Source-linked AI summary
Acoustic effects of medical, cloth, and transparent face masks on speech signals
Ryan M. Corey, Uriah Jones, Andrew C. Singer
TL;DR
Face masks can make speech harder to understand, especially for listeners with hearing loss. This paper measures acoustic attenuation across medical, cloth, transparent, and other face coverings using a head-shaped loudspeaker and a live human talker, finding high-frequency attenuation that is strongest in front of the talker and little effect on most lapel microphones.
Problem
Speech can be difficult to understand when talkers wear masks, especially for listeners with hearing loss, motivating measurements of masks’ acoustic effects.
Method
The study measures mask-related acoustic attenuation and directivity with a head-shaped loudspeaker, a live human talker, and microphones at listener distance and near the face.
Results
Masks generally attenuate frequencies above 1 kHz, with effects varying by mask type and cloth material and weave; transparent masks perform poorly acoustically, while most masks have little effect on lapel microphones.
Takeaways & Limitations
Lapel microphones and related amplification systems may support verbal communication with masked talkers, including when clear window masks preserve visual cues.
Abstract
from arXiv · showhide
Face masks muffle speech and make communication more difficult, especially for people with hearing loss. This study examines the acoustic attenuation caused by different face masks, including medical, cloth, and transparent masks, using a head-shaped loudspeaker and a live human talker. The results suggest that all masks attenuate frequencies above 1 kHz, that attenuation is greatest in front of the talker, and that there is substantial variation between mask types, especially cloth masks with different materials and weaves. Transparent masks have poor acoustic performance compared to both medical and cloth masks. Most masks have little effect on lapel microphones, suggesting that existing sound reinforcement and assistive listening systems may be effective for verbal communication with masks.
I. INTRODUCTION
Face masks can make speech harder to understand, particularly for listeners with hearing loss, while their acoustic effects vary across mask designs. This study measures those effects to identify masks and technologies that support speech transmission.
- Speech can be difficult to understand when a talker wears a mask, especially for listeners with hearing loss.
- Prior research largely examined surgical masks and N95 respirators, which attenuated higher-frequency sounds by 3–12 dB.
- Cloth-mask effectiveness at blocking respiratory droplets depends on fabric material, weave, and thickness.
- The study measures medical, cloth, transparent, and plastic face coverings using a head-shaped loudspeaker, a live human talker, and microphones near the face.It also examines how masks affect sound reinforcement and assistive listening systems.
II. METHODS
The experiments compare masked and unmasked speech using both a head-shaped loudspeaker and a live human talker. Measurements include listener-distance recordings, directional sweeps, repeated speech recordings, and microphones placed on or near the face.
- A microphone two meters from the talker simulated sound heard by a conversation partner.Additional lavalier microphones were placed at headset, lapel, cheek, and forehead positions.
- A head-shaped loudspeaker produced logarithmic frequency sweeps and was rotated in 15 degree increments to characterize mask-related directivity.
- A human talker made 30-second readspeech recordings, repeated three times non-consecutively with each mask.
- Both sources were measured without a face covering to establish a baseline, then with twelve face coverings.
A. Acoustic attenuation of face coverings
Masks generally preserve frequencies below 1 kHz but attenuate higher frequencies, with acoustic performance varying substantially by mask type and cloth composition. Transparent masks perform particularly poorly at high frequencies.
- Table I reports logarithmically weighted average attenuation from 2 kHz to 16 kHz.
- Most masks had little effect below 1 kHz but attenuated higher frequencies by different amounts.The surgical mask and KN95 peaked around 4 dB attenuation, while the N95 attenuated high frequencies by about 6 dB.
- Cloth-mask performance varied widely with composition and weave, with breathable fabrics transmitting more sound.Jersey- and plain-weave 100% cotton masks were comparable to the surgical mask, while tightly woven denim and bedsheets performed worst acoustically.
- Transparent masks blocked around 8 dB for the human talker and 10–14 dB for the loudspeaker at high frequencies.Their acoustic losses affect high-frequency sound cues important for speech despite preserving visual cues.
B. Effect of face coverings on speech directivity
Mask-related high-frequency attenuation is strongest in front of the talker and weaker to the side and behind. The findings suggest that masks may redirect sound energy laterally rather than simply absorb it.
- For all masks tested, acoustic attenuation was strongest in front of the talker.Sound transmission to the side and behind was less strongly affected, and the shield amplified sound behind the talker.
- The spatial pattern suggests that masks may deflect sound energy to the sides rather than absorb it.Microphones placed to the side of the mask may therefore be useful for sound reinforcement.
C. Effect of microphone placement
Masks attenuate high-frequency sound for distant listeners but have different effects on microphones placed on or near the face. Lapel microphones show little effect from most masks, while cheek microphones can show higher levels with spectral distortion.
- The listener and headset microphones experience similar high-frequency attenuation from the PVC window mask.
- The cheek microphone records higher sound levels under the mask, but with spectral distortion that varies with mask shape.
- Lapel and forehead microphones show small, mostly uniform attenuation across speech frequencies.
- Only the shield substantially affects speech spectra captured by a lapel microphone; most other masks have little effect.
IV. CONCLUSIONS
Face masks attenuate high-frequency sound in front of the talker, while mask materials and designs differ substantially in acoustic performance. Clear-window masks and shields perform poorly acoustically, but clear masks used with lapel microphones can preserve visual cues without strongly affecting sound capture.
- Masks attenuate high-frequency sound in front of the talker, with the strongest attenuation above 4 kHz.
- Polypropylene surgical masks perform best acoustically, while loosely woven 100% cotton masks are a well-performing alternative.
- Tightly woven cotton and cotton/spandex blends should be avoided when speech transmission is a concern.
- The study did not evaluate respiratory-droplet blocking, and acoustic performance may trade off against viral protection.
- Clear-window masks and shields perform worse acoustically than opaque cloth masks, but clear masks do not strongly affect lapel microphones.
ACKNOLWEDGMENTS
The acknowledgments identify the makers of two masks and describe the research fellowship support.
- Mask 5 was sewn by Ms. Catherine Somers, and mask 9 was sewn by Mr. Austin Lewis.
- The research was supported by an Intelligence Community Postdoctoral Research Fellowship appointment at the University of Illinois.