How to Even Out a Quiet Podcast Guest and Fix Interview Echo

When Listeners complain that they have to turn their volume up whenever your guest talks and turn it back down when you reply, reaching for a standard volume booster or limiter makes things worse. Raising the gain of the whole file pushes the louder host straight into distortion while the guest is still too quiet.

A single-threshold compressor has the opposite trap: if you set the threshold low enough to catch a guest speaking 14 dB below the host, you either squash the host's voice flat or pull up every breath, chair creak, and air-conditioner hum during the pauses.

Audo's Balance audio (/audio-leveler) solves this by separating who speaks when:

• On a single mixed file (one MP3, WAV, or MP4). Balance audio detects each speaker's turns, measures each person's loudness independently, and brings every voice to the same target (−16 LUFS for podcasts by default) while leaving silent pauses untouched. Switching on 'Even out each voice too' also steadies a speaker who leans toward and away from their mic.
• On separate host and guest tracks (multitrack). Drop both tracks into Balance audio at once. It lines them up by their shared acoustic bleed so they stay in sync, matches both speakers' loudness, and gives back the balanced mix (or separate leveled tracks).

During a live call, 'echo cancellation' in Zoom, Teams, Meet, or Riverside solves one specific problem: acoustic feedback, where your voice comes out of your guest's laptop speakers and re-enters their microphone a split second later.

It does nothing for room reverb—the hollow, boxy ring caused by your guest's own voice bouncing off bare drywall, glass windows, and a hard desk into their microphone. Because those reflections arrive within milliseconds of every syllable, removing them requires speech-aware post-processing.

After the call, drop the recording into Audo's Echo remover (/remove-echo) or Audio enhancer (/audio-enhancer). Both strip the room reflections while keeping the speaker's voice close and clear, and the strength slider lets you leave 10–20% of the room in so the voice never sounds unnaturally dead.

If your recording has both a reverberant room and a quiet guest, run the two steps in this order (or chain them in one go inside Audo's browser audio editor with 'Open in editor'):

• Step 1: Strip echo and background noise first. Run Echo remover (/remove-echo) or Audio enhancer (/audio-enhancer) before leveling. That way the leveler measures dry speech rather than boosting the tail of the room echo.
• Step 2: Balance the speakers. Click 'Open in editor' on the cleaned result and run Level voices (or open /audio-leveler directly) to bring the host and guest to −16 LUFS.
• Step 3: Shorten long pauses or cut filler words (optional). In the same browser editor, shorten dead air or remove 'um' and 'uh' before downloading your finished MP3 or WAV.

Audo's Balance audio and Echo remover are built to rescue mixed recordings and two-track interviews in seconds without manual automation lanes.

If you recorded 3 or more separate local ISO tracks in Riverside or a field recorder, have frequent overlapping crosstalk where two people laugh or argue at the exact same moment, and already mix in Reaper, Logic Pro, or Audition, editing the separate tracks in a multitrack DAW gives you manual clip-gain and mute control over simultaneous speech.