Professional Voiceover Recording Cleanup

Voiceover quality standards: what clients and platforms expect

Professional voiceover delivery has specific technical requirements that vary by platform. ACX (Amazon's audiobook platform) requires a noise floor below -60dB, peaks between -3dB and -6dB, and RMS (average loudness) between -18dB and -23dB. Broadcast standards (for TV and radio commercials) typically require a noise floor below -60dB with no audible artifacts. E-learning platforms and corporate narration clients generally accept slightly looser specs but still expect clean, noise-free audio without audible room tone.

Meeting these standards from a home studio is the central challenge for voiceover artists. Professional studios invest tens of thousands of dollars in acoustic treatment, high-end microphones, and quiet HVAC systems. Home studios typically have untreated rooms, consumer-grade microphones, and noise sources that can't be fully eliminated. Post-processing bridges the gap — cleaning up what the room and equipment introduced.

Common home studio problems for voiceover

• ●Room tone — the ambient sound of your recording space, including HVAC, distant traffic, and building vibration. Professional studios have near-silent room tone; home studios always have some. The question is whether it's audible in the recording.
• ●Mouth clicks and lip smacks — close-mic voiceover recording picks up the small wet sounds of mouth movements between words. These are largely inaudible in conversation but extremely noticeable in isolated narration where every sound is exposed. Hydration helps; some professionals use a thin layer of green apple or olive oil on the lips.
• ●Breath noise — audible inhales and exhales between phrases. Some breath sounds are natural and expected in narration; heavy, sharp breaths are distracting. The goal is reduction, not elimination — completely removing all breath sounds makes narration sound robotic.
• ●Plosive pops — explosive bursts of air from 'p' and 'b' sounds hitting the microphone diaphragm directly. Pop filters (mesh screens positioned between mouth and mic) prevent most plosives. Existing plosives can be reduced with a high-pass filter but not fully eliminated.
• ●Sibilance — harsh, piercing 's' and 'sh' sounds caused by microphone proximity and bright frequency response. A de-esser (specialized dynamic EQ targeting 5-9kHz) is the standard fix.

The voiceover cleanup chain

1. 1Noise removal first — remove background room tone, HVAC hum, and ambient noise. This is where AI tools like Voice Isolate are most useful: upload the raw recording and remove environmental noise in a single pass without affecting voice quality.
2. 2De-breath — reduce (not eliminate) audible breath sounds between phrases. Dedicated tools like iZotope RX Breath Control automate this; manually, you can lower the volume of breath sections by 6-10dB in your DAW.
3. 3De-click — remove mouth clicks and lip smacks. iZotope RX Mouth De-click is purpose-built for this. In Audacity, the Click Removal effect handles some clicks but is less effective on mouth sounds specifically.
4. 4De-ess — tame harsh sibilance. Apply a de-esser targeting the 5-9kHz range with moderate reduction. Over-de-essing makes 's' sounds like 'th' — use just enough to remove the harshness.
5. 5Compression and normalization — apply gentle compression (2:1 ratio, slow attack) to even out volume, then normalize to your delivery spec (e.g., -20dB RMS for ACX).

When to re-record vs when to clean up

Re-record when: there's heavy clipping or distortion (unrepairable), severe room echo (difficult to fully remove and sounds unnatural when aggressively processed), significant background noise events you can't mask (phone ringing, door slamming during a critical phrase), or consistent mouth/throat issues throughout the recording (dry mouth, congestion) that affect every take.

Clean up when: the performance is good but the room has audible ambient noise (AI tools handle this well), there are occasional mouth clicks or pops (fixable with de-click tools), breath sounds are slightly too prominent (easily reduced), or the noise floor is slightly above spec (noise reduction brings it into compliance). Time spent on cleanup is justified when re-recording means re-scheduling studio time, re-booking talent, or losing a good performance.