Commercial microphone windscreens cut wind-induced noise by roughly 15 to 25 dB, and some exceed 25 dB, according to the US National Bureau of Standards’ benchmark test of ten windscreens. That margin matters because wind noise climbs fast: hearing aid microphones saturate at about 12 m/s with levels up to 116 dB SPL (Zakis, Journal of the Acoustical Society of America 2011), and a behind-the-ear microphone position reached 94.84 dB(A) at just 20 km/h in a 2025 flow simulation (Riedel et al., Scientific Reports 2025). The human cost shows up in surveys: 42% of hearing aid wearers report dissatisfaction in windy situations (MarkeTrak VIII, cited in Korhonen et al. 2017). Our police motorcycle helmet wind noise roundup covers what wind does to the ear under a helmet; this page is about the microphone and what you put around it. We aggregated data from the Journal of the Acoustical Society of America, the Noise Control Engineering Journal, Canadian Acoustics, Energies, Scientific Reports, hearing-aid research in the Journal of the American Academy of Audiology and Hearing Review, and recent neural wind-noise reduction papers.
TL;DR
- Commercial windscreens reduce wind-induced noise by about 15 to 25 dB (NBS, Microphone Windscreen Performance 1979).
- Hearing aid microphones saturate at 12 m/s, with wind noise up to 116 dB SPL (Zakis, JASA 2011).
- A 610 mm secondary windscreen added 5 to 15 dB of low-frequency protection over a 76 mm foam ball at 4 m/s (Bonsma et al., Canadian Acoustics 2017).
- 42% of hearing aid users are dissatisfied in windy situations (MarkeTrak VIII, cited in Korhonen et al., JAAA 2017).
- Wind noise at a behind-the-ear microphone position reached 94.84 dB(A) at 20 km/h on a KEMAR manikin (Riedel et al., Scientific Reports 2025).
- Wind noise differed by up to 20 dB between hearing aid designs in some cases (Köhler and Georganti, Sonova, DAGA 2017).
- A fine mesh over a recessed microphone cut boundary-layer wind noise by 8 dB (van Dercreek et al., TU Delft, AIAA Scitech 2019).
- A wind noise algorithm improved phoneme identification by 17% to 35% across four speech levels (Lee, Hearing Review 2016).
- A diffusion-based neural model lifted PESQ from 1.70 to 3.07 on wind-corrupted speech (Lemercier et al., ITG Speech Communication 2024).
- A low-complexity neural wind-noise reducer runs with 249K parameters and about 73 MHz of compute (Eftekhari et al., Fraunhofer, 2025).
1. How Wind Noise Scales With Wind Speed
Wind noise is not sound arriving from a distance. It is pseudo-sound: pressure fluctuations created by turbulent air hitting the microphone itself, which is why a windscreen can remove it without blocking the voice you want. The physics is unforgiving. Because the pressure scales with the square of wind velocity, every doubling of wind speed adds about 12 dB in theory, and measurements often climb faster than theory (Korhonen, Wind Noise Management in Hearing Aids, Seminars in Hearing 2021, citing Strasberg 1988 and Morgan and Raspet 1992).
The benchmark rotating-arm tests show what that means in practice. Going from 2 m/s to 4 m/s, a gentle breeze to a light one, raised unscreened microphone noise from 59.6 dB to 75.7 dB, a jump of 16.1 dB (75.7 minus 59.6), already beyond the 12 dB square-law estimate (Hosier and Donavan, NBS, Microphone Windscreen Performance, NBSIR 79-1599, 1979). By 14 m/s the same bare microphone read 107.6 dB, which is why no amount of gain staging rescues an unprotected capsule in a gale.
