Creating high-quality audiobooks demands more than a pleasant voice and good narration skills. One of the most persistent challenges narrators face is minimizing mouth noise and breath sounds that can distract listeners, break immersion, and signal amateur production. These subtle clicks, smacks, and sharp intakes of air accumulate over hours of listening, turning an otherwise engaging performance into an exhausting experience. By combining disciplined recording techniques with precise editing practices, you can achieve a clean, professional sound that keeps your audience focused on the story. This guide walks through every stage—from pre-recording preparation to final mastering—so you can produce audiobooks that sound polished, consistent, and free from distracting oral noises.

Preparing Your Recording Space and Gear

The foundation of a quiet recording lies in your environment and equipment. Even the best editing cannot fully rescue a track spoiled by echo, background hum, or poorly captured mouth sounds. Invest time in setting up a space that minimises acoustic problems before you hit record.

Acoustic Treatment

Hard surfaces like bare walls, windows, and floors create reflections that smear the audio and make mouth clicks more prominent. Use materials that absorb sound, such as acoustic foam panels, dense blankets, or even a portable vocal booth. Place absorbent panels at ear height and behind your recording position. A reflection filter around the microphone also helps, but it should be paired with room treatment rather than relied upon alone.

Microphone Selection and Placement

A cardioid or supercardioid condenser microphone works well for narration because it rejects sound from the sides and rear. However, cheaper condenser mics can exaggerate sibilance and mouth noise. Dynamic microphones, like the Shure SM7B or Electro-Voice RE20, are less sensitive to high-frequency clicks and breaths, making them popular choices for voice-over work. Regardless of type, position the microphone about six to twelve inches from your mouth, slightly off-axis (not directly inline with your lips). This angle reduces plosives and the pickup of direct mouth noises without sacrificing clarity.

Pop Filter and Windscreen

A pop filter is essential for minimising plosive bursts (P and B sounds) that can cause low-frequency thumps in the recording. Use a double-layer fabric pop filter mounted on a flexible arm. For additional protection against breaths, add a foam windscreen over the mic capsule. Some narrators combine both, but be careful not to position the pop filter too close to the microphone—leave a gap of two to four inches to allow air to dissipate.

Maintaining Consistent Mic Technique

Once your setup is in place, rehearse maintaining constant distance and angle throughout each recording session. Even small movements change the tonal balance and make mouth noises more variable, complicating editing. Mark your standing position on the floor with tape and use a microphone boom arm that stays firmly in place. Keep your head steady, and avoid turning your face away from the mic while reading.

Mastering Breathing and Mouth Noise at the Source

Nothing replaces good vocal hygiene and technique. Editing can clean up a recording, but preventing noises before they happen saves hours of post-production work.

Hydration and Diet

Dry mouth tissues produce more audible clicks and lip smacks. Drink water at room temperature throughout your recording session—avoid cold water, which can cause throat tension. Herbal teas with honey are also beneficial. Avoid dairy products, caffeine, and sugary drinks before recording, as they can thicken saliva or cause sticky mouths. Chewing a green apple or using a mouth moisturizer designed for singers can help reduce friction noises.

Vocal Warm-Ups and Breath Control

Warm up your voice with gentle humming, lip trills, and tongue twisters for five to ten minutes before recording. This relaxes the jaw and tongue, reducing involuntary clicks. Practice diaphragmatic breathing—breathing from your belly rather than your chest—to make breaths quieter and more controlled. Inhale through your nose and exhale softly through your mouth, aiming for a silent air exchange during natural pauses in the text.

Pausing Naturally

Many mouth noises occur when narrators try to squeeze in a breath while continuing to speak. Instead, plan natural pauses at punctuation marks and phrase boundaries. Take a silent breath during these pauses, allowing the microphone to capture less direct air flow. Mark your script with breath indicators (like a slash) so you know exactly where to inhale without rushing.

