crinacle's IEM FR measurement database

[csglinux]

Is it possible resonances could be the effect of the coupler/tubing and not present when inserted into a real ear?

No. There can be discrepancies in amplitude and frequency, but if there's a resonance in the coupler, the same thing will happen in your ear canal.

would it though? are resonances solely the result of the air volume? If so, I'd say you're right. But I thought resonances are dependent on material type, volume, and geometry - probably other things I'm missing.

My point is, a perfectly nice cylinder really is nothing like an ear canal when you think about it.

Resonance is just energy trapped or accumulated at a particular wavelength. The length (and volume, if you have transverse modes) plays a role, but the whole point of the coupler is to mimic the length and volume of the average ear canal. Likewise, the materials used (@crinacle's vinyl coupler) will mimic the inside of the ear canal - which also isn't perfectly reflecting. Measurement differences aren't an error. The fact that you'll hear something slightly differently to me, doesn't mean your ears have an error - it just means they're slightly different.

There are much larger sources of discrepancy from compensation curves for mic, soundcard, and shift to some (arguably debatable) target loudness curve, and the variation in mic sensitivity as a function of frequency.

 

[SilverEars]

If the purpose is to identify resonance peaks, that could be done by varying insertion depth within a single coupler, no? (And that could be useful information to prospective IEM owners, because they'd know they can also shift those peaks with shallower/deeper insertion.) But again, this isn't an error.

Is there another source of "error" I'm overlooking?

This is quite interesting. I'm curious how much variance there is in people's canal shape and the bend. And at same insertion depth, how much variance there can be.

How will we know the differences? We'd need to get a good respentation samples of the canal.

How is the avg figured out?

If there is a way that is undoubtably foolproof to figure out the resonance due to depth and points in the canal, you can know based on most common shape or median. What extremes and differences are possible and how common the features can be, and thus how the FR would reflect as result I'm curious about.

 

[bartzky]

That's actually a pretty old topic and has been discussed and researched for decades now.

Is it possible resonances could be the effect of the coupler/tubing and not present when inserted into a real ear

That coupler would be pretty bad then, wouldn't it? A good standard coupler represents the average human ear below 10 kHz or even above, if go for the new (not so cheap) GRAS. That being said, not everyone's ear is average. It might be that the measurement show a 7 kHz peak, but you hear it at 8 kHz and your neighbor at 6.5 kHz. So the peak may be shifted but it will appear.
Usually we expect two more or less relevant peaks caused by the closed ear canal. One at about 6-9 kHz and another one at around 12-15 kHz depending on the deepness of the fit. The deeper the fit the higher the frequency will be.
If you look through my database you will see those two peaks in nearly every measurement. They sometimes may be masked by other peaks nearby or shifted higher due to very deep fit though.

My point is, a perfectly nice cylinder really is nothing like an ear canal when you think about it.

One of the first couplers, if not the first, was created about 100 years ago by Bell for telephone testing. It was made of natural caoutchouc and looked and felt similar to a real ear. But ageing was a serious problem. The clever solution: A coupler doesn't have to look or feel like a real ear, it just needs to sound like an ear! So people started making couplers from durable materials like stainless steel that doesn't look like ears, but have the same acoustical properties.

 

[james444]

This is quite interesting. I'm curious how much variance there is in people's canal shape and the bend. And at same insertion depth, how much variance there can be.

How will we know the differences? We'd need to get a good respentation samples of the canal.

How is the avg figured out?

If there is a way that is undoubtably foolproof to figure out the resonance due to depth and points in the canal, you can know based on most common shape or median. What extremes and differences are possible and how common the features can be, and thus how the FR would reflect as result I'm curious about.

Here's a video of Paul Barton touching on this subject. Length, diameter and shape can vary dramatically:




Here's a study that found a significant gender related difference in resonant peaks, due to the fact that females tend to have shorter ear canals than males:

http://www.scielo.br/scielo.php?pid=S1808-86942014000100041&script=sci_arttext&tlng=en


And here's a most extensive study on all sorts of things that may affect FR, like ear canal length and volume, ear canal shape, ear drum impedance and the stapedius reflex. They even built two kinds of adjustable ear canal simulators to study a large variety of different sizes of artificial earcanals:

https://pdfs.semanticscholar.org/77de/b9c44b683417477f62873a9b82bde27e378a.pdf

 

[castleofargh]

Here's a video of Paul Barton touching on this subject. Length, diameter and shape can vary dramatically:




Here's a study that found a significant gender related difference in resonant peaks, due to the fact that females tend to have shorter ear canals than males:

http://www.scielo.br/scielo.php?pid=S1808-86942014000100041&script=sci_arttext&tlng=en


And here's a most extensive study on all sorts of things that may affect FR, like ear canal length and volume, ear canal shape, ear drum impedance and the stapedius reflex. They even built two kinds of adjustable ear canal simulators to study a large variety of different sizes of artificial earcanals:

https://pdfs.semanticscholar.org/77de/b9c44b683417477f62873a9b82bde27e378a.pdf

the last link is really cool, while approaching the questions mostly with the ear canal as a nice tube(which is a simplified approach), I had never seen such extensive testing.

now one question came to me: how is it possible for me to have about 1000 more "likes" than you have? you're consistently such a helpful and all around cool guy while I'm being a jerk on half my posts. is it a bug? were you evil before and embraced the carebear philosophy only in the last few years where I've known you? or is it that deep down, Headfiers just like jerks more? there is a mystery right there.

