In film, television, and video production, dialogue is the backbone of storytelling. Clean, intelligible dialogue tracks are essential for maintaining audience immersion and a professional sound. Unfortunately, unwanted mouth noises—such as smacks, clicks, pops, and wet swallowing sounds—frequently degrade the clarity of dialogue. These artifacts can distract listeners, break suspension of disbelief, and signal an amateur production. While prevention during recording is the first line of defense, post-production offers a powerful toolkit for removing or minimizing these noises. This article explores the nature of mouth noises, presents a range of proven techniques for elimination, outlines a practical workflow, and recommends best practices for both recording and editing to achieve pristine dialogue tracks.

Understanding Mouth Noises

Mouth noises encompass a variety of transient sounds produced by the vocal tract during speech. They can originate from the lips, tongue, cheeks, and throat. Common types include:

  • Lip smacks – sharp, short sounds caused by lips parting quickly.
  • Tongue clicks – produced when the tongue releases from the roof of the mouth.
  • Saliva pops – moist, percussive noises from accumulated saliva between the cheeks and gums.
  • Wet mouth sounds – slushy, resonant noises often heard on close-miked voiceovers.
  • Breathy pops and plosives – explosive sounds made by sudden air puffs against the microphone capsule.
  • Throat clearing or swallowing – low-frequency rumbles and clicks from the larynx.

These sounds occur naturally due to dry mouth, nervousness, poor mic technique, or even the actor’s natural physiology. They are most audible in quiet sections of a track or when the microphone is placed very close to the mouth. Recognizing each type is the first step in choosing the most effective removal technique.

Essential Techniques for Removing Mouth Noises

Modern digital audio workstations (DAWs) and dedicated plugins provide multiple strategies for tackling mouth noises. The best approach often combines several methods, depending on the severity and context of the noise.

1. Spectral Editing with iZotope RX

Spectral editing is one of the most powerful methods for removing mouth noises without affecting the underlying dialogue. In tools like iZotope RX, audio is displayed in a visual frequency/time graph. Mouth noises typically appear as short, vertical lines (transients) in the upper frequencies (2–10 kHz) or as blurry smudges in the mid-range. By zooming into the spectral view, you can select the offending noise with a marquee tool and either reduce its gain, mute it, or apply a spectral repair algorithm such as “Replace” or “Attenuate.” The advanced algorithms in RX, especially the Mouth De-click and Spectral De-noise modules, can automatically detect these transients and reduce them with precision. A dedicated tutorial from iZotope explains the step-by-step process for dialogue cleanup, making it accessible for editors of all skill levels.

2. Dedicated De-Click and De-Mouth Plugins

Specialized plugins are designed explicitly for removing clicks, pops, and mouth smacks. Waves X-Click and iZotope RX’s De-click module both analyze the audio for transient spikes and apply targeted attenuation. The user can adjust the sensitivity (threshold) and the strength of the reduction. For mouth noises that are repetitive but subtle, automatic de-click plugins can save hours of manual editing. However, care must be taken not to over-process, which can introduce artifacts or dull the clarity of consonants. Many professionals recommend a two-pass approach: first, an automatic de-click with conservative settings, followed by manual spectral editing of remaining problematic noises.

3. Manual Volume Automation and Clip Gain

For minor mouth smacks that occur between words, manual volume automation is a simple, non-destructive technique. Within a DAW (such as Pro Tools, Logic Pro, or Cubase), draw volume automation points to lower the gain during the noise and restore it immediately after. This method works best when the noise is isolated in a quiet gap. Alternatively, using clip gain (gain applied to an entire audio clip) in conjunction with crossfades can attenuate the noise while keeping the natural dynamics of the surrounding speech. Manual automation demands careful listening and a steady hand, but it preserves the original audio quality perfectly because no filtering or processing is applied.

4. Advanced Spectral Repair Tools

Beyond basic spectral editing, some tools offer even more sophisticated repair options. For instance, iZotope RX’s Spectral Repair module includes algorithms like “Replace,” which reconstructs the missing audio by interpolating from surrounding frequencies, and “Attenuate,” which reduces the noise while preserving the underlying voice. These are particularly useful when a mouth noise overlaps with a spoken syllable. Similarly, Adobe Audition’s Spot Healing Brush (found in the spectral frequency display) allows you to “paint over” a noise, and the software will intelligently fill the gap. Such tools require practice to avoid creating unnatural-sounding results, but they are indispensable for high-end dialogue restoration.