Recording Techniques That Reduce Unwanted Sounds

Even with perfect preparation, some mouth noise will appear. The following techniques capture cleaner audio, making later editing lighter.

Off-Axis Placement and Angle

Speaking directly into the center of the microphone capsule picks up the most direct sound, including saliva clicks and breath blasts. Instead, position the microphone so that you speak past it, not into it—rotate the mic 15–30 degrees to the side. This off-axis placement reduces the intensity of plosives and sibilance while retaining vocal warmth. Experiment with the angle using a test recording, then adjust until you find the sweet spot.

Use a Breath Screen or Mesh

Some narrators use a small mesh screen placed close to their mouth, separate from the pop filter. This acts as a baffle that diffuses the rush of air during exhalation. You can improvise with a fine metal mesh mounted on a gooseneck stand. This trick significantly lowers breath volume without muffling the voice.

Record in Short Segments

Long takes increase fatigue, which leads to heavier breathing, mouth dryness, and more inadvertent noises. Record in passages of five to ten minutes, then pause to drink water and reset your posture. This approach also makes editing easier because you have discrete files to process. Use a DAW that supports regions or markers, such as Audacity or Adobe Audition, to label each segment for quick navigation.

Monitor Levels and Background Noise

Keep your input level moderate—aim for peaks around -6 dB to -3 dB with an average around -12 dB. Too much gain introduces noise floor hiss, while too little forces you to boost in post, amplifying mouth clicks. If your environment has low-frequency rumble (HVAC, traffic), use a high-pass filter at around 80 Hz during recording (if your interface or software allows) to cut that rumble without affecting voice quality.

Editing Workflow to Remove Mouth Noise and Breaths

Once you have recorded clean takes, the editing phase refines them into a seamless audiobook. A systematic workflow ensures you do not miss problematic spots.

Step 1: Apply Noise Reduction

Start by capturing a noise print of your room silence (a few seconds of quiet with no speaking). Use your DAW’s noise reduction tool to sample that print and apply gentle reduction across the entire track—typically 12–18 dB reduction with a moderate frequency smoothing. Avoid aggressive settings, as they can create artifacts that sound worse than the noise. This step removes consistent background hum but will not fix mouth clicks.

Step 2: Spectral Editing for Clicks and Pops

Mouth clicks show up in the spectrogram as short, bright vertical lines or dots, often between 2 kHz and 8 kHz. Use spectral editing tools (available in Audition as the Spectral Frequency Display, or in Audacity via the Spectrogram view) to manually select and delete these events. Zoom in tightly, draw a selection around the click, and press delete. For very small clicks, use a noise gate or DeClicker plugin (like Waves WLM or the free iZotope RX Elements) to automate removal. However, manual spectral editing yields the most transparent results, preserving the natural voice.

Step 3: Breath Reduction

Breaths are not inherently bad—some even add realism and pacing. But loud, sharp breaths are distracting. Apply a compressor or expander with a fast attack and release to reduce the volume of breaths without silencing them entirely. In Audition, use the “Essential Sound” panel and set the “Voice” mode to “Clean Up” for built-in breath control. Alternatively, create a custom preset: compress with a ratio of 3:1, threshold around -20 dB, attack 1 ms, release 50 ms. Then go through the track and manually lower the gain on any remaining inhales that spike above the surrounding speech. The goal is to make breaths feel natural but not intrusive.

Step 4: Manual Editing of Problematic Areas

For stubborn clicks, lip smacks, or tongue pops, zoom into the waveform and use a subtle fade or cut. With the waveform expanded, locate the transient click—it often appears as a jagged spike. Highlight a few milliseconds around the spike, apply a crossfade, or lower the gain by 6–12 dB. For breaths between sentences, select the breath region and reduce its volume to around 50% of the sentence level. Avoid completely deleting all breaths, as absolute silence between words sounds unnatural and robotic.

