How loud is loud speech and can hearing aids process a shout?

In a recent blog post, we looked back on a publication by Pearsons, Bennett, and Fidell (1977), who measured speech levels in a wide range of environments and established audiologic benchmarks for soft, average, and loud speech levels.

The review confirmed that, even with modern acoustic analyses, average speech levels fall around 65 dBA, with the quietest parts of speech reaching as low as 42 dBA (Šrámková et al., 2015).  In this post, we shift focus toward louder speech and ask, “How loud is loud and shouted speech?”

Benchmarks for loud and shouted speech

Returning to the benchmarks from Pearsons and colleagues, the levels for loud and shouted speech were recorded in an anechoic chamber without additional background noise.

• Loud speech from male talkers measured 76 dBA, while shouted speech increased by 13 dB to 89 dBA.
• For female and child talkers, loud speech levels were in the range of 72 dBA and shouted speech reached 82 dBA.
• Averaging across these examples, the range from casual conversation to shouted speech spans approximately 30 dB.

In today’s noisy world, 89 dBA is loud but does not present a listening risk, and does not approach the loudest human sounds, which fall far short of the loudest airborne animal sounds. As examples, opera singers routinely exceed 100 dBA and a lion’s roar will exceed 110 dBA.

The Lombard effect

Spoken language may also reach higher levels when talkers are exposed to background noise. This phenomenon is called the Lombard effect and occurs when talkers raise their voice as background noise levels increase (Brumm & Zollinger, 2011).

Interestingly, the Lombard effect is not unique to humans and happens for most vertebrates – many species of songbirds will sing louder in cities than in the countryside. The effect is a deeply embedded mechanism that serves the purpose of maintaining successful communication as background noise increases.

Bottalico and colleagues (2022) documented the Lombard effect in a group of older adults with hearing loss who were listening in a simulated restaurant setting. They found that speech levels increased by .51 dB for each 1 dB increase in background noise. In other words, increasing background noise from 80 dBA to 86 dBA would motivate a talker to increase their vocal level by 3 dB.

How Phonak hearing aids manage loud speech

An essential function of hearing aids is amplifying quiet speech while avoiding uncomfortably loud sounds. This is achieved through a process of amplitude compression, which applies more amplification to quiet sounds and comparatively less amplification to louder sounds. These systems avoid over-amplification and maintain listening comfort while restoring audibility for soft speech. For the loudest sounds (e.g., the opera singer or roaring lion), hearing aids have a protective output limiting mechanism.

However, this review shows that even in background noise, most spoken language levels remain below output limiting thresholds and within the range of amplitude compression. Audiologically, this highlights the importance of a well-designed compression system.

Phonak’s audiologists and engineers considered these factors when designing APD 2.0, the adaptive compression architecture in Lumity hearing aids.

The highlights of this system include adaptive time constants that combine the advantages of slow and fast-acting compression, and an additional knee point (i.e., level where compression begins to reduce gain) for loud speech inputs. APD 2.0 also provides a gain prescription that interacts with this adaptive compression architecture to maintain constant gain levels for inputs exceeding the loud speech knee point; this helps to maintain linearity and normalize loudness growth functions, each with a goal of ensuring loud signal clarity.

To learn more about the Lumity platform, we invite you to visit our website.

---

References

1. Pearsons, K. S., Bennett, R. L., & Fidell, S. A. (1977). Speech levels in various noise environments. Office of Health and Ecological Effects, Office of Research and Development, US EPA.
2. Šrámková, H., Granqvist, S., Herbst, C. T., & Švec, J. G. (2015). The softest sound levels of the human voice in normal subjects. The Journal of the Acoustical Society of America, 137(1), 407-418.
3. Brumm, H., & Zollinger, S. A. (2011). The evolution of the Lombard effect: 100 years of psychoacoustic research. Behaviour, 148(11-13), 1173-1198.
4. Bottalico, P., Piper, R. & Legner, B. (2022). Lombard effect, intelligibility, ambient noise, and willingness to spend time and money in a restaurant amongst older adults. Scientific Reports, 12, 6549.