[Survey] Can sound quality be measured?
Nov 15, 2018 at 2:33 PM Post #2 of 5
Asking a question like that around here is throwing a lighted match into a barrel of gasoline. :)

The Sound Science forum is probably a better place, although you were probably limited to this section if your post count is very low.

Yes, there are many tools that can be used to classify sound quality - frequency response, noise, distortion. There are even some people making attempts at imaging and soundstage measurements (not sure about those). Does it measure everything the ear can hear and the brain can perceive? No.
 
Nov 16, 2018 at 3:30 AM Post #3 of 5
Clicked on the link and couldn't answer the first four questions, because the questions are loaded with the incorrect assumption there IS a noticeable difference.
 
Nov 16, 2018 at 11:30 AM Post #4 of 5
I really wish I could answer your questionnaire... but I don't think I honestly can.
I'm going to be brutally honest: this doesn't seem very well thought. I really, really don't want to be rude in saying it (and I considered not saying it at all), but since you're asking these questions publicly you're likely to get bad answers to bad questions. It's not the respondents' fault, they're just trying to help. I'm assuming that, because you're a university student and this is university research, you're being serious about this-- so I'll give serious answers.

Questions 1-4: correlation ≠ causation. A headphone's greater performance does not cause it's price to be higher, and its greater price does not cause its performance to be greater. However, high performance headphones do generally cost more than low performance ones, though there are exceptions. The HD 650 and HD 6XX are (allegedly) the same headphone; HD 650 sells for $499, the HD 6XX sells for $199; does that mean the HD 650 sounds better?

Question 5 is pretty interesting and I think for most people (if they're being honest) the answer would probably be "a lot more than I would like." Myself included.

Question 6: First to niggle: you could (and should) change "qualitative factors" to read "qualities." Next, I wonder if you actually meant "quantitative factors," as in factors that can be quantified or measured? Third, perhaps there are people that disagree, but each of those things (with the exception of the first two, but that's another story that I shouldn't get into) contributes equally to the quality of a headphone.

Question 7: The problem with this question is twofold: presentation and interpretation. Which specs are being shared, and how well are they understood? If all of a headphone's quantifiable measurements are presented, and the person reading them knows how to interpret all of those measurements, then, yes, the measurements will be representative. If, on the other hand, only a few specs are given and they're poorly understood, then, no, the measurements are not representative. The problem is that many (most? all?) specs can be misleading if they're not properly understood.

Question 8: This question makes my brain hurt. It seems to me the answers should read, Yes, Maybe, Uncertain, Maybe Not, and No. Does this question mean, "can qualities that are previously unquantified be quantified?" Or does it mean, "can subjective qualities be measured?" You could restate the first meaning as "can something previously un-measured be measured?" The answer is "probably." The second meaning is a lot more difficult. For instance, can "pretty" and "ugly" be quantified? I find Leonardo's paintings to be pretty, but Monet's paintings to be ugly; you find Monet's paintings to be pretty, but Leonardo's paintings to be ugly. We're using the same words, but obviously they have different meanings for us. In order to measure "pretty" and "ugly" you'd have to have different metrics depending on who is doing the measuring. Audio is similar: something that sounds "warm" to me sounds "bloomy" to somebody else; something that sounds "warm" to somebody else sounds "dark" to me. The difficulty isn't that the qualities of sound are subjective, it's that our interpretations of them are.

Question 9: Okay, I'm gonna be pedantic again: research doesn't determine anything, it helps us understand it. With that in mind, would I be interested in research that helps us understand people's subjective interpretations of a given stimulus? YES!!!!! OMG THAT WOULD BE $(&^*#@ AMAZING!!!

Question 10: Refer to my response to question 8, above. First you'd have to know what each descriptor means to each person using that word. To do this even remotely realistically, you'd have to have a universal understanding of the language being used: in order to meaningfully measure them, the words "warm," "bloomy," "dark," "pretty," and "ugly" have to have the same meaning and apply to the same things for all people... At which point they're no longer subjective qualities open to interpretation, they're objective qualities that are quantifiable.

Question 11: See answers 1-10, above.

Assuming that you've read through this whole response, again, I'm really not trying to be mean or rude or to pick apart your thesis just for fun. This is, in a weird way, coming from a place of sympathy: I have a lot of unpleasant memories of high school and college wherein teachers and professors just kind of told me, "you're wrong," without giving much explanation. This just left me feeling lost and kind of stupid. I guess this falls under the saying there are no stupid questions, only stupid answers. Maybe it could be better phrased as there are no stupid mistakes, only stupid reactions to mistakes. To me, education is equally about asking questions, making and learning from mistakes, and learning new ideas. I think your questionnaire was full of mistakes, my hope is to help you learn from those mistakes.