5. EQ and High-Pass Filtering (with Caution)

Some mouth noises, especially low-frequency pops and rumbles, can be reduced using equalization. A high-pass filter set between 60 and 80 Hz can remove subsonic rumbles from throat clearing or breath pops. A notch filter tuned to a specific resonant frequency (e.g., 200–400 Hz for certain wet sounds) may also help. However, EQ should be used sparingly because it affects the entire audio signal—cutting too much can make the voice sound thin or unnatural. The best strategy is to combine gentle high-pass filtering with more targeted spectral tools for higher-frequency smacks.

Step-by-Step Workflow Example

To illustrate how these techniques come together, consider a typical dialogue clip with moderate mouth click problems:

  1. Listen and assess – Play the clip at normal speed and then loop sections with noticeable noises. Identify the type and location of each noise.
  2. Apply automatic de-click – Run the Mouth De-click module in iZotope RX with a moderate threshold. Listen to the processed clip to confirm no artifacts are introduced.
  3. Manual spectral cleanup – Open the spectral view and zoom into each remaining noise. Use the “Attenuate” or “Replace” spectral repair function for noise overlapping speech; use the “Gain” function to reduce isolated noises in silence.
  4. Check for volume mismatches – After removal, listen for any unnatural dips in volume. Use volume automation to smooth out transitions.
  5. Final pass with EQ – If the clip still has low-frequency rumble, apply a high-pass filter at 60 Hz. Avoid additional processing unless absolutely necessary.
  6. Reference the original – A/B compare the processed and unprocessed clip to ensure quality has not degraded.

Preventive Measures During Recording

Preventing mouth noises at the source is always preferable to fixing them in post. Implementing good recording practices can dramatically reduce the need for time-consuming editing.

Hydration and Vocal Care

Dry mouth often leads to lip smacks and tongue clicks. Encourage talent to drink room-temperature water frequently and avoid caffeine or dairy products before recording. Using a mild lubricant like a small piece of apple or gum can help maintain moisture. Good vocal warm-ups also reduce unnecessary throat noises.

Microphone Selection and Placement

Close-miking amplifies mouth noises. While a proximity effect is sometimes desired, it also makes clicks and smacks more pronounced. Position the microphone slightly off-axis (angled) and at a distance of 6–12 inches from the mouth. A cardioid or supercardioid pattern can help reject off-axis noises. For voiceover, a large-diaphragm condenser microphone with a built-in pop filter often works well.

Use of Pop Filters and Windscreens

A pop filter placed three to four inches from the microphone capsule reduces plosive blasts and also catches some moisture before it hits the grille. For outdoor or windy conditions, a foam windsock provides additional protection. These simple accessories are low-cost but highly effective at reducing unwanted transient sounds.

Common Mistakes to Avoid

  • Over-processing with de-click plugins – Applying too much de-clicking can destroy the natural attack of consonants (especially T, K, and P sounds) and make dialogue sound mushy or unnatural.
  • Ignoring the context of the scene – In some cases, slight mouth noises add realism. Removing every single saliva click may make a dialogue track feel sterile. Exercise judgment based on the aesthetic of the project.
  • Relying solely on EQ – EQ cannot fix transient clicks without affecting the voice. Always use spectral tools as the primary method.
  • Cutting noise in the middle of speech – When using spectral selection, be careful not to select frequencies that are part of the voice’s formants. Audition the selection carefully before applying.

Conclusion

Removing mouth smacks and other unwanted mouth noises is a critical skill for dialogue editors and sound designers. By combining a thorough understanding of the noise types with modern spectral editing tools, dedicated de-click plugins, manual volume automation, and careful EQ, you can achieve clean, professional dialogue tracks. Prevention during recording also plays a vital role—simple steps like proper hydration, microphone placement, and the use of pop filters can save hours in post. With practice, these techniques become second nature, allowing you to focus on the art of storytelling rather than the annoyance of mouth clicks. For further reading, consult the official documentation of major audio repair tools such as iZotope RX, Adobe Audition spectral editing guide, and the Audacity noise reduction manual for practical examples.