Noise reduction and a noise gate are not interchangeable. Noise reduction aims to lower unwanted sound while you speak. A noise gate mainly turns down or mutes audio when you are not speaking. Choose the tool by where the noise occurs—not by which control makes the waveform look quieter.
Quick answer
Choose noise reduction when a steady background layer remains audible during speech. Choose a noise gate when unwanted sound is mainly noticeable in pauses and the speaker is consistently louder than the room. Use both only lightly, and preview difficult sentences before committing to a full video.
- Fan, HVAC, hum, hiss, or a steady road bed under speech: start with noise reduction.
- A noisy room mainly between sentences: try a gentle gate or expander.
- A few keyboard clicks, handling bumps, or door slams: use manual edits or clip gain.
- Background voices or TV dialogue: use cautious cleanup and improve the source; a gate cannot separate overlapping voices.
- Clipped, distant, or echoey words: find a better source, move the microphone closer, or rerecord.
What noise reduction does
Noise reduction tries to lower unwanted background sound while preserving dialogue. It is most useful when the interference has a reasonably stable pattern, such as fan wash, air conditioning, electrical hum, computer noise, camera hiss, road noise, or some wind.
The real test is not whether the background becomes silent. It is whether the voice is easier to understand without sounding metallic, hollow, watery, or flattened. The useful amount depends on the noise and how much change to the remaining signal is acceptable.
- Overlapping people, music, TV dialogue, and podcasts are less predictable.
- Clicks, barks, dishes, doors, and keyboard strikes are isolated events rather than a steady bed.
- Changing wind bursts, sirens, and passing vehicles can overlap speech abruptly.
- Severe echo, clipping, and a far-away microphone may require a better recording.
What a noise gate does
A noise gate responds to level. Below its threshold it turns a source down or off; above the open threshold it lets the source through. This makes it useful for pauses in a voiceover, tutorial, stream, or talking-head video.
A gate does not know the difference between a fan and a voice. A keyboard or nearby speaker can open it if they are loud enough. Once you begin talking, the gate remains open and the continuous background remains audible.
- Open threshold: too high and soft words or sentence starts will not open the gate.
- Close threshold: too high and the voice can cut in and out during quiet phrasing.
- Attack: too slow and beginning consonants disappear.
- Hold: too short and the result can chatter between syllables.
- Release: too fast and word endings vanish while room tone pumps.
Noise gate vs expander
An expander is often gentler for spoken video. Instead of switching audio fully off below a threshold, it lowers quieter material by a controllable amount. That keeps some natural room tone and avoids obvious on/off transitions.
Use a light expander when pauses need tidying without sounding artificial. Use a gate only when the pauses genuinely need to be quieter and normal quiet speech still opens it reliably. Neither replaces noise reduction for a fan that continues while the speaker talks.
Use noise reduction first when noise is always present
Choose this path for a screen recording with laptop fan noise, an interview beside air conditioning, a phone video with road wash, or a voiceover with constant microphone hiss.
- Keep the untouched source file.
- Find a representative section with normal speech, a quiet sentence, and a pause.
- Apply the lightest useful noise-reduction setting or preview.
- Compare Original and Cleaned at the same playback position and volume.
- Listen for clearer words, not simply a lower noise floor.
- Stop if consonants smear, the voice becomes robotic, or room tone pumps.
- Use the video-noise preview when steady noise is the main issue.
Use a gate or expander when the problem is in the gaps
Choose this path when speech is already clear but pauses reveal room tone, a computer fan, or other low-level background sound. Set microphone gain sensibly first; an extreme threshold cannot repair a distant microphone.
- Start with a gentle expander when the editor offers one.
- Set the gate’s open threshold below normal speech and its close threshold below the open threshold.
- Test normal phrases, quiet words, breaths, and pauses.
- Increase hold or release if word endings vanish or the result chatters.
- Recheck quiet speech after adding any other enhancement.
- Listen on headphones and on the speakers your audience is likely to use.
Use manual edits for isolated noises
For a chair scrape, keyboard clack, dog bark, handling bump, or door slam, edit the moment instead of increasing full-track processing. Duplicate the audio or project, split around the event only when it does not overlap a key word, and use modest clip gain with short fades.
If the event overlaps speech, preserve intelligibility first. Some remaining noise is usually better than a missing syllable.
Can you use noise reduction and a gate together?
Yes, provided each stage solves a different problem. Begin with a clean source and sensible microphone level. Apply light noise reduction for a steady bed under speech, then use a gentle expander or gate only if pauses remain distracting. Repair isolated hits manually and add a limiter only to prevent clipping after level changes.
Restraint matters more than a rigid order. If real-time suppression, headset processing, operating-system voice isolation, and editor enhancement are all active, disable layers one by one to find the smallest effective chain.
Common mistakes and better fixes
Match each problem to the narrowest useful tool rather than forcing one processor across the entire recording.
- A hard gate cannot remove fan noise during speech; use light noise reduction for the steady bed.
- A threshold that makes the room completely silent may remove breaths, soft words, and sentence endings; lower it or use an expander.
- Maximum denoise can make speech metallic or watery; use the lowest useful amount.
- One setting may not fit every section when noise or microphone distance changes.
- Neither tool reliably separates another speaker; use isolated tracks, a better source, or rerecording.
- Compression can make background sound more obvious, so repair steady noise before raising levels.
Practical examples
For a screen recording with laptop fan noise, use light noise reduction under narration and add a gentle expander only if pauses remain distracting. For occasional keyboard clicks, repair isolated hits locally instead of forcing a gate or denoiser beneath words.
For an interview with air conditioning and unequal speaker levels, preview steady-noise cleanup and check each speaker before using a gentle expander. For a phone video with traffic and nearby voices, a steady road layer may improve, while passing cars and overlapping conversation may require a cleaner source or replacement narration.
Final check before export
Judge the result by intelligibility and naturalness, not by silence between words.
- Identify whether the problem is steady noise, noisy pauses, isolated events, overlap, echo, or clipping.
- Use noise reduction for sound that continues under speech.
- Use a gate or expander only for pauses, with quiet words tested.
- Choose the least aggressive setting that improves intelligibility.
- Compare the same sentence before and after processing.
- Check breaths, S and T sounds, sentence endings, and room transitions.
- Preserve the original and avoid stacking strong processing layers.
Sources and update notes
This article was reviewed on September 6, 2026. Product interfaces and third-party features change; check the linked official documentation before relying on exact tool behavior.
