Head-Fi.org › Forums › Equipment Forums › Sound Science › Hearing tests
New Posts  All Forums:Forum Nav:

Hearing tests

post #1 of 5
Thread Starter 
Intro: At a recent meet I conducted a series of hearing tests mostly for my own curiosity. The test consisted of a few survey questions followed by two tests. The survey questions helped med figure out the demographic taking the test. The first test is aimed at "finding the smallest difference in sound level you can spot". The second test tries to "find the smallest difference in pitch (frequency) you can hear."

Hypothesis: My basic hypothesis was this: "Do people on Head-Fi have better hearing then most other people?" We as a group spend a LOT more on headphone and related equipment than the average person. I have friends that still mock me fore what I paid for some of my headphones. We justify the extra cost, but is this cost justification based on superior hearing or is this one way we become "cool" in the world? Also, our most basic method for determining if one piece is better than another is our hearing, but is this correctly founded?

It is also my hypothesis that the ability to judge audio equipment is related to how well you can tell sonic differences. One way to judge a sonic difference is with intensity (dB) and another way is with pitch (Hz).

Preliminary Analysis: I tried to standardize as much as I could before the test began. Everyone used the same equipment, sat in the same seat, wrote with the same pen, etc. Everyone was within 3dB of each other as a few wanted it turned up and a few said that it was too loud, but basically everyone was very close on the main volume knob on the amp.

Equipment
Headphones: Grado RS-1
Amp: Corda Opera
DAC: Corda Opera
Source: Wave files from the test played through Internet Explorer with all software volumes at 100%.

Problems
No test is perfect, and this one was no exception. First of all the room was noisy and that noise varied depending on when the test was administered. With my dB meter I measured the ambient noise from 60dB to 72dB depending on the time. The test was administered with open headphones because that is all I own.

I did not consider the noise to be much of a problem for one simple reason: This environment is the exact same environment that everyone else was in when they judged the validity of audio equipment. So if noise effected the test then it also equally effected the ability to judge if one piece of audio equipment is better than another. If it was not noisy enough to hear subtle difference from one audio cable to another, etc. then it was also not too noisy for this test.

I concede that ideally I would have liked to have to go a very quiet place. For this reason I did not try to do any frequency range tests as those are even more heavily dependent on volume levels. Also, the second test measured the ability to determine the difference between two tones (if any). If you could hear the tone over the ambient then you should be able to hear if it changes given a constant dB volume.

Demographic



Age Distribution
We were mostly made up of 21-30 year olds. This comes not no big surprise to me. In the second Age chart I am just displaying each age as a bar graph.

Audio Preference
I surveyed each participant as asked if they preferred audio equipment that was neutral (accurate), warm (pleasing), or both. A majority like both with a few more leaning towards warm sound.

Gender Makeup
The high majority of people surveyed were male. In fact it should be noted that all the female participants were there because of some male friend that brought them.

Dominate Hand
I was curious if the correlation between your dominate hand and side of your brain that is dominate would have any effects on these tests. Most people surveyed were right handed.

"One (the right brain) is visual and processes information in an intuitive and simultaneous way, looking first at the whole picture then the details. The other (the left brain) is verbal and processes information in an analytical and sequential way, looking first at the pieces then putting them together to get the whole." Therefore I wondered if this would effect the way we hear or perceive sound.

Hearing Opinion
I surveyed everyone asking what their own personal opinion of their own hear was. I asked them to rate this on a scale for 1 to 10 with 5 being average. The mean was 6.3 meaning slightly more than average we as a surveyed group consider our hearing slightly better than average. It was interesting that some chose as low as 1 and others chose as high as 9.

Hearing Opinion vs. Audio equipment preference
I made this chart by summing all hearing opinions (1-10 scale) for each Audio Equipment preference and comparing the three. Since each are split identically and the both category is right near mean as well there is no statistical correlation. Therefore based on your opinion of your own hearing you cannot say whether you will like neutral, warm, or both type of Audio Equipment.

Results



dB Test:

This test played a tone at 440Hz followed by a tone either +6dB, -6dB, or no change. Each participant was tested at this level 10 times. The scores would be tallied and a the correct guess out of a total possible 10 would be shown. Then they would proceed to the next level (3dB difference) and so on.

Since there were three choices for each guess out of ten choices a pure guess would average out to be 3 out of 10 or 30%. So any mean significantly above a random guess would have at least some statistical relevance. My imposed cutoff was 7/10 correct and you stop the test because any further and random guessing starts getting too significantly high. It is interesting to note that one person actually guessed 5/10 correct at the hardest level after getting only 1 correct the level before.

The two curved lines in the middle are polynomial trend lines of the highest and lowest scores. As a group our mean is somewhere bound in between those two curves. As a group everyone was able to hear the difference between 6dB and 3dB. 87.5% of the group could hear 1dB of difference. Only 25% could hear 0.5dB of difference. No one could hear below this statistically speaking as it would approach a pure guess.





Pitch Test:
This test played a tone at 440Hz followed by a tone either +50c, -50c, or no change (c=Cents). Each participant was tested at this level 10 times. The scores would be tallied and a the correct guess out of a total possible 10 would be shown. Then they would proceed to the next level (20c difference) and so on.

