# Remove Wind Noise: Use 80 Hz High-Pass Filter or Fitted Windshield Based on Recording Conditions

Hannah Morgan · October 3, 2026

> Remove wind noise with an 80 Hz high-pass filter at 12 dB/octave for low rumble, or fit a windshield when gusts repeatedly hit the mic directly.

| Takeaway | Detail |
| --- | --- |
| Apply an 80 Hz high-pass filter first when wind noise is low-frequency rumble | Use a gentle 12 dB/octave high-pass at 80 Hz if speech remains intelligible and rumble is below roughly 80 Hz |
| Choose a fitted windshield first when airflow repeatedly hits the microphone | If gusts still overload or contaminate the voice, stop increasing the high-pass and address physical wind protection |
| Directional microphones mounted on cameras include built-in wind noise reduction | These mics integrate a windshield, shock mount, high-pass filter, and two-stage level control for outdoor use |
| High-pass filters are adjustable in audio applications for low-frequency management | DSP settings allow activation of High-Pass Filter (HPF) functions to control bass response in recordings |

This guide explains how to remove wind noise from outdoor speech recordings using either an 80 Hz high-pass filter or a fitted windshield.

It provides specific thresholds and conditions for choosing the correct first corrective step based on the type and severity of wind damage.

![Remove Wind Noise](https://static.mm-ais.com/article-images-ai/remove-wind-noise-use-80-hz-high-pass-fi-ai-d2759526.jpg)

## Why Wind Becomes Recordable Noise

Wind does not arrive at a microphone as sound. It arrives as moving air, and outdoor air is rarely smooth. It curls into eddies as it passes the mic body, the shock mount, the camera cage, and the person holding the rig, and those eddies carry local pressure that changes from one instant to the next. A microphone diaphragm responds to pressure differences across its surface, not to anyone's intention to record speech, so turbulence is converted into voltage exactly as if it were an acoustic event.

Directional capsules offer only partial relief. A directional mic works by comparing the pressure at its front and rear ports, and it rejects some of the energy arriving from the sides and rear. Wind does not respect that geometry, because it is not a plane wave arriving from a defined angle. A gust can reach the front port and the rear port at different moments and at different pressures, so the same capsule that rejects a gust from one direction may accept one from another, and a single sustained gust can switch between the two as it churns. Repositioning helps; it rarely settles the problem.

Turbulence also has a shape in the frequency domain. Large, slow eddies produce the biggest pressure swings, and slow pressure swings are low-frequency energy. That is why wind on a recording sounds like rumble, thumps, and a low roar rather than hiss. The energy lands where adult speech fundamentals sit, where male vocal weight is carried, and where room and handling noise accumulate. A gentle high-pass at 80 Hz, 12 dB per octave, removes much of the turbulence below the fundamental while leaving the consonant range and most of the voice's intelligibility intact. It treats the overlap, not the whole spectrum.

The cheapest way to confirm the mechanism is to watch the unprocessed waveform while a gust passes.

| What the raw waveform shows during a gust | What it means |
| --- | --- |
| Slow, large, asymmetrical swings that shift the baseline | Turbulent pressure, not speech, is moving the diaphragm |
| Peaks flattening while the voice still looks normal | Wind pressure has already consumed the available headroom |
| The baseline recentering as the gust passes | The disturbance is airflow, not the source |

If the low end swells, the baseline wanders, and the peaks flatten while the voice stays visible inside the excursion, you are watching air pressure drive the diaphragm. Read that before you reach for any filter: it separates a low-frequency disturbance layered beneath the voice from a diaphragm that is being physically pushed.

![Why Wind Becomes Recordable Noise — Remove Wind Noise](https://static.mm-ais.com/article-images-pixabay/remove-wind-noise-use-80-hz-high-pass-fi-905d54af.jpg)

## What Available Evidence Supports

The most direct evidence for combining physical wind protection with electronic filtering comes from Tella's 2026 microphone guide, which lists a camera-oriented option that includes an integrated shock mount, windshield, and high-pass filter alongside two-stage level control. That product description treats wind protection and filtering as a single workflow rather than competing fixes, which supports the idea that both layers belong in the same setup. However, the grounding material reports neither measured attenuation figures nor specific recording conditions, so it should be read as evidence of available equipment, not as proof of a universal cutoff frequency or wind-noise performance value.