| Metric | Value | Source |
|---|---|---|
| Overall wind noise at an unscreened microphone, grazing flow, at 2 m/s vs 4 m/s vs 14 m/s | 59.6 dB vs 75.7 dB vs 107.6 dB | NBS (now NIST), NBSIR 79-1599, 1979 |
| Theoretical increase in wind noise level for each doubling of wind speed | About 12 dB | Korhonen, Seminars in Hearing 2021 (citing Strasberg, JASA 1988) |
| Wind speed at which hearing aid microphone outputs saturated, and peak level | 12 m/s, up to 116 dB SPL | Zakis, JASA 2011 |
| Shape of the wind noise spectrum below vs above 300 Hz | Roughly flat below 300 Hz, falling 26 dB per octave above | Korhonen, Seminars in Hearing 2021 (citing Dillon et al., NAL 1999) |
| Difference below 20 Hz between still air and 5 m/s wind, tripod microphone with a single standard windshield | Up to 30 dB | Energies, Experimental Verification of Windshields 2022 |
The steep low-frequency tilt explains why wind ruins bass first: the energy sits below 300 Hz, so a high-pass filter removes much of the rumble but also the warmth of the voice. Most recent available data: Hosier and Donavan, 1979; the physics has not changed, but no newer open dataset repeats this speed sweep.
2. Foam, Fur and Basket Windscreens: Measured Attenuation
Material choice moves the result by more than 10 dB. Small foam balls are the weakest option: a 3.81 cm (1.5 inch) foam windscreen delivered 8.1 dB at 20.9 km/h, about half the low end of the commercial range, while purpose-built weatherproof designs reached 22.2 dB at 32.2 km/h (Hill, Blotter, Leishman and Sommerfeldt, Development of a Weatherproof Windscreen for a Microphone Array, Noise Control Engineering Journal 2006).
Foam density matters as much as foam presence. When engineers tested five foams from 20 to 60 pores per inch (PPI, a count of the open cells along one inch of material), the middle value of 40 PPI worked best, because foam that is too open lets turbulence through and foam that is too dense creates its own wake behind the screen (Zhao et al., On the wind noise reduction mechanism of porous microphone windscreens, JASA 2017). Basket-style designs, a rigid open frame that holds still air around the capsule, scale by stacking layers: a five-layer semi-spherical metal mesh windscreen reached 18.2 dB of wind noise reduction (Zhao et al., Experimental investigations on the wind noise reduction of semi-spherical metal mesh windscreens, ICSV 2018).
| Metric | Value | Source |
|---|---|---|
| Wind-induced noise reduction of commercial windscreens, depending on wind speed and flow direction | About 15 to 25 dB (some above 25 dB) | NBS (now NIST), NBSIR 79-1599, 1979 |
| 3.81 cm foam windscreen, wind noise reduction at 20.9 km/h vs 32.2 km/h | 8.1 dB vs 15.0 dB | Hill et al., Noise Control Engineering Journal 2006 |
| Weatherproof spider-leg windscreen design, wind noise reduction at 20.9 km/h vs 32.2 km/h | 8.9 dB vs 22.2 dB | Hill et al., Noise Control Engineering Journal 2006 |
| Best-performing foam among five porosities tested (20 to 60 PPI) | 40 PPI | Zhao et al., JASA 2017 |
| Five-layer semi-spherical metal mesh windscreen, wind noise reduction | 18.2 dB | Zhao et al., ICSV 2018 |
| Basket windshield used alone, wind noise attenuation at 100 Hz | 37 dB (vendor’s own claim) | Rycote, Modular Windshield Kit product sheet 2014 |
| Extra attenuation from adding a fur cover over the basket, and combined total, at 100 Hz | Plus 13 dB, 50 dB total (vendor’s own claim) | Rycote, Modular Windshield Kit product sheet 2014 |
The fur question has a surprisingly thin evidence base. The BYU team describes fur as very effective and quiet in airflow but much more expensive than foam, without a number, and the only fur figure we could trace is Rycote’s own single-frequency claim in its Modular Windshield Kit sheet. Treat any headline that puts a precise dB gap between fur and foam with caution until an independent lab publishes a side-by-side curve. Wind blasts from a talker’s own breath are a different, closer-range problem, covered in our pop filter and plosive attenuation statistics.
