Audiologist Hearing Test

[Flint]

In my other thread about having custom sleeves made for my Shure SE846 IEM earphones, I mentioned that the audiologist I worked with also performed a hearing test on me. I wanted to share what I learned:

The audiologist put some funky testing IEM-like things in my ears. She used the same squared off ear-canal sleeves musicians often use with commercial IEMs, but the earphone device was different, a box with two tubes and a cable coming off it. The tubes were for the sound and the cable plugged into her computer. After having the audio testing thing inserted and connected, she handed me a really handy thumb trigger which was also connected to her PC and she sat behind me and ran the test. I was to press the trigger whenever I heard the test noise - a “tat-tat-tat-tat-tat” at different sine wave pitches.

As she started the test I thought I was doing pretty well, but after just a few seconds she spoke into her PC mic and told me she was using a different test tone, a narrow-band noise with the same repeating character. The narrow-band noise was centered at different frequencies and different volume levels and for about 5 minutes she tested my right ear before proceeding to my left ear.

After the test was completed and the apparatus was out of my ears, she told me she had to switch to the narrow-band noise because I clearly had tinnitus (ringing in my ears) and when the test tone sine wave happened to be at the exact frequency as one of the frequencies my ringing is, I cannot hear it at low levels. She said that was common with musicians and construction workers.

Then she printed and handed me the results of my tests, and this is where it gets odd, to me.

She explained that from 500Hz to 2,00Hz my hearing ability was above the average for people aged 45-55, and comparable to a person aged 25-35. “Great!” I say.

From 3,000Hz to 5,000Hz my hearing ability is on average for a person my age.

Around 6,000Hz my hearing is lower than the average for people my age.

Above 8,000Hz my hearing is the best she has measured, comparable to a teenager.

Why do I find this surprising?

Well, I have played drums since I was a 9 year old boy. I played for several hours just about every day. It wasn’t until I was working full time as a session drummer and playing very loud gigs that I started wearing earplug or early IEMs to protect my hearing. I also was a hunter and shooter. I would spend my summers hunting rabbit with a .22 rifle and a .38 revolver and deer with a 380 rifle and shoot skeet with a 12-gauge shotgun. I was shooting since before I was a drummer, and it was rare that I wore earplugs while hunting (I did wear hearing protection when target or skeet shooting). I installed some of the biggest and loudest car stereos in Southern NM back in the early 1980s. In my studio and other studios I worked in, I would turn up the control room speakers to hear all the minutest details of the recordings I was making. And at home and with headphones I prefer to listen louder than just about anyone I know who claims to be an Audiophile. I have had permanent tinnitus since I was about 16 and it gets a little worse every year, but it had never given me headaches like some people complain about. Even now, when I tune my drums or write a drum part for a song, I don’t wear earplugs as I need to hear what my drums sound like in order to know what to do to make them right for the recording.

My hearing should be shot. I mean, really shot… like nearly deaf. My dad, who destroyed his hearing from shooting without protection most of his life, was what I envisioned my future being.

That said, ever since I was about 25 years old I have always carried “musicians” earplugs around with me, like the Etymotic 12dB models, which I put in at concerts or loud bars. I wear my sound blocking IEMs on planes, when riding the Metro, or even when walking around the city, which allows me to turn down the music and still hear all the detail over the noise of the world.

But I am shocked!

Near perfect hearing above 10kHz? She tested me clear up to 20kHz and said it was amazing – something I bragged about to people with an internal birdy telling me I was full of shit because that shouldn’t be possible.

She didn’t test below 500Hz, but I am confident in that.

Maybe… just maybe, that’s why I am so sensitive to shrill speakers or headphones which artificially boost the output above 10kHz to make them sound “clearer” or “more detailed”. For me they are like needles piercing my eardrums. Maybe my slight deficit at 6kHz is why I am so keen to have the best possible midrange performance possible – that is the frequency where consonants are recognized by the listener (I will use that on my wife when I am ignoring her speaking to me – “Honey, you know I am hard of hearing at the speech recognition frequencies.”).

Whatever the case, I am amazed my hearing tested as well as it did. I hope I can take care to protect it going forward as my love of music is unlimited and I cannot imagine a world without it.