To verify a candidate microphone or accessory, check whether its windshield is physically fitted to the capsule and whether its high-pass control can be set near 80 Hz without introducing voice thinning. A unit that offers both features in one housing passes the first test; a unit that relies on a separate windscreen and a software filter passes only if the filter slope is at least 12 dB/octave and the windscreen reduces gust peaks by a visible margin on a meter. These checks separate products that support real-world adjustment from those that look complete on paper but fail once wind conditions change.

Compare the two approaches by listening for the point where gusts stop overloading the voice. If speech remains intelligible and the only defect is rumble below roughly 80 Hz, a gentle 12 dB/octave high-pass at 80 Hz is sufficient. If gusts still contaminate the voice above that band, stop increasing the filter and add or upgrade the windshield instead. This comparison should be done with a controlled test recording before committing to a full take.

The method stops working when low-frequency damage extends above 80 Hz or when repeated airflow reaches the microphone diaphragm directly. In those cases, no amount of filtering below 80 Hz will restore clarity, and the only reliable fix is a better-fitted windshield or a move to a sheltered position. Before committing to a purchase, confirm that the high-pass control is accessible during recording and that the windshield can be swapped or cleaned without tools.

Before committing to a purchase, confirm that the high-pass control is accessible during recording and that the windshield can be swapped or cleaned without tools. These two checks separate products that support real-world adjustment from those that look complete on paper but fail once wind conditions change. This verification step ensures the equipment can adapt to changing outdoor conditions during a recording session.

![What Available Evidence Supports — Remove Wind Noise](https://static.mm-ais.com/article-images-pixabay/remove-wind-noise-use-80-hz-high-pass-fi-12509bca.jpg)

## Compare Windshield and High-Pass

Physical wind protection and electronic filtering solve different parts of the same problem, and the choice between them should follow what the recording actually shows. This section alone ranks physical wind protection against electronic filtering for a recorded take. If the defect is steady low-frequency rumble and speech remains intelligible, an 80 Hz high-pass filter is the correct first corrective move. If airflow repeatedly reaches the microphone and low-frequency damage extends above 80 Hz, a fitted windshield is the better first choice.

| Option | Best condition | Main failure mode | Action |
| --- | --- | --- | --- |
| 80 Hz high-pass | Steady low-frequency rumble; speech remains clean | Leaves midrange turbulence and can thin a low voice | Apply first, then audition |
| Fitted windshield | Repeated gusts, outdoor movement, or contamination above 80 Hz | Adds weight, changes resonance, or fits poorly | Add before the next take |
| Noise reduction | Steady residual bed after capture | Removes speech texture or pumps | Use only after |

Monitor first: if speech remains intelligible and rumble below roughly 80 Hz is the defect, apply a gentle 12 dB/octave high-pass at 80 Hz. Listen for thinning in lower-register voices, and pull the cutoff back toward 60 Hz if the tone collapses. If gusts still overload or contaminate the voice, stop increasing the filter slope and add a windshield instead.

A fitted windshield earns priority when the recording shows broadband contamination above 80 Hz or when the take involves movement. The physical barrier removes energy before it reaches the diaphragm, which prevents the midrange turbulence that a high-pass filter cannot touch. Check the fit: a loose windshield shifts resonance and can sound worse than no windshield at all.

Noise reduction belongs last, not first. It works on a steady residual bed captured after the take, and it carries the risk of removing speech texture or introducing pump artifacts. Apply it only after the windshield and high-pass have done their work, and keep the reduction under 6 dB to preserve natural decay.

The ranking is conditional, not absolute. A quiet indoor voice recorded with a low-cut switch engaged may need no windshield at all, while a reporter walking between buildings needs both. Use the table as a decision tree: identify the dominant failure mode, match it to the best condition, and act in the order the table prescribes.

![Compare Windshield and High-Pass — Remove Wind Noise](https://static.mm-ais.com/article-images-pixabay/remove-wind-noise-use-80-hz-high-pass-fi-dd1dd433.jpg)

## Set the Filter by Ear and Meter

Begin with a 12 dB/octave high-pass filter set to 80 Hz. This slope removes enough low-frequency rumble to make a decisive difference while preserving the vocal fundamental, which for most adult voices sits between 85 Hz and 180 Hz. Setting the cutoff here gives the clearest test of whether wind is the dominant problem and whether electronic filtering alone can solve it.

