Calculation of 1/f Fluctuation from Sound Signal and Comfort Evaluation

Round 1

Reviewer 1 Report

This study aims to relate comfort to 1/f fluctuations. They defined the 1/f fluctuation coefficients and conducted a subjective evaluation experiment on comfort. Currently, the results cannot sufficiently support the conclusions. There are two major points needed improvement:

1. The general noise, e.g., the pink noise, with various 1/f fluctuation coefficients, should be generated as an example to verify the calculation method.

 

2. The subjective experiment was not scientifically designed:

a) one original music with various 1/f fluctuation coefficients is not sufficient;

b) the pink noise (with various 1/f fluctuation coefficients) should be adopted as the stimuli;

c) for pairwise comparison method, all stimuli should be rated at least twice, and the errors should be calculated. 

Author Response

Thank you for your points out.
We submitted the response sheet and revised manuscript.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper proposes different calculation methods for music fluctuation parameters. In particular a total fluctuation, a frequency fluctuation and an amplitude fluctuation coefficient are proposed.

The objective is to find a relationship between fluctuation  coefficient and subjectively evaluated comfort.

The most important finding from the point of view of the authors is that a larger fraction of the 102 subjects enrolled in the study chosen a music with fluctuation coefficient close to -1, or, in other words, 1/f fluctuation behavior.

The subjects were asked to associate adjectives to their impression in listening to music. The subjects who preferred music with fluctuation coefficient far from -1 chose the adjectives powerful, strong, clear. The subjects who felt more  comfortable listening to music with 1/f fluctuation associated to their choice the adjectives sleepy and vague.

I think that this paper addresses an issue that could be interesting. The experimental design is satisfying also if I would suggest to propose a randomized trial. In other words, I think that it would be better if music A (fluctuation coefficient far from -1) and  music B (fluctuation coefficient close to -1) were presented in a randomized order.

The paper needs an extensive revision of the English language. Sometimes it is very difficult to understand the meaning of the text. I tried to rephrase some sentences especially in the methods section but I would suggest a complete revision made by a native English scientist.

What raises my major perplexities is the fact that there is no attempt to provide a statistical analysis of the data. The authors found that a larger fraction of subjects chose the music of type B but they also found a difference between males and females. I would have expected that the authors had presented the result of a Pearson chi square test to assess if the difference between males and females is statistically significant.

I would also propose to treat more quantitatively the relation between the music A or B choice and the adjectives used to define the impression. A multivariate logistic regression could be used, for example, in which the adjectives, or maybe a subset of them, are explaining factors of the choice. Other variables of interest are sex and age. I would like to see the how the choice A or B are represented on the impression parameters space

I think that missing a statistical analysis of the data makes the paper too weak to be published.

 

 

 

 

 

 

Author Response

Thank you for your points out.
We submitted the response sheet and revised manuscript.

Author Response File: Author Response.pdf

Reviewer 3 Report

Ref. Review: Applsci-1777324_v1_reviewer

Paper Title: Calculation of 1/f Fluctuation from Sound Signal and Comfort Evaluation

Dear Authors and Editor,

This manuscript has a clear work structure, and the topic is relevant to understanding sound qualities in human auditory perception. Some items must be clarified before publication acceptance in the Applied Science journal.

Please observe the following items:

11) Lines 49 – 53, pp. 2: “To maximize the sound absorption coefficient of the soundproof material, it is necessary to make the thickness of the sound-absorbing material 1/4 or more of the wavelength of the noise. Therefore, when the frequency of the noise decreases, it is necessary to make the soundproof material thicker.” Please add a reference for the part of the text.

22)  In 2.1 Fluctuation Coefficient of the Whole Sound Signal, please show the decomposition of equation 1 until you can reach λm.

33)   In 2.2. Fluctuation Coefficient of the Frequency, please show the decomposition of equation 2 until you can reach λf.

44) In 2.3. Fluctuation Coefficient of the Amplitude, please show the decomposition of equation 1 until you can reach λa.

55) Figure 6. 1/f fluctuation coefficient λ of each fluctuation for each piece of music: The authors are mentioning the concentration of observations around -1. Please also explain the whole figure regarding frequency and amplitude.