3. Windscreen Size: Bigger Helps, Up to a Point
The oldest rule of thumb ties attenuation to the enclosed volume. An empirical fit by Bleazey, reproduced in the NBS report, gives attenuation A = 6.77 log V + 10.4 dB, with V in cubic inches, valid from 2 to 13 m/s. The math is sobering: multiplying the volume by ten adds only 6.77 dB (because log 10 = 1), so doubling the diameter, which multiplies volume by eight, buys about 6.1 dB (6.77 x log 8 = 6.77 x 0.903).
Modern work explains why. A porous windscreen is most effective where its diameter is 2 to 4 times the turbulence wavelength, meaning the size of the swirling eddies carried by the wind (Zhao et al., Spatial decorrelation of wind noise with porous microphone windscreens, JASA 2018). Larger eddies, which carry the deepest rumble, need very large screens, which is why wind turbine acousticians stack a big secondary screen over a standard foam ball. A field-tested setup of that kind, described under IEC 61400-11:2012 (the international standard for measuring wind turbine noise), uses a ground board of at least 1000 mm, a primary half-sphere of at least 90 mm and a secondary foam sphere of at least 500 mm (Wszołek et al., HETMAN project deliverable D3.2, 2022).
| Metric | Value | Source |
|---|---|---|
| Empirical attenuation formula for windscreen volume V in cubic inches (2 to 13 m/s) | A = 6.77 log V + 10.4 dB | Bleazey fit, cited in NBS NBSIR 79-1599, 1979 |
| Diameter range where porous windscreens attenuate wind noise most effectively | 2 to 4 times the turbulence wavelength | Zhao et al., JASA 2018 |
| Best single-layer semi-spherical mesh windscreen among the sizes tested | 20 cm diameter (mid-sized) | Zhao et al., ICSV 2018 |
| Neumann foam windscreen of 50 mm vs 90 mm diameter, wind noise attenuation at 20 km/h | 15 dB vs 26 dB (vendor’s own claim) | Neumann, WS 81 and WS 87 product pages |
| 610 mm secondary windscreen plus primary vs a 76 mm foam ball alone, below about 160 Hz at 4 m/s | 5 to 15 dB more reduction | Bonsma, Gara and McCabe, Canadian Acoustics 2017 |
| Same secondary combination vs a 178 mm foam ball alone, lowest frequencies | Up to 5 dB or more better | Bonsma, Gara and McCabe, Canadian Acoustics 2017 |
| IEC 61400-11 setup: ground board, primary windscreen and secondary windscreen minimum diameters | 1000 mm, 90 mm, 500 mm | HETMAN D3.2, 2022, describing IEC 61400-11:2012 |
| Maximum expanded measurement uncertainty at 2.1 m/s, ground board with double windscreen vs microphone at 4 m height | 7.3 dB vs 11.3 dB | Energies, Experimental Verification of Windshields 2022 |
The vendor figures point the same way (the 90 mm Neumann WS 87 claims 26 dB against 15 dB for the 50 mm WS 81), but they are single-condition manufacturer numbers at 20 km/h (about 5.6 m/s). The Canadian team also flagged the trade-off: giant screens are fragile, hard to transport and catch wind loads on elevated mounts (Bonsma et al., Insertion Loss and Wind-Induced Noise Results for Secondary Windscreens, 2017). The uncertainty row comes from a 2022 field validation of wind turbine windshields published in Energies. For why turbine measurements push so hard into low frequencies, see our wind turbine infrasound statistics.
4. The Price of Protection: Insertion Loss, Treble Loss and Durability
Every windscreen also dims the sound you want a little. Engineers call this insertion loss: the drop in level of an ordinary sound when the windscreen is put on in still air. At speech frequencies the penalty is tiny: 0.1 to 0.6 dB between 250 Hz and 4 kHz in the NBS benchmark, and under 0.4 dB for the BYU weatherproof designs, well inside the 1 dB budget the authors set (Hill et al. 2006).