If rumble remains audible after engaging the 80 Hz filter, raise the cutoff to 100 Hz and listen again. This is a small step, but it often clears residual wind without thinning the voice. If consonants lose crispness, chest resonance fades, or low-volume speakers sound distant, return to 80 Hz. The goal is intelligibility first, not maximum rumble removal.

Watch the meters during gusts. If the input signal reaches 0 dBFS or shows visible clipping, do not adjust the filter yet. Lower the preamp gain or move the microphone farther from the wind path before making any other change. A clipped signal cannot be fixed in post, and raising the high-pass frequency will not recover lost peaks.

Use this threshold to judge the 80 Hz setting: if speech remains clear and rumble below 80 Hz is the only defect, the filter is working. If gusts still overload the capsule or low-frequency damage extends above 80 Hz, the microphone is being struck by direct airflow and a fitted windshield is the better first choice.

| Filter Setting | When to Use | Check |
| --- | --- | --- |
| 80 Hz, 12 dB/octave | Rumble below 80 Hz is the main defect | Speech remains intelligible |
| 100 Hz, 12 dB/octave | Rumble persists after 80 Hz pass | Consonants and chest resonance intact |
| Stop and add windshield | Gusts still overload or damage extends above 80 Hz | No clipping during gusts |

Do not increase the high-pass frequency beyond 100 Hz unless the voice is clearly male and the fundamental is well above 100 Hz. For most speakers, pushing past this point removes too much natural warmth and makes the recording sound thin. If higher cutoffs are needed, pair them with a physical windshield rather than relying on filtering alone.

![Set the Filter by Ear and Meter — Remove Wind Noise](https://static.mm-ais.com/article-images-pixabay/remove-wind-noise-use-80-hz-high-pass-fi-303efc9d.jpg)

## Know Where the Method Stops

An 80 Hz-only treatment is invalid when the take contains damage a filter cannot undo or wind contamination that reaches the vocal range. Check the recording itself, not just whether the lowest rumble becomes quieter: the speech must remain intact, and the unwanted sound must be confined to a range the filter can address. If either condition fails, stop treating the cutoff as a cure.

Listen for a gust that has clipped the recording. A harsh crack, flattened peak, or abrupt distortion that stays in the voice after filtering is a sign the waveform has already been changed; removing frequencies below 80 Hz cannot restore the missing detail. Reduce the recording gain or improve the shielding, then repeat the recording. Do not keep raising the cutoff in an attempt to erase distortion that is embedded in the speech.

Wind can also leave a rough, speech-like hash above the filter’s cutoff. Temporarily audition the 200–500 Hz region and compare it with the full recording. If the turbulence is still audible there, an 80 Hz high-pass will not make it disappear just because the noise is strongest at lower frequencies. Add physical protection when that midrange contamination remains, then check that the voice is clean rather than relying on a more aggressive filter alone.

Choose the final cutoff for the actual speaker and recording, not by treating 80 Hz as universally optimal. A speaker’s vocal range, the microphone’s response, wind speed, and the room all affect how much filtering the take can tolerate. After addressing clipping or persistent broadband turbulence, adjust the cutoff cautiously and listen for a voice that has become unnaturally thin. If the voice loses body before the wind becomes acceptable, return to the recording conditions and improve the shielding instead.

![Know Where the Method Stops — Remove Wind Noise](https://static.mm-ais.com/article-images-pixabay/remove-wind-noise-use-80-hz-high-pass-fi-cf63e88a.jpg)

## Test One Outdoor Recording

Make the comparison controlled before changing anything. Use one speaker, one microphone position, and one outdoor exposure: record 20 seconds of normal speech followed by 20 seconds of silence without moving the microphone or speaker. Retain the raw file, and note the preamp setting, input gain, and any input pad. Those details make the before-and-after comparison repeatable rather than a judgment based on two different performances.

Duplicate the raw file so the original remains untouched. Leave one copy unprocessed, and apply a gentle 12 dB/octave high-pass filter at 80 Hz to the other. Level-match the copies by ear before switching between them; a louder copy can seem cleaner even when it is not. Listen first to breathy consonants such as *f*, *s*, and *h*. They should remain clear, while low rumble and movement from handling should recede. If the filter dulls the consonants or makes the voice feel thin, the setting is already doing too much.

