Knowledge of the relation between people’s preferences and room acoustic parameters is critical for choosing the correct acoustic treatment and for creating good sound environments for people. In the frame of a PhD, people’s preferences regarding different acoustic environments were investigated. Their judgments were related to various room acoustics parameters. The results from the study can be used to achieve appropriate acoustic design.
The experiments
Different acoustic configurations were used, altering the acoustic properties in a furnished room. The configurations used both absorptive and reflecting treatment. The room acoustic measurements reverberation time, T20, speech clarity, C50, and sound strength, G, were measured. Speech was played from a loudspeaker in a fixed position and recorded at three different positions in the room.
In the listening test, performed in a neutral room, people were asked to judge the quality of the speech. For each sample of speech, they also chose predefined attributes, related to the clarity of the sound. In addition, the participants in the listening test rated the different samples of speech.
The participants were asked to imagine themselves sitting in a room listening to information. It should be noted that none of the participants in the listening tests had any insight into the study, and all sounds were encoded. This means that the participants could not relate any sounds to any of the configurations.
Results
Sound quality
Relating the responses from people in the listening tests to objective room acoustic measurements showed a good correlation between sound quality and speech clarity, C50. The evaluation showed a low correlation to reverberation time, T20.
A linear regression study showed R2, i.e., degree of explanation, of 65% or more between perceived sound quality and C50 for the frequencies 1000 Hz, 2000 Hz and 4000 Hz. For the frequency 2000 Hz, R2 was as much as 85%, meaning 85% of the answers can be explained by the regression line. This is a high number for subjective testing. The figure to the left shows scatterplots and regression lines for sound quality in relation to reverberation time, T20, and speech clarity, C50. Blue dots are the responses and red lines represent the linear regression.
Attributes
As regards the attributes, the correlation to reverberation time was the highest of the different room acoustic parameters measured. For the example at frequency 2000 Hz, the correlation coefficient was r=‑0.9 between attribute and reverberation time. However, a fairly good correlation could also be seen to speech clarity, again for the example of frequency 2000 Hz, with the correlation quality between attribute and C50 being r=0.7. (If r=-1 or r=1 the two variables are completely correlated. If r=0, no correlation is found between the variables). In the study where people rated the different sounds, the best one, for all positions, was the configuration with the most absorption. This configuration also had the highest C50.
None of the studies could show a good correlation to sound strength, G. However, it should be noted that this parameter showed only small variations between the different test positions and configurations. The use of other attributes could also have produced different results on correlation and regression. It can thereby not be concluded that G is not an important parameter to consider.
People’s preferences
The results from this study show that for rooms where people listen to information, it is very important to consider the room acoustic parameter C50, as this parameter corresponds very well to people’s sound preferences. Acceptable sound quality was obtained for C50=2 dB, but for satisfactory sound quality, as much as C50=8 dB was needed. A better sound environment increases the chances of hearing properly, also supporting learning and the performance of tasks. These results are therefore useful for the acoustic design of ordinary rooms such as classrooms and offices.
This study is explained in more detail in a paper published in MDPI International journal of Environmental Research and Public Health, which you can find by following this link https://www.mdpi.com/1660-4601/18/23/12274