The cost moves to the treble as screens get bigger or denser. The 178 mm foam ball in the Canadian tests lost about 2 dB at 10 kHz and 3.5 dB at 16 kHz (Bonsma et al., Insertion Loss and Wind-Induced Noise Results for Secondary Windscreens, 2017), and wrapping a mesh basket in fabric traded clarity for wind protection (Zhao et al. 2018). That is the real engineering choice: a thick cloth cover added 3.0 to 7.0 dB of wind noise reduction but pushed insertion loss to 1.1 to 2.2 dB, while adding mesh layers instead improved protection and kept insertion loss low.
| Metric | Value | Source |
|---|---|---|
| 178 mm foam windscreen insertion loss at 10 kHz vs 16 kHz | About 2 dB vs about 3.5 dB | Bonsma, Gara and McCabe, Canadian Acoustics 2017 |
| Single-layer mesh windscreen with a thick cloth cover: extra wind noise reduction vs resulting insertion loss | 3.0 to 7.0 dB vs 1.1 to 2.2 dB | Zhao et al., ICSV 2018 |
| High-frequency loss at 15 kHz, 50 mm vs 90 mm Neumann foam windscreen | 2 dB vs 3 dB (vendor’s own claim) | Neumann, WS 81 and WS 87 product pages |
| Sand trapped in 4 minutes of blowing, standard foam vs cone-shaped weatherproof windscreen | 38.5 g vs 0.1 g | Hill et al., Noise Control Engineering Journal 2006 |
| Typical lifetime of foam windscreens under outdoor UV exposure | Less than one year | Hill et al., Noise Control Engineering Journal 2006 (citing earlier work) |
Durability is the hidden statistic. Foam that clogs with grit or crumbles in sunlight within a year quietly loses the attenuation printed on its spec sheet, which is why permanent outdoor monitors favor shaped, weatherproof designs over a bare foam ball. Most recent available data for foam durability: Hill et al., 2006.
5. Wind Noise in Hearing Aids, Hearables and Phones
Body-worn and pocket devices cannot carry a 90 mm windscreen, so microphone placement becomes the windscreen. A 2025 flow simulation found 94.84 dB(A) at the behind-the-ear microphone position of a KEMAR manikin at 20 km/h, against 70.98 dB(A) deep inside the ear canal, a 23.86 dB difference (94.84 minus 70.98) set by position alone (Riedel et al., Investigation of the physical driving mechanisms of wind noise in hearing devices, Scientific Reports 2025). Doubling the speed from 10 to 20 km/h raised the level at that position from 75.61 to 94.84 dB(A), or 19.23 dB.
Design spreads are just as large. In a Sonova wind tunnel study of three behind-the-ear and three in-the-ear designs, differences between designs reached up to 20 dB in some cases, mainly at the front microphone of behind-the-ear devices (Köhler and Georganti, Wind noise variability of different hearing aid designs and ear geometries, DAGA 2017). Directional microphone modes tend to make wind worse, because their low-frequency correction also boosts uncorrelated noise, and wind tunnel tests found directional settings generally noisier than omnidirectional ones (Chung, Mongeau and McKibben, JASA 2009); our directional microphone hearing aid SNR roundup covers the benefit side of that trade.