66)  Figure 8 should be near the text.

77) Lines 223 – 224, pp. 7: “On the other hand, from results (b) and (c), each fluctuation coefficient was far from -1.” What was the reason? Please provide more details.

88) The authors are showing Figure 7, but there is no explanation for this figure. Please inform in the text what you are showing in this figure and its interpretation.

99) Figure 12. Sample of the participant selected an impression item in a free answer format: Please provide the references for selecting these adjectives.

110) 4. Conclusions: please inform the limitations of this study.

Author Response

Dear, Reviewer 3

Thank you for pointing out our manuscript.

Please see the attached file.

Author Response File: Author Response.pdf

Reviewer 4 Report

The comparison between the revised paper and old versions shows the efforts of the authors to improve the paper. The current version can be accepted for publication. 

Author Response

Dear, Reviewer 4

Thank you for checking our manuscript.

Round 2

Reviewer 1 Report

I don't think the authors have well solved my questions:

a) one original music with various 1/f fluctuation coefficients is not sufficient; b) the pink noise (with various 1/f fluctuation coefficients) should be adopted as the stimuli; c) for the pairwise comparison method, all stimuli should be rated at least twice, and the errors should be calculated. 

Author Response

Dear Reviewer

Please see the attached file.

Thank you for your cooperation.

Author Response File: Author Response.pdf

Reviewer 2 Report

I think that the paper was improved after the first revision process.

The language is still very difficult to be understood and the text is hard to be followed

I tried to insert some suggestions to clarify what is reported by the authors. I am attaching the .pdf version of the ms with the embedded comments. I would suggest in any case a large revision of the text by a colleague who was native English speaker.

A paragraph was inserted in the method section in which the accuracy of the three methods proposed for the calculation of the three fluctuation coefficients should be evaluated. This accuracy test is performed in the case of the pink noise for which the whole signal spectrum slope is by definition inversely proportional to the frequency. Calculating the spectrum in the case of the pink noise the authors found the expected 1/f behavior. They just checked that the pink noise behaves as such.

This step in the methodology is pleonastic and does not add any information as not the fact that the three fluctuation coefficients are different and measure different aspects as it is shown in Figure 8.

In what sense do the authors think they evaluated the accuracy of the calculation of the frequency and amplitude  coefficients?

I did not understand if only the frequency fluctuation coefficient was used to define music of type A or B or if also the other two fluctuation coefficients, the whole signal and the amplitude fluctuation coefficients were used in the comfort experiment.

Could you please clarify?

I think that the paper was improved after the first revision process.

The language is still very difficult to be understood and the text is hard to be followed

I tried to insert some suggestions to clarify what is reported by the authors. I am attaching the .pdf version of the ms with the embedded comments. I would suggest in any case a large revision of the text by a colleague who was native English speaker.

A paragraph was inserted in the method section in which the accuracy of the three methods proposed for the calculation of the three fluctuation coefficients should be evaluated. This accuracy test is performed in the case of the pink noise for which the whole signal spectrum slope is by definition inversely proportional to the frequency. Calculating the spectrum in the case of the pink noise the authors found the expected 1/f behavior. They just checked that the pink noise behaves as such.

This step in the methodology is pleonastic and does not add any information as not the fact that the three fluctuation coefficients are different and measure different aspects as it is shown in Figure 8.

In what sense do the authors think they evaluated the accuracy of the calculation of the frequency and amplitude  coefficients?

I did not understand if only the frequency fluctuation coefficient was used to define music of type A or B or if also the other two fluctuation coefficients, the whole signal and the amplitude fluctuation coefficients were used in the comfort experiment.

Could you please clarify?

Author Response

Dear Reviewer 2

Please see attached file.

Thank you for your cooperation.

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

As the authors argued, 'This paper focused on frequency fluctuation and comfortability for fundamental consideration'. It is necessary to have a standard sound stimulus,  e.g. noise with different 1/f fluctuation coefficients - from revised pink noise to pink noise. 

Unless new evidence from the new experiment has been added, I don't recommend publishing this article.

Author Response

Dear, Reviewer 1

Thank you for pointing out our manuscript.

Please see the attaced word file

Author Response File: Author Response.docx