Initial Study on the Reverberation Time Standard for the Korean Middle and High School Classrooms Using Speech Intelligibility Tests

Round 1

Reviewer 1 Report

The paper is devoted to establishment of the acoustical standard for classrooms in Korea. A lot of native standards are known (not only in the US and UK) and the proposed criteria correlate with them. Taking into account the world experience we can hope that the new standard will comply with good practice. However, the proof of the proposal is not perfect. I would like to pay attention to the following points:

1. All references date back to 2013 and earlier. Many papers devoted to classroom acoustics have been published in the last decade. Probably recent achievements can be taken into account and affect the results of the study. In any case, please, give a brief review of recent results.
2. Acoustic conditions for a speaker are not studied. It is well known that in rooms with bad acoustics the speakers are forced to raise their voice. Long term speech with the raised voice can cause uncomfortable feelings and sometimes to diseases of the throat.
3. The suggested standard RT < 0.8 s allows for sufficiently anechoic classrooms, for example 0.2-0.3 s. But a such condition is not comfortable for a lecturer, also the speech level will be low in the back rows of the classroom. The relevancy of the short reverberation should be proved.
4. You do not discuss how the acoustic criteria may change in bigger or smaller rooms relative to the simulated room. Are the obtained results suitable for the classrooms with different sizes and proportions?

And a few minor remarks:

1. There are two types of bars in Figure 1. What is the difference between them?
2. Table 3 contains only NRC values. Is the reverberation time calculated using NRC values? Usually, frequency dependent absorption coefficients are used, and hence the reverberation time depends on frequency. But we see single RT values in Figure 4. Did you somehow average it of use the value at certain frequency?
3. The similar problem with STI. This index is not local, it values vary from seat to seat. But we see only one value in Figure 4. How did you calculate it?
4. 6 is not clear. What do the abbreviations (PB, SRT, CVC) mean?
5. Please, clarify the capacity of the simulated classroom (section 3.1).

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

This is a review of buildings-1287543, titled: "Establishment of the acoustical performance standard of the Korean classrooms using speech intelligibility test". Setting acoustical standards for learning environments ensures that both the students as well as the teachers can learn and work effectively. The authors have presented measurements of background noise levels and reverberation times of 15 classrooms in Cheonju. Listening tests were also designed to simulate different reverberation conditions to evaluate speech intelligibility in Korean. Although the subject should be of interest to the education sector, there are many areas in the paper which must be improved. The standard of the English language must be improved significantly.

Firstly, the paper is not replicable at all. Details about all the measurements and the equipment used are not reported, i.e. details of the classrooms, measurement equipment for background noise and reverberation time, listening test equipment, listening test setup, etc. Fifteen classrooms also seem to be a small sample to set the standard for the entire country. 

Secondly, the subjective listening test was conducted with adults only, whereas the standard is meant for students. And no statistical analysis was conducted.

Thirdly, the recommended background noise level of 35 dBA was not based on any experiments. 

Lastly, the introduction of facemask wearing due to COVID-19 would also impact speech intelligibility and there should be a discussion on this.

Therefore, this paper cannot be accepted and there must be an overhaul of the experimental procedure.

Some detailed comments are as follows:

1. The studies referenced in the introduction are outdated and recent studies should be examined and included.
2. Details of measurement equipment are missing.
3. Details of measured classrooms are missing.
4. Details about participants are missing, e.g. age? gender?
5. Listening test setup and equipment missing.
6. Only one SNR was tested, and no background noise was simulated.
7. How was the speech simulated? As a point source? Or talking head model?
8. Figure 6 is difficult to see, the plot should be redrawn with colors or markers for greater clarity. Captions are also lacking in detail.
9. The caption of Figure 7 is not sufficiently informative. Both the axes should be labelled in more detail and with appropriate units. Are the wicks referring to standard error?
10. Figure 8 is too crowded. The plot should be redrawn with colors or markers for greater clarity, and legends should be included. Captions are also lacking in detail.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear Authors,

the introduction is quite confusing. You are addressing so many issue (background noise, intelligibility, reverberation time, C50...) please discuss one item at a time.
On the reverberation time and in general on the differences between reverberant classroom and optimized ones, some paramount papers are missing like

10.1016/j.apacoust.2021.108051
10.4103/1463-1741.70506 
10.1044/1092-4388(2009/08-0111) 

Materials and Methods. I really could not understand the intelligibility test you performed. why suing 72 adults? why not children? the schools results are related to children, not to adults...
How they graded the speech intelligibility on this test? are there surveys?

Since you quoted Bradley in your introduction, why do you carried out only RT parametric tests and not also STI and C50? 
Why did you not consider also impaired individuals? It is known that ranges suitable for impaired individuals suit also normo - ones. Please check this as example https://doi.org/10.3390/app11093942. Please Include some discussion on this topic. Since you are proposing some new guidelines for you country (very good of you, by the way!) please do not forget to include unfortunate persons!

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have improved the paper, it is clearer now. The new title is more suitable to the content of the article and assume further development of the acoustic standard. There are only a couple of points to be reflected.

What frequencies do you mean in lines 250-251? If they are 500 and 1000 Hz (as in Table 10), that you should give the absorption coefficient values for 500 and 1000 Hz in Table 4 instead of NRC.

You are still insisting on the criteria RT < 0.8 s, so anechoic rooms seem to be suitable for lectures and classes. Please, take care about lecturers! According to my own experience it is very difficult to talk in the anechoic rooms for a long time. My voice gets tired, whereas the listeners in the back rows have problems with the speech level. I think that the minimum threshold for the reverberation time can be proposed.

Author Response

Please see attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors have addressed most of the issues raised. The key change is the scope of the paper, which is much improved and appropriately reflects the content presented. The details of the sites and experimental procedures are now in order.

Given the lack of timeliness of the experiments (i.e. done in 2013), the authors should acknowledge that the results must be supplemented with additional experiments in the near future to validate the findings from 2013.

Also, it seems that there is no mention of ethical approval for the human experiments, is that not needed in Korea?

Lastly, there are some editorial issues to be ironed out:

1. Stylistic problems. I.e. starting a sentence with numbers
2. Figure quality. The figures are of poor quality and should be replotted.

Author Response

Please see attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear authors,

the paper is much more improved. Thank you. Please be aware that reference 26 is wrong. Name and last name are the other way around (you inverted name and last name, it should be Caniato M. Marzi A. Gasparella A. ) Please correct.

Author Response

Please see attachment

Author Response File: Author Response.pdf