| Metric | Value | Source |
|---|---|---|
| Hearing aid wearers dissatisfied with their hearing aids in windy situations | 42% | MarkeTrak VIII (Kochkin 2010), cited in Korhonen et al., JAAA 2017 |
| Wind noise level difference between microphones within one behind-the-ear hearing aid vs between different behind-the-ear devices (front microphone, rear microphone) | Up to 12 dB vs up to 6 dB (front) and up to 8 dB (rear) | Zakis, JASA 2011 |
| Wind noise at the behind-the-ear microphone position vs deep in the ear canal, KEMAR, 20 km/h | 94.84 dB(A) vs 70.98 dB(A) | Riedel et al., Scientific Reports 2025 |
| Worst-to-best broadband difference within the same hearing aid style, A-weighted | 5 to 7 dB | Köhler and Georganti (Sonova), DAGA 2017 |
| Reduction of physical wind noise from a hearing aid microphone cover | 18 dB SPL (vendor’s own claim) | Widex brochure 2012, cited in Korhonen, Seminars in Hearing 2021 |
| Boundary-layer wind noise reduction from a fine mesh over a recessed microphone, wind tunnel | 8 dB | van Dercreek et al. (TU Delft), AIAA Scitech 2019 |
Phones sit at the bottom of this hierarchy: a single small port, no windscreen, and wind arriving from any angle. A March 14, 2026 study tested tubular front cavities of 1 to 3 mm diameter behind a smartphone-style housing at 4, 6 and 8 m/s, but did not yet measure how much they cost in speech sensitivity (Experimental Investigation of a Tubular Front Cavity for Wind Noise Suppression in MEMS Microphones of Mobile Devices, Micromachines 2026). The mesh result from TU Delft is the closest independent number for the acoustic meshes used in phone ports. Most recent available data for the device comparisons: Zakis, 2011, and Köhler and Georganti, 2017.
6. Digital Wind-Noise Reduction: What the Algorithms Deliver
When hardware runs out, signal processing takes over, and the hearing aid evidence is the most rigorous. A wind noise attenuation algorithm cut wind noise by 13 to 18 dB at 1 kHz for frontal wind at 4 to 7 m/s, and by 19 dB at 2,200 Hz at 10 m/s (Korhonen et al. 2017, as summarized in Seminars in Hearing 2021). In the same trial of 15 experienced wearers, the largest gain in phoneme identification was 48.2% for speech at 65 dB SPL in a 5 m/s wind (Korhonen, Kuk, Seper, Mørkebjerg and Roikjer, Evaluation of a Wind Noise Attenuation Algorithm, JAAA 2017). A separate evaluation with 10 adults found phoneme gains of 17% to 35% across four speech levels (Lee, Efficiency of a Wind Noise Attenuation Algorithm, Hearing Review 2016).
Neural models are now aimed at recordings and phones. PESQ, a standardized objective speech-quality score, rose from 1.70 for noisy input to 3.07 with a diffusion-based model on simulated wind noise, and the same model family generalized to real wind-tunnel recordings it had never seen (Lemercier et al., Wind Noise Reduction with a Diffusion-based Stochastic Regeneration Model, 2024). The open question is cost: a Fraunhofer model matched a leading low-complexity baseline with 249K parameters and about 73 MHz of compute, a budget that fits mobile chips (Eftekhari et al., Low-Complexity Neural Wind Noise Reduction for Audio Recordings, 2025).
| Metric | Value | Source |
|---|---|---|
| Wind noise attenuation of a hearing aid algorithm at 1 kHz, frontal wind 4 to 7 m/s, and at 2,200 Hz at 10 m/s | 13 to 18 dB, and 19 dB | Korhonen et al. 2017, as summarized in Korhonen, Seminars in Hearing 2021 |
| Largest phoneme identification gain with the algorithm on vs off (speech 65 dB SPL, wind 5 m/s, 15 wearers) | 48.2% | Korhonen et al., JAAA 2017 |
| Phoneme identification gain, algorithm on vs off, all phonemes vs vowels only (10 adults) | 17% to 35% vs 22% to 46% | Lee, Hearing Review 2016 |
| PESQ score, noisy input vs diffusion-based model, simulated wind-noise test set | 1.70 vs 3.07 | Lemercier et al., ITG Speech Communication 2024 |
| DNSMOS (a neural estimate of listener opinion), noisy vs diffusion model, unseen real wind recordings | 1.89 vs 3.56 | Lemercier et al., ITG Speech Communication 2024 |
| PESQ with vs without ultrasound airflow sensing added to a DCCRN speech enhancer | 2.480 vs 2.265 | Yuan et al. (Carnegie Mellon), DeWinder 2024 |
| Size and compute of a low-complexity neural wind-noise reducer | 249K parameters, about 73 MHz | Eftekhari et al. (Fraunhofer), 2025 |
The pattern across studies is consistent: algorithms recover meaningful intelligibility at moderate wind, but none of them undoes saturation. Once a hearing aid microphone clips at around 12 m/s (see section 1), the information is gone, which is why the best systems still start with port placement and a physical screen. The ultrasound result comes from the DeWinder preprint. For latency and quality figures on general-purpose neural denoisers, see our neural noise suppression statistics.