I really hope that my answers have been helpful. If you have more questions, feel free to ask; hopefully I'll be able to answer.
 
Nov 16, 2018 at 1:28 PM Post #5 of 5
I really wish I could answer your questionnaire... but I don't think I honestly can.
I'm going to be brutally honest: this doesn't seem very well thought. I really, really don't want to be rude in saying it (and I considered not saying it at all), but since you're asking these questions publicly you're likely to get bad answers to bad questions. It's not the respondents' fault, they're just trying to help. I'm assuming that, because you're a university student and this is university research, you're being serious about this-- so I'll give serious answers.

Questions 1-4: correlation ≠ causation. A headphone's greater performance does not cause it's price to be higher, and its greater price does not cause its performance to be greater. However, high performance headphones do generally cost more than low performance ones, though there are exceptions. The HD 650 and HD 6XX are (allegedly) the same headphone; HD 650 sells for $499, the HD 6XX sells for $199; does that mean the HD 650 sounds better?

Question 5 is pretty interesting and I think for most people (if they're being honest) the answer would probably be "a lot more than I would like." Myself included.

Question 6: First to niggle: you could (and should) change "qualitative factors" to read "qualities." Next, I wonder if you actually meant "quantitative factors," as in factors that can be quantified or measured? Third, perhaps there are people that disagree, but each of those things (with the exception of the first two, but that's another story that I shouldn't get into) contributes equally to the quality of a headphone.

Question 7: The problem with this question is twofold: presentation and interpretation. Which specs are being shared, and how well are they understood? If all of a headphone's quantifiable measurements are presented, and the person reading them knows how to interpret all of those measurements, then, yes, the measurements will be representative. If, on the other hand, only a few specs are given and they're poorly understood, then, no, the measurements are not representative. The problem is that many (most? all?) specs can be misleading if they're not properly understood.

Question 8: This question makes my brain hurt. It seems to me the answers should read, Yes, Maybe, Uncertain, Maybe Not, and No. Does this question mean, "can qualities that are previously unquantified be quantified?" Or does it mean, "can subjective qualities be measured?" You could restate the first meaning as "can something previously un-measured be measured?" The answer is "probably." The second meaning is a lot more difficult. For instance, can "pretty" and "ugly" be quantified? I find Leonardo's paintings to be pretty, but Monet's paintings to be ugly; you find Monet's paintings to be pretty, but Leonardo's paintings to be ugly. We're using the same words, but obviously they have different meanings for us. In order to measure "pretty" and "ugly" you'd have to have different metrics depending on who is doing the measuring. Audio is similar: something that sounds "warm" to me sounds "bloomy" to somebody else; something that sounds "warm" to somebody else sounds "dark" to me. The difficulty isn't that the qualities of sound are subjective, it's that our interpretations of them are.

Question 9: Okay, I'm gonna be pedantic again: research doesn't determine anything, it helps us understand it. With that in mind, would I be interested in research that helps us understand people's subjective interpretations of a given stimulus? YES!!!!! OMG THAT WOULD BE $(&^*#@ AMAZING!!!

Question 10: Refer to my response to question 8, above. First you'd have to know what each descriptor means to each person using that word. To do this even remotely realistically, you'd have to have a universal understanding of the language being used: in order to meaningfully measure them, the words "warm," "bloomy," "dark," "pretty," and "ugly" have to have the same meaning and apply to the same things for all people... At which point they're no longer subjective qualities open to interpretation, they're objective qualities that are quantifiable.

Question 11: See answers 1-10, above.

Assuming that you've read through this whole response, again, I'm really not trying to be mean or rude or to pick apart your thesis just for fun. This is, in a weird way, coming from a place of sympathy: I have a lot of unpleasant memories of high school and college wherein teachers and professors just kind of told me, "you're wrong," without giving much explanation. This just left me feeling lost and kind of stupid. I guess this falls under the saying there are no stupid questions, only stupid answers. Maybe it could be better phrased as there are no stupid mistakes, only stupid reactions to mistakes. To me, education is equally about asking questions, making and learning from mistakes, and learning new ideas. I think your questionnaire was full of mistakes, my hope is to help you learn from those mistakes.

I really hope that my answers have been helpful. If you have more questions, feel free to ask; hopefully I'll be able to answer.


I appreciate the effort you put into critiquing the questions. There were a few goals I had in mind with this survey. The main objective was to get people's general opinion on what perceived sound quality is; and to gauge the value of researching potential ways to objectify sound quality, or to find measurements that correlate with subjective interpretations. In hindsight, I should have focused more on this. The other questions weren't really necessary and as you put it, not enough thought was put into them. Also, some of these questions might have worked better if I was more transparent. If I detailed the intent of the question the interpretation wouldn't be as vague.
 

Users who are viewing this thread

Back
Top