Susie Valentine, Ph.D. – Susie_Valentine@starkey.com
Martin McKinney Ph.D. – Martin_McKinney@starkey.com
Dania Rishiq – Dania_Rishiq@starkey.com
Tao Zhang, Ph.D. – Tao_Zhang@starkey.com
Starkey Hearing Technologies
6600 Washington Ave. S.
Eden Prairie, MN 55378
Popular version of paper 2pPPb23
Presented Tuesday afternoon, June 4, 2013
ICA 2013 Montreal
Stop and listen for a moment: what do you hear? Are the sounds that you hear annoying to you? Everyone has an idea of what sounds are annoying. Sounds typically reported as highly annoying include scraping sounds, chalkboard scratching, animal cries, etc. (Kumar, Forster, Bailey, & Griffiths, 2008). Listeners with hearing impairment (HI), especially those who utilize hearing aids, tend to find more types of sounds annoying, including paper rustling, dishes clanging and water running. These sounds are often less annoying to normal-hearing listeners and this difference may be one of the reasons that HA users can become easily fatigued in noisy conditions. For hearing aid users, a common approach to deal with this complaint is to reduce high frequency gain. While this approach may mitigate the complaint, it can create audibility issues for speech. The audibility issue created by reducing high frequency gain can often lead to speech understanding problems especially in the presence of noise, thus making this approach to addressing annoying sounds less favorable. A more effective approach is to determine the underlying cause of annoyance and then design an algorithm to selectively reduce it. There are many different properties of a sound that can lead to annoyance: loudness, sharpness, roughness, as well as, combination of attributes. While existing literature on annoyance perception for HI listeners is scarce, a previous attempt was made to investigate this perception using real-world recordings (Vishnubhotla, et al, 2012). The study showed a large variability of annoyance ratings across listeners, indicating that annoyance is different across people. However, the use of real world sounds adds additional confounds in that subjective associations with the sound sources could influence annoyance ratings. For example, the sound of an electric razor was presented to the participants; someone who has a negative view of shaving may rate an electric razor as highly annoying based on their opinion about shaving not necessarily about the properties of the sound.
In this study, we use abstract psychoacoustic stimuli designed carefully to avoid possible confounding subjective associations. Examples of sounds used in this study are provided. In the first phase of the study, listeners with bilateral hearing loss evaluated the range and order of annoyance of 72 unique stimuli. The did this by auditioning 6 different sounds at a time and ordering them from least to most annoying. This task was repeated a total of 36 times, resulting in each stimulus being presented a total of 3 times. The results of this phase helped determine a subset of stimuli, 39 unique stimuli, to be used for the second phase, an annoyance magnitude estimation task. During the magnitude estimation task, participants listened to a reference stimulus and a test stimulus, judged which stimulus was more annoying and then provided a numeric value of ‘how much more annoying the stimulus is’ (e.g. twice as annoyance, ten times as annoying). This magnitude estimation method was used to measure the annoyance of each stimulus in bilateral and unilateral hearing impaired listeners. The stimulus that was chosen to be the reference signal was ranked in the middle of the annoyance continuum from the first phase results. All stimuli were presented via insert earphones, at three different levels or loudness.
The results show that the number 1 driver of annoyance is loudness. In general, as a stimulus gets louder the annoyance rating goes up. This result was consistent across the bilateral and the unilateral listeners. Additionally, narrowband stimuli were rated as more annoying than broadband stimuli. Within the narrowband category the stimuli with the highest frequency rose to the top of the annoyance continuum. This indicates that sounds that only contain a small range of frequencies and specifically a range of frequencies with a high pitch are perceived as being more annoying than other sounds. In a real world environment this could represent keys clanging, dishes clanging together or many others. Sounds that fluctuate over time or change more often are rated as more annoying than steady-state sounds. For instance, paper rustling is more annoying than static noise from a TV that is not receiving a signal. Below are two sound samples, the first sound sample (sound A) is a sound that was rated by our listeners as being minimally annoying, whereas sound B represents the other end of the annoyance scale as the sound rated as being the most annoying.
The next step in this study is to look at various annoyance models and determine if any model that exists for normal-hearing listeners can account for these data. Once the annoyance perception of listeners with hearing loss can be modeled fairly accurately, the model can inform the development of advanced signal processing algorithms. The idea behind this type of algorithm is that the hearing aid will recognize the annoying sound and then alter the output of the hearing aid to reduce the annoyance of the sound. This would thus make listening through a hearing aid less bothersome to some HI listeners and take away the increased annoyance level of the sounds. Currently, hearing aids are designed to reduce various types of noise through gain reductions, such as wind noise or machine. That type of reduction is a level reduction in a specific or a broad range, which would make understanding speech more difficult, however providing listening comfort and reducing listening fatigue. To further enhance listening in noise, small specific adjustments could be used to reduce the specific property that makes the sound annoying, leading to a more pleasant listening experience without the reduction in speech understanding.
(1) Kumar, S., Forster, H. M., Bailey, P., & Griffiths, T. D. (2008). Mapping unpleasantness of sounds to their auditory representation. Journal Of The Acoustical Society Of America, 124(6), 3810–3817. doi:10.1121/1.3006380
(2) Vishnubhotla, S., Xiao, J., Xu, B., McKinney, M. and Zhang, T. (2012). Annoyance perception of environmental noises by hearing impaired listeners. Journal of the Acoustical Society of America, 131(4), 3536