 

[PaulyT]

Very interesting, thanks for the writeup. I know I have some tinnitus too - I hear a constant low-level ringing, or really more of a steady high-pitched whine, which gets worse when I'm tired. I've never been tested like that though across the frequency spectrum, only the basic tones of a simple hearing test.

This sort of thing always makes me question the desire to get completely flat response in a speaker/headphone system. I mean, if our individual hearing is all over the place, why even bother? I guess the counter-argument is that you want to hear your speakers the same way you hear "natural" sounds - that even if your internal hearing response is different from another person's, your own ears will respond the same way to speakers vs other sounds and you want them to match. Ok fine. But to me it just means that our hearing, our perception of sounds, can adjust to our ears' different responses (e.g. our brain compensates for changes in ears' response), that having something that's flat by a measurement mic just isn't all that critical. We just make a big deal of it because it's easy to measure.

Ok ok I'm fanning flames here, but these are the sorts of thoughts I always have when we talk about stuff like this. Just being honest.

 

[Flint]

I whole-heartedly support the argument that a speaker should be completely neutral. When a speaker is reproducing, say, an acoustic guitar, it should sound the same as if there were someone playing the acoustic guitar in the room. That means a relatively flat response is necessary.

Dr. Floyd Toole did hundreds of scientific listening perception studies when he was working with the Canadian Government research ministry and found that most people found a flat on axis response and flat power response gave the listeners a better perceived experience when listening to natural music, like acoustic instruments recorded to sound as they do in real life. That led to the creation of the "Canadian sound" which led to his working at JBL and Harman International and creating many lines of great speakers which were both affordable AND sounded pretty darn good to most people.

In my experience, people who like the artificial coloring of boosted treble or bass are wanting an experience they cannot get in real life - more detail than normal, more visceral boom than normal, etc. They are using audio to create an unreal experience which is like crack cocaine, LSD, or Ecstasy on their auditory senses.

 

[PaulyT]

... found that most people found a flat on axis response and flat power response gave the listeners a better perceived experience when listening to natural music

Ok but WHY? Why do people prefer that? Is it simply because that's what they're used to, at the moment? What if they listened to non-flat response for some period of time (minutes, hours)? Would their experience then be any better than with a flat response?

Dunno why I'm arguing this. I really do basically agree with you, that flat response is better. But I just think we can get used to non-flat response, and that maybe there are other things (dynamics, proper acoustics, etc.) that are ultimately more important over time. They're just harder to measure and understand.

 

[Deleted member 133]

Dr. Floyd Toole did hundreds of scientific listening perception studies when he was working with the Canadian Government research ministry...

Cripes! You make it sound positively commie-derived Monty-Pythonesque!

For the record, it is the National Research Council of Canada. It's the same as the US' National Research Council except that it's a government agency and it actually does research!

For the record, the NRC that Floyd Toole worked at many years ago is gone; dead, destroyed by Stephen Harper-backed Albertans who knew nothing about research, or running a research organization, or any other organization for that matter. It's chief architect disappeared from sight (on some form of long-term medical leave according to rumour) after Harper was trounced by Trudeau (who vowed to unshackle Canada's scientific community) in last year's federal election, and a rebranding designed by the incompetent Albertans was cancelled at the eleventh hour this year, and right now the whole place remains pretty much in a state of chaos (which it had been for the last five years or so). It's in critical condition and will take years to recover - if it's lucky. Textbooks should be written about the travesty carried out by the cowboys from the West on the NRC.

Floyd Toole would have never been able to do the incredible work that he did in today's NRC.

Jeff

ps. My apologies to Albertans - at least any who are not incompetent like the bunch that came to Ottawa. Well not really; they actually never really came to Ottawa: they just commuted into town every so often - maintaining their Alberta residences.

 

[Flint]

You'd need to read the papers Dr. Toole created from his research. It took place in the 1980s and 1990s and was very, very thorough - real proper research as opposed to limited listening tests with a few audiophiles.

There were many things he compared, including response, power response, dynamic range, harmonic distortion, phase delay, etc. The ultimate result was a reference design speaker called the Energy 8 which had an 8" woofer, a 1.1" soft dome tweeter (ralatively new design at the time) and a linkwitz-riley crossover at 1,800Hz. That led to the formation of the Energy speaker company and many speaker companies which got access to all the research he did like Paradigm and the other Canadian big boys.