Step 5: Use De-Esser for Sibilance

Sibilance (sharp S and T sounds) can also be mistaken for mouth noise. A de-esser plugin detects high-frequency energy and tames it. Apply a de-esser set to around 5–8 kHz with moderate reduction (2–4 dB). This reduces the harshness without draining the voice of presence.

Advanced Editing Tips for a Professional Finish

Beyond the basic workflow, these advanced techniques help you achieve studio-quality sound.

Compression and Equalization

Use a gentle compressor to even out volume variations between soft and loud passages. This also helps bring down the volume of occasional heavy breaths without manual gain riding. Set a ratio of 2:1 to 3:1, threshold around -18 dB, attack 10 ms, release 100 ms. Follow the compressor with a high-pass filter at 80 Hz to remove low-end rumble, and a slight high-shelf boost above 10 kHz if the voice sounds dull. Avoid excessive EQ, as boosting can amplify mouth noises.

Noise Gate Best Practices

A noise gate mutes audio below a set threshold, useful for removing background noise in pauses. But if set too aggressively, it cuts off the tail of words and breaths, causing clipping. Use a gate with a fast attack (1 ms) and a moderate release (100 ms) and threshold set just above the noise floor. For audiobooks, a gate is best used only on sections with long silences; for continuous narration, manual editing is cleaner.

Automation for Dynamic Control

Write volume automation to manually smooth out problematic spots. Listen through the entire track at low volume to identify spots where mouth noise or breath spikes break the flow. On those sections, draw automation nodes to lower the gain by 3–8 dB over a few hundred milliseconds. This is tedious but yields a polished result that automated plugins cannot match.

Normalization and Loudness Standards

After editing, normalize your track to meet audiobook platform standards. Most distributors (ACX, Findaway Voices) require RMS levels around -20 dB to -25 dB, with peak levels below -3 dB. Use a loudness meter plugin (like Youlean Loudness Meter) to measure integrated loudness and true peak. Adjust gain or apply a limiter sparingly to hit the target. Then export in mono 16-bit 44.1 kHz WAV for distribution.

Common Pitfalls and How to Avoid Them

Even experienced narrators fall into traps. Here are the most frequent mistakes and workarounds.

Over-Editing: Removing Too Much Breath

Deleting every breath leaves gaps that sound abrupt and unnatural. The ear expects a subtle rise in noise before a word. Instead of removing breaths, reduce their volume to around 40–50% of the speech level. This preserves flow while eliminating distraction.

Too Much Noise Reduction

Applying heavy noise reduction can cause a watery, echoing effect called “swishing” or “artifact noise.” Always apply the lightest reduction that still cleans up the track. If you need more, combine gentle reduction with manual editing rather than increasing the reduction amount.

Inconsistent Volume Between Segments

Recording across multiple days or at different times can produce volume and tonal inconsistencies. Use automation to match levels when splicing takes. Also, re-eq each segment to match a reference track from your first recording session. A consistent headphone reference helps, but trust your ears and a waveform overlay.

Skipping the Spectrogram

Many editors rely solely on waveform view, which hides clicks and pops. Always use spectrogram view when editing for mouth noise—it reveals high-frequency disturbances that waveforms miss. With practice, you can spot a click as easily as a misspelled word.

Final Checks and Mastering

Before export, run a final QC pass with fresh ears. Listen to a few minutes of the book at a normal listening volume, then at a very low volume (where mouth clicks become more obvious). Check for abrupt cuts, missing breaths, or unnatural transitions. Use a crossfade of 10–20 ms between spliced regions to avoid clicks at edit points. Finally, apply a limiter with a ceiling of -3 dB to catch any stray peaks, and export your master.

By methodically addressing mouth noise and breath sounds from the recording stage through to final mastering, you remove the distractions that undermine listener engagement. This attention to detail not only enhances the storytelling but also establishes you as a reliable, professional narrator. With practice, these techniques become second nature, and your audiobooks will consistently earn praise for their clean, immersive sound.