The pitch tests results were a little bit wider. This might imply that it was either a better indicator or less effected by the ambient noise.

Again, since there were three choices for each guess out of ten choices a pure guess would average out to be 3 out of 10 or 30%. So any mean significantly above a random guess would have at least some statistical relevance. My imposed cutoff was 7/10 correct and you stop the test because any further and random guessing starts getting too significantly high. It is interesting to note that one person actually guessed 6/10 correct at the hardest level after getting only 1 correct the level before.

The two curved lines in the middle are polynomial trend lines of the highest and lowest scores. As a group our mean is somewhere bound in between those two curves. As a group everyone was able to hear the difference between 50c changes. 93.75% of the group could hear 20c pitch changes. 68.75% could hear a 10c pitch change. Only 31.25% could hear a 5c pitch change. No one could hear below this statistically speaking as it would approach a pure guess.

The two tests can be found here. If you took the test and would like your scores send me a PM. I did not publish names with any scores on purpose.

Second Phase of testing
Next, I plan on next testing my Barbershop singing chorus' hearing and comparing the results. Finally, I want to sample a random "normal people" group for final comparison.

Questions?
Feel free to PM me. I would appreciate all input.
post #2 of 5
Nicely done!

My only suggestion would be using a set of closed back phones to improve isolation and help with the ambient noise issue.
post #3 of 5
Quote:
Originally Posted by flashnolan View Post

Hypothesis: My basic hypothesis was this: "Do people on Head-Fi have better hearing then most other people?" We as a group spend a LOT more on headphone and related equipment than the average person. I have friends that still mock me fore what I paid for some of my headphones. We justify the extra cost, but is this cost justification based on superior hearing or is this one way we become "cool" in the world? Also, our most basic method for determining if one piece is better than another is our hearing, but is this correctly founded?

It is also my hypothesis that the ability to judge audio equipment is related to how well you can tell sonic differences. One way to judge a sonic difference is with intensity (dB) and another way is with pitch (Hz).
I think it is an interesting test (actually, quite an impressive piece of work), but I don't think that listening to test tones and attempting to discern differences in intensity and pitch is necessarily a good indication of a person's ability to discern differences in how a particular musical passage sounds when it is played through different equipment. It's probably not a perfect analogy, but would the ability of oenophiles to tell the differences between various wines be fairly tested if they were confined to dipping the tip of their tongue in various wines for a couple of seconds and then asked to state how the wines were different?

Note that I'm not saying "audiophiles" actually have better hearing, or anything about the extent to which people can discern subtle differences in musical passages when played through different equipment, etc. I'm just saying that listening to test tones is different than listening to music, IMO.

Nevertheless, it is an interesting test and a nice presentation.
post #4 of 5
Thread Starter 
Quote:
Originally Posted by PhilS View Post
I think it is an interesting test (actually, quite an impressive piece of work), but I don't think that listening to test tones and attempting to discern differences in intensity and pitch is necessarily a good indication of a person's ability to discern differences in how a particular musical passage sounds when it is played through different equipment. It's probably not a perfect analogy, but would the ability of oenophiles to tell the differences between various wines be fairly tested if they were confined to dipping the tip of their tongue in various wines for a couple of seconds and then asked to state how the wines were different?

Note that I'm not saying "audiophiles" actually have better hearing, or anything about the extent to which people can discern subtle differences in musical passages when played through different equipment, etc. I'm just saying that listening to test tones is different than listening to music, IMO.

Nevertheless, it is an interesting test and a nice presentation.
This is exactly why I have put this into my hypothesis - it it just that right now an unproven hypothesis.

I think your analogy has some merit. There is definitely a big difference listening to 2ms of an audio clip as opposed to 1s or 10s. When audio is "sliced" too thin it sounds weird just like the wine might taste different then actually drinking it. Then again real wine testers don't drink wine anyway - they spit it out!

If you stick your tongue into wine you can still tell a lot about it. Is it acidic, sweet, etc.? The same is with listening to a test tone. It might not be able to tell me the full body taste of the sound, but if I cannot tell it is sweet or sour how can I be a proper judge of the overall taste? I think listening to a test tone is an important major component that is testable to determine hearing quality, but I agree it is not a 100% indicator. There is more involved which the "tongue dip" cannot test for.

Plus your illustration breaks down due to the fact that only your tongue is tasting the wine and not involving your sense of smell which is the major part of tasting anyway. When listening to a test tone you are using all of your hearing faculties.
post #5 of 5
Thread Starter 
Quote:
Originally Posted by Joelby View Post
Nicely done!

My only suggestion would be using a set of closed back phones to improve isolation and help with the ambient noise issue.
Well then buy me a pair!

Seriously though a closed pair would properly represent the testing environment of the meet with a closed pair of headphones. The open pair I used properly represented the testing environment of this meet for a person with open headphones. My thinking is this: how can anyone properly compare open headphones at a meet?

I have heard many people name extremely subtle differences that they can hear when comparing to different things while they are wearing a pair of open headphones. There are only a two conclusions as I see it:
A. They really did not hear any differences and only thought they did
B. They could hear differences and the ambient noise in the open headphones was not enough of a detriment effect

So in either choice for open headphones this is a valid test.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Hearing tests