 

[SilverEars]

Did I hear this right? Ear drum size differences being different? Here I'm making a difference between ear canal and ear drum. Is he actually meaning ear drum? If so, what consequinces that has? Because, there is only one consistant microphone to emulate the various drums.

 

[1TrickPony]

the last link is really cool, while approaching the questions mostly with the ear canal as a nice tube(which is a simplified approach), I had never seen such extensive testing.

now one question came to me: how is it possible for me to have about 1000 more "likes" than you have? you're consistently such a helpful and all around cool guy while I'm being a jerk on half my posts. is it a bug? were you evil before and embraced the carebear philosophy only in the last few years where I've known you? or is it that deep down, Headfiers just like jerks more? there is a mystery right there.

Moderators.
Are.
Overrated.



;-x

 

[castleofargh]

Moderators.
Are.
Overrated.



;-x

I don't remember when I became quasi-modo, but indeed you're making a good point.
lol, I didn't even consider people's attraction toward uniforms. silly me. being so directly associated with the Sound Science section didn't feel like it won me unilateral adoration from Headfiers TBH. I felt like more people hated my guts even before I got a chance to prove them right.
whatever the impact or impacts, this completely messes up my idea that we were similar for having a close enough number of posts.

Did I hear this right? Ear drum size differences being different? Here I'm making a difference between ear canal and ear drum. Is he actually meaning ear drum? If so, what consequinces that has? Because, there is only one consistant microphone to emulate the various drums.

the eardrum varies in size with people and it grows with us from child to adult. it's likely that different eardrums will react differently on various variables. one of the tools used to check the state of the eardrum and other things is tympanometry, where they basically check how the eardrum moves in response to a given stimulus. it's obviously used to look for issues, but variations between individuals are still expected. even the acoustic reflex doesn't have the same strength or triggers at the exact same level of sound pressure for different people.
what everybody is aiming at with measurement standards is a model for average human behavior, because that's what more people will be close to. but we are physiologically different and will keep on being until we get a well deserved clone army.

 

[vladstef]

Right now, I think that it would be pointless to try and go into details about canal shapes, eardrum shapes etc. especially since an 'average' already exists. The method described in 'An Acoustic Basis for the Harman Listener Target Curve' by innerfidelity is very good imho and enough. Even perfectly flat speakers in a perfect room with perfect positioning will sound different depending on particular ears but us individuals are used to our ears and how things sound and this should translate well to headphones.

The only issue for me would be peaks in treble which form due to our particular canals. I recently got IMM6 and experimented a bit with couplers - I tried to get the particular peak in treble to align with how I'm hearing it when using sine gen on my ears. To be honest, it was nowhere near compared to how @crinacle measures IEMs in this way - 10 to 14mm of distance from eartip to microphone is way too short if tube is around 8mm, it takes around 22-23mm for the peak to align. Now, this is obviously my personal experience with this and I'll repeat that I am just trying to contribute to this conversation - I still trust this database of measurements a lot more than any of mine due to how little experience I actually have. Having said this, I would like to point out that in terms of volume, 20+mm of tubing is much closer to 711 coupler and I am very conflicted - is my eardrum that deep inside my head or is my canal wider than 8mm?

And my last point about all the variation about canals - the only way one could have a perfect response would be to measure their own canal and adjust fr response accordingly - the condition being that you are getting a perfect fit with your iems in any given moment. Until we have some sort of futuristic technology that could do this every time you insert an IEM, all of this is pointless and even then, there will be purists who wouldn't want their music tainted by additional software changes, me included.

Just to repeat again because I am really not trying to start a fight - I am fairly inexperienced and my 'knowledge' could very well be misinformation/misunderstanding and I am open to suggestions and other facts that would change my mind and educate me.

 

[1TrickPony]

I don't remember when I became quasi-modo, but indeed you're making a good point.
lol, I didn't even consider people's attraction toward uniforms. silly me. being so directly associated with the Sound Science section didn't feel like it won me unilateral adoration from Headfiers TBH. I felt like more people hated my guts even before I got a chance to prove them right.
whatever the impact or impacts, this completely messes up my idea that we were similar for having a close enough number of posts.

Bah. You never came off like some prick with a badge. Another former moderator had so much crap come his way because he was simply calling for it imo.

You probably deserve your everything that came your way. Haha.

 

[ForceMajeure]

del

 

[james444]

Moderators.
Are.
Overrated.



;-x

Likes.
Are.
Overrated.

Bah. You never came off like some prick with a badge. Another former moderator had so much crap come his way because he was simply calling for it imo.

You probably deserve your everything that came your way. Haha.

+1

And my last point about all the variation about canals - the only way one could have a perfect response would be to measure their own canal and adjust fr response accordingly - the condition being that you are getting a perfect fit with your iems in any given moment. Until we have some sort of futuristic technology that could do this every time you insert an IEM...