Summary: Microphone Wind Noise and Windscreen Attenuation by the Numbers
| Metric | Value | Source |
|---|---|---|
| Wind noise reduction of commercial windscreens | About 15 to 25 dB | NBS (now NIST), NBSIR 79-1599, 1979 |
| Theoretical wind noise increase per doubling of wind speed | About 12 dB | Korhonen, Seminars in Hearing 2021 |
| Hearing aid microphone saturation point | 12 m/s, up to 116 dB SPL | Zakis, JASA 2011 |
| 3.81 cm foam windscreen at 20.9 km/h | 8.1 dB reduction | Hill et al., Noise Control Engineering Journal 2006 |
| Five-layer metal mesh windscreen | 18.2 dB reduction | Zhao et al., ICSV 2018 |
| Basket windshield plus fur cover at 100 Hz | 50 dB (vendor’s own claim) | Rycote product sheet 2014 |
| Secondary 610 mm windscreen benefit vs 76 mm foam ball, below about 160 Hz | 5 to 15 dB | Bonsma et al., Canadian Acoustics 2017 |
| IEC 61400-11 primary and secondary windscreen minimum diameters | 90 mm and 500 mm | HETMAN D3.2, 2022 |
| Hearing aid users dissatisfied in wind | 42% | MarkeTrak VIII, cited in Korhonen et al., JAAA 2017 |
| Behind-the-ear microphone position at 20 km/h, KEMAR | 94.84 dB(A) | Riedel et al., Scientific Reports 2025 |
| Hearing aid design-to-design wind noise differences | Up to 20 dB | Köhler and Georganti (Sonova), DAGA 2017 |
| Fine mesh over a recessed microphone | 8 dB less boundary-layer noise | van Dercreek et al. (TU Delft), AIAA Scitech 2019 |
| Largest phoneme identification gain from a wind noise algorithm | 48.2% | Korhonen et al., JAAA 2017 |
| Phoneme identification gain in a separate evaluation | 17% to 35% | Lee, Hearing Review 2016 |
| PESQ before vs after diffusion-based wind noise reduction | 1.70 vs 3.07 | Lemercier et al. 2024 |
| Low-complexity neural wind-noise reducer | 249K parameters, about 73 MHz | Eftekhari et al. (Fraunhofer), 2025 |
Methodology and Sources
We ran 13 web searches and read the primary documents directly, using open repositories and the OpenAlex and Europe PMC databases where publisher pages were blocked. Vendor figures are labeled as the vendor’s own claim and kept to a minority of each section.
- Hosier and Donavan, National Bureau of Standards (now NIST): Microphone Windscreen Performance, NBSIR 79-1599, 1979 (also the source of the Bleazey volume formula).
- Hill, Blotter, Leishman and Sommerfeldt (Brigham Young University): Development of a Weatherproof Windscreen for a Microphone Array, Noise Control Engineering Journal, 2006.
- University of Technology Sydney windscreen papers by Zhao, Cheng, Qiu, Burnett and Liu: porous windscreen mechanism, JASA 2017; spatial decorrelation, JASA 2018; semi-spherical metal mesh windscreens, ICSV 2018.