Next, judge the 20-second silence rather than only the spoken portion. If the filtered copy is cleaner but the silence still contains audible gusting, try 100 Hz once and repeat the same level-matched comparison. Keep that change only if the voice remains natural and the remaining noise decreases. If speech hash appears, or consonants lose their edge, restore the 80 Hz setting instead of continuing upward. This gives you a practical upper limit based on the recording, not a preset number applied without listening.

Use the result to decide what to do with the next take. When speech stays intelligible, the main defect sits below roughly 80 Hz, and the filtered silence is acceptably quiet, keep the gentle 80 Hz high-pass as the first correction. When gusts still overload the input, remain audible through the silence, or contaminate the voice after the single 100 Hz check, stop increasing the filter. Fit a windshield and repeat the same 20-second speech-and-silence test at the unchanged position and preamp setting.

Save both filtered trials with descriptive names, such as raw_80Hz and raw_100Hz, and keep the unprocessed file beside them. A useful pass is not merely the version with less bass; it is the version in which consonants survive, rumble and handling movement are reduced, and the silence no longer reveals a distracting gust pattern.

## Apply Four Final Decision Rules

Use these four checks to turn the current recording into a conditional choice for the next take. If the defect is chiefly steady rumble below 80 Hz and no peaks clip, apply an 80 Hz high-pass filter. Compare the filtered voice with the bypassed version at a similar listening level. Keep the filter only if the speech stays natural and the rumble is reduced; a cleaner meter reading alone is not a reason to accept a thinner voice.

If the noise rises and falls with gusts, fit a windshield before recording again. Make a short test clip and listen through the loudest gusts, not just the quiet gaps. Check that the voice remains clear and that the gusts no longer dominate the take. If they still contaminate speech, do not assume a stronger filter is the next fix; reassess the setup before committing to a longer recording.

If either correction distorts the voice, return to the raw recording rather than stacking more processing on the damaged result. Rebalance gain, microphone distance, and shielding, then make another test clip. Listen for a stable voice level and natural tone while checking that peaks remain unclipped. Change one part of the setup at a time so the next comparison shows which adjustment helped, and keep the raw file available for that comparison.

Use a 100 Hz trial only as a check against the 80 Hz result. If it improves clarity without thinning the speaker, retain 100 Hz for that recording. If the voice sounds hollow, return to 80 Hz. If neither setting improves the speech, stop filtering and treat the remaining problem as broadband airflow rather than low-frequency rumble. At that point, adjust the recording setup instead of raising the cutoff again, then verify the change with another short listen.

## What to do next

| Step | Action | Why it matters |
| --- | --- | --- |
| 1 | Monitor a test recording to confirm that speech remains intelligible and the defect is low-frequency rumble. | A high-pass filter is appropriate only when it can reduce rumble without masking the voice. |
| 2 | If that condition is met, activate the High-Pass Filter (HPF) in the audio application’s DSP settings and use a gentle 12 dB/octave high-pass at 80 Hz. | This targets rumble below roughly 80 Hz while avoiding unnecessarily aggressive filtering. |
| 3 | Play back the recording with speech present, and keep the setting if the rumble falls away while the voice remains clear. | The filter should remove the low-frequency defect without weakening intelligibility. |
| 4 | If airflow repeatedly hits the microphone, fit a windshield before recording again. | Physical wind protection addresses direct airflow instead of asking the high-pass filter to remove every trace of it. |
| 5 | For outdoor use with a directional camera microphone, use its built-in wind noise reduction and integrated windshield and shock mount. | Directional camera microphones combine wind protection, mounting, filtering, and two-stage level control for outdoor recording. |
| 6 | If gusts still overload or contaminate the voice, stop increasing the high-pass and add or refit physical wind protection before recording again. | Once filtering cannot separate the voice from gust noise, stronger physical protection is the correct next move. |

## Frequently Asked Questions

**What frequency should I set for the high-pass filter to reduce low-frequency wind rumble?**

Apply an 80 Hz high-pass filter first when wind noise is low-frequency rumble.

**What slope should I use for the high-pass filter if speech is still intelligible?**

Use a gentle 12 dB/octave high-pass at 80 Hz if speech remains intelligible and rumble is below roughly 80 Hz.

