggressive noise reduction in the OS, avoid virtual sound devices that resample, and keep a wired cardioid mic close and slightly off-axis from the mouth. Stable input gain and minimal processing let the portal capture the detail it’s now capable of.

Opus helps, but I’d pair it with a quick, repeatable workflow: 1) ensure mono capture and consistent mic distance. 2) enable diarization and verify speaker tags. 3) run macros with PHI-safe templates only. 4) proof in two passes—content, then PHI. 5) finalize with timestamped audit. Cleaner audio plus disciplined checks reduces rework and HIPAA risk.