While it is true that frequency response is one important characteristic, other also play a big role like dynamic response, phase shift, quality of crossovers, cabinet acoustics, and so on. But while most of the other components are very difficult for a layman to understand, frequency response is pretty easy. That's why Audio magazine and later Stereophile do the tests they do, which always include a frequency response plot. Power response is too variable for a simple chart, but JBL tried to standardize a simple chart back in the late 1980s with little success. Today a polar plot is most often used for power response if you want to really understand how the speaker performs in a room.

But it really isn't hard to understand - if you build a two-way speaker with a tweeter which is 10dB louder than the woofer (as they tend to be before applying a good crossover pad), the sound will not be pleasing - even if you get everything else absolutely perfect. There was a period when graphic EQs were a big deal to every buyer, during my generation of teen audio nuts, and the result was chaos which darn near all of us outgrew and the demand for such a feature died. Speakers got better sounding every year and the need to make up for their deficiencies grew weaker. It took longer with car audio, but even that isn't about stupid uneven response curves anymore, just loudness and bass from what I can tell.

Now, by flat response, that doesn't mean the same output at 20Hz as at 20,000Hz. That means lack of peaks and dips in the curve from bass to treble. In reality, the ideal room response (often called the room curve in theater design) is the idea where the bass is louder than the treble and the response is linear, but tilted downward, as you go from 20Hz to 20,000Hz. This is a reflection of the room interactions, dispersion, power loss over distance, resistance per frequency in the air, etc. When measuring a speaker in an anechoic environment what looks like a ruler flat response often turns into a slightly downward slope curve in most real world installations. In my case, and that of a few others I work with or know, the ideal is a curve which slopes downward about 1dB per octave from 100Hz to 16,000Hz. And ideally as smooth and free from dips and peaks as possible.

While that curve is "ideal" it cannot come at the cost of poor phase, dispersion, dynamics, high THD, etc. It is all one huge balancing act, but the science, materials, and manufacturing skills invested in driver design has vastly improved the quality of the woofers, midranges and tweeters available on the market where many of the important aspects of the sound must be addressed. They have also become more affordable, so it doesn't cost an arm and a leg to get good sound from affordable parts. Modeling software, digital processing, and testing tools have also gotten much better, making it easier to design a good speaker system using those vastly superior transducers. All in all it is a matter of cost, time, quality, and target performance characteristics which separate one speaker from another. And since there are thousands of choices and even factories which can affordably produce custom transducers at moderate quantities, the options and variations are limitless.

So, what we have at our fingertips are dozens of ways to measure the important aspects of what sounds good to most people (and the key word is "most"). Of those measurements are group delay (phase), phase, waterfall plot (dynamics), Impulse response (dynamics and offset delay), frequency response, power response (dispersion) and polar plots (dispersion).

There you go.

 

[Flint]

Cripes! You make it sound positively commie-derived Monty-Pythonesque!

For the record, it is the National Research Council of Canada.

It worked! I was too lazy to go look it up so I threw a red herring out there for you to correct for me and clear it all up. Thanks, man.

I also think the work that was done by Dr. Toole at the NRCC was invaluable, and probably didn't need to be continued as we are still learning from it and applying it today.

 

[malsackj]

The other side is the human, genetics, and what they are hearing. Each of us are different and have person choices as to what is good sound. This is what makes the market so diverse with products.

The problem is also lack of the general populous to learn and understand the ear and hearing mechanism, protect it for good hearing. We can not fix it with a new pair of eye glasses.

 

[Flint]

The other side is the human, genetics, and what they are hearing. Each of us are different and have person choices as to what is good sound. This is what makes the market so diverse with products.

The problem is also lack of the general populous to learn and understand the ear and hearing mechanism, protect it for good hearing. We can not fix it with a new pair of eye glasses.

Good points.

On the first - what each of us considers good sound - I am reminded that the characteristic of a sound which each of us focuses on to determine accuracy, or realism, or pleasing-ness, that thing we chose to pay the most attention to, that is what makes many of us so different. For me it is the detail in the crucial midrange characterized by high dynamic accuracy and lack of extreme aberrations in frequency balance. When that is right I am happy. For others is it bass extension or "enhanced" detail (often through boosted treble and a dip in the response in the midrange). I get that and appreciate it.