**When should I choose a fitted windshield over adjusting the high-pass filter?**

Choose a fitted windshield first when airflow repeatedly hits the microphone.

**What should I do if increasing the high-pass filter doesn't stop wind from overloading the recording?**

If gusts still overload or contaminate the voice, stop increasing the high-pass and address physical wind protection.

**Do directional microphones mounted on cameras have built-in wind noise reduction?**

Directional microphones mounted on cameras include built-in wind noise reduction.

**What components are integrated into directional mics for outdoor use?**

These mics integrate a windshield, shock mount, high-pass filter, and two-stage level control for outdoor use.

## Quick answers

| When should you apply an 80 Hz high-pass filter first? | Apply an 80 Hz high-pass filter first when wind noise is low-frequency rumble. |
| --- | --- |
| What type of high-pass filter should be used if speech remains intelligible and rumble is below roughly 80 Hz? | Use a gentle 12 dB/octave high-pass at 80 Hz if speech remains intelligible and rumble is below roughly 80 Hz. |
| What should be chosen first when airflow repeatedly hits the microphone? | Choose a fitted windshield first when airflow repeatedly hits the microphone. |
| What should you do if gusts still overload or contaminate the voice? | If gusts still overload or contaminate the voice, stop increasing the high-pass and address physical wind protection. |
| What do directional microphones mounted on cameras include? | Directional microphones mounted on cameras include built-in wind noise reduction, integrating a windshield, shock mount, high-pass filter, and two-stage level control for outdoor use. |

Also worth reading: **Remove Reverb from Audio Recordings with AI**: [Remove Reverb from Audio Recordings](https://audobox.com/blog/remove_reverb_from_audio_recordings_with_ai.php) · **Kill Room Echo in Recordings Without Soundproofing**: [Kill Room Echo in Recordings](https://audobox.com/blog/kill_room_echo_in_recordings_without_soundproofing.php) · **How to Restore Old Audio Recordings at Home**: [How to Restore Old Audio](https://audobox.com/blog/how_to_restore_old_audio_recordings_at_home.php)

### Related reading

- [Spectral Repair vs Neural Nets: Field Recording Noise in 2026](https://audobox.com/blog/spectral-repair-vs-neural-nets-field-recording-noise-in-2026.php)
- [Remove office background noise: 2026 65 dBA Krisp wins vs Zoom Teams](https://audobox.com/blog/remove-office-background-noise-2026-65-dba-krisp-wins-vs-zoom-teams.php)
- [Clean outdoor audio with AI wind noise removal](https://audobox.com/blog/clean_outdoor_audio_with_ai_wind_noise_removal.php)
- [Compare AI de-reverb tools: 3 tests, same podcasts, pick the 2026 winner](https://audobox.com/blog/compare-ai-de-reverb-tools-3-tests-same-podcasts-pick-the-2026-winner.php)
- [Podcast Sibilance Control: Artificial Intelligence vs Multiband—4 dB, Keep or Switch?](https://audobox.com/blog/podcast-sibilance-control-artificial-intelligence-vs-multiband4-db-keep-or-switch.php)
- [One Record per Interview: Verify the Corrected 2026 Rollup of 95](https://audobox.com/blog/one-record-per-interview-verify-the-corrected-2026-rollup-of-95.php)

### Latest

- [Choose a Music Stem Splitter Without Files: Check 4-Stem, 44.1 kHz WAV](https://audobox.com/blog/choose-a-music-stem-splitter-without-files-check-4-stem-441-khz-wav.php)
- [Compare AI de-reverb tools: 3 tests, same podcasts, pick the 2026 winner](https://audobox.com/blog/compare-ai-de-reverb-tools-3-tests-same-podcasts-pick-the-2026-winner.php)
- [Podcast Sibilance Control: Artificial Intelligence vs Multiband—4 dB, Keep or...](https://audobox.com/blog/podcast-sibilance-control-artificial-intelligence-vs-multiband4-db-keep-or-switch.php)

Canonical: https://audobox.com/blog/remove-wind-noise-use-80-hz-high-pass-filter-or-fitted-windshield-based-on-recording-conditions.php
Markdown: https://audobox.com/blog/remove-wind-noise-use-80-hz-high-pass-filter-or-fitted-windshield-based-on-recording-conditions.php/index.md