- Bonsma, Gara and McCabe (HGC Engineering): Insertion Loss and Wind-Induced Noise Results for Secondary Windscreens, Canadian Acoustics, 2017.
- Experimental Verification of Windshields in the Measurement of Low Frequency Noise from Wind Turbines, Energies, 2022.
- Wszołek et al.: HETMAN project deliverable D3.2, 2022, describing the measurement setup of IEC 61400-11:2012.
- Zakis: Wind noise at microphones within and across hearing aids, JASA, 2011.
- Korhonen: Wind Noise Management in Hearing Aids, Seminars in Hearing, 2021 (review citing Strasberg 1988, Dillon et al. 1999 and a 2012 Widex brochure).
- Korhonen, Kuk, Seper, Mørkebjerg and Roikjer: Evaluation of a Wind Noise Attenuation Algorithm, Journal of the American Academy of Audiology, 2017 (also the MarkeTrak VIII 42% figure).
- Lee: Efficiency of a Wind Noise Attenuation Algorithm, Hearing Review, 2016.
- Riedel, Becker, Näger, Czwielong and Schoder: Investigation of the physical driving mechanisms of wind noise in hearing devices by computational fluid dynamics, Scientific Reports, 2025.
- Köhler and Georganti (Sonova): Wind noise variability of different hearing aid designs and ear geometries, DAGA 2017.
- van Dercreek, Manjunath, Ragni and Snellen (TU Delft): Design and Evaluation of Microphone Cavity Geometries for Wind-Tunnel Acoustic Measurements, AIAA Scitech 2019.
- Qingdao University of Technology authors: Experimental Investigation of a Tubular Front Cavity for Wind Noise Suppression in MEMS Microphones of Mobile Devices, Micromachines, March 14, 2026.
- Lemercier, Thiemann, Koning and Gerkmann: Wind Noise Reduction with a Diffusion-based Stochastic Regeneration Model, ITG Conference on Speech Communication, 2024.
- Yuan, Han, Kumar and Raj (Carnegie Mellon University): DeWinder: Single-Channel Wind Noise Reduction using Ultrasound Sensing, 2024.
- Eftekhari, Chetupalli, Shetu, Habets and Thiergart: Low-Complexity Neural Wind Noise Reduction for Audio Recordings, 2025.
- Vendor claims: Neumann WS 81 and WS 87 product pages; Rycote Modular Windshield Kit sheet, 2014.
- Data watch: Much of the windscreen evidence is old (the 1979 benchmark, the 2006 BYU study, the 2011 hearing aid wind tunnel study) and is flagged as most recent available data; no newer open study repeats a full speed sweep on an unscreened microphone. We found no peer-reviewed side-by-side measurement of fur vs foam, so the only fur numbers here are Rycote’s single-frequency (100 Hz) claims, which are not comparable with Neumann’s broadband figures at 20 km/h. Saturation thresholds differ across sources: Zakis measured saturation at 12 m/s, while Lee (Hearing Review 2016) cites Dillon et al. saying a 5 m/s wind can saturate microphones at 90 to 100 dB SPL; the two figures come from different devices and test setups, so we report the measured JASA value. The Seminars in Hearing review reports 49.6% phoneme identification with the algorithm and an 8.4 dB SNR benefit, while the JAAA abstract states a largest gain of 48.2%; these are different measures and we cite the JAAA figure. Lee also reports 58% satisfaction in wind from MarkeTrak, consistent with the 42% dissatisfied figure. In Riedel et al. 2025 a few in-text attenuation values do not match the paper’s Table 1; we used Table 1 only. IEC 61400-11 itself is paywalled, so its windscreen dimensions are taken from the HETMAN deliverable’s description of the 2012 edition; the 2018 amendment may differ.
Last updated: November 6, 2026. We update this roundup quarterly, and the next refresh is expected when the next edition or amendment of IEC 61400-11 and the next MarkeTrak hearing aid satisfaction survey are published.