From what I've been gathering, there's quite a lot of research going on regarding personalization of HRTFs. The main driving force not being demanding audiophiles, but the requirements for spatial audio in virtual and augmented scenarios. The thing is that averaged HRTFs have proven insufficient to achieve precise sound localization with headphones / IEMs.

Here's an interesting article on the development of an electronic travel aid for the blind. The team started out using HRTFs from the CIPIC Database, but soon realized they needed personalized HRTFs to achieve satisfactory sound localization: https://www.eurasip.org/Proceedings/Eusipco/Eusipco2007/Papers/d4p-h05.pdf

And here's Microsoft's site on spatial audio and HRTFs: https://www.microsoft.com/en-us/research/project/spatial-audio/

Bottom line: I'm pretty confident that we'll get that futuristic technology at some point down the line...

 

[money4me247]

now one question came to me: how is it possible for me to have about 1000 more "likes" than you have? you're consistently such a helpful and all around cool guy while I'm being a jerk on half my posts. is it a bug? were you evil before and embraced the carebear philosophy only in the last few years where I've known you? or is it that deep down, Headfiers just like jerks more? there is a mystery right there.

From my personal experience, here are the top ten secrets of getting an insane like count on head-fi:
1) Reviewing! write formal reviews of something hotly anticipated & have the review end up on the front page (this is the most high yield). even better bonus is that when you do it often enough, you will have random people offering you "free stuff for your unbiased opinion." Don't worry, the close relationships and free products definitely never turns reviewers into more-or-less unofficial undercover advertisers for increasing product hype/visibility. All the good reviews have that "my unbiased opinion" disclaimer and I definitely for sure remember reading front page review that just completely trash the product being showcased which ensures that there is a fair balance of positive and critical reviews being showcased.
2) Flame wars: jump into as many as possible meaningless head-fi debates that randomly rage throughout the forums with simple & calm truisms that helps calm everyone down "everyone's impressions are valuable, everyone hears differently, there is room for both objective & subjective thinking. interesting points from both sides." ...etc, etc. (definitely second most high yield place to farm for likes)
3) Hype the hype: jump into the fan-thread of a product with exorbitant and unrealistic praise of that specific product, bonus points if you say that it beats out XYZ more expensive and more popular product. moar bonus points if you do it close to launch or even pre-launch when no one else has access to the product but when there are lots of people following the thread.
4) Unhype the hype: jump into a thread overflowing with hype and write 'honest' impressions (aka don't go crazy with hype & taper the unrealistic expectations & tactfully mention some flaws... these types of impressions can be strangely harder to find than you realize). bonus points if you accidentally call into question someone else's audiophile cred, which can lead to #10.
5) Versus mode: write a lot of comparisons, especially when someone is requesting it. especially helpful is direct head-to-head comparisons of competitor products in a neutral tone with list of pros/cons of each which can help people who are trying to decide between products. Everyone loves a good smack-down comparison fight. Bonus points if you say something outrageous & move onto step #10.
6) Be helpful: give advice when asked (bonus points if you are able to base the advice more on the asker's wants/needs rather than your personal preferences/likes)
7) Be unhelpful: give advice when not being asked. bonus points for making controversial comments on items that you have not heard and/or have no experience with. super bonus points if you make a subtle dig at someone's audiophile credibility and/or hearing ability and/or preferences and/or belief system. this will often lead to #10.
8) Data points: post measurements with a disclaimer, this used to be a lot less common & would often lead to #10. even better if you mention doing a blinded ABx comparison and/or volume-matching for good measure. will lead to #10 the fastest when it directly contradicts someone else's personal impressions.
9) Omnipresent: continually follow a product thread & make appropriate comments according to the eternal hype & unhype cycle
10) Farming during conflict: accidentally start random flame wars & then do #2-9

Good luck & have fun!

 

[crinacle]

And so the reset button has been pressed. Equipment swap, software swap, format enhancements.

iPhone dongle (rated at 0.1ohm output impedance) -> TRRS splitter -> high efficiency capsule microphone -> IEC60318-4 compliant inner ear simulator

App used is AudioTools. Expensive, but much more customisable than FFTPlot. Due to the octave scaling, I'll have to use pink noise as opposed to the white noise that's used on the decade scaling of FFTPlot. Those who require the Periodic Pink Noise file that I've created, just holler.

Measurements will be taken at roughly 3mm away from reference plane. This will create a resonance point at around 8000Hz that should generally be ignored.

Some new sample data to show y'all what's about to go down from now on.

Hyla CE-5
NCM Bella
Alpha & Delta D6
AAW W900 Universal
TFZ King Pro
Rose Mini2
Rose BR5 Mk2

Yes, I am aware that the bass looks spiky. Live with it, just extrapolate a smooth line in your mind, I can't fix that myself.

Good thing now is that I also have text data that can be saved so I can recall any previous graphs to be overlaid on command. That's something pretty sweet. Those who want the raw data just feel free to ping me.

Future measurements under the new format and equipment will be consolidated under this imgur album.

 

[csglinux]

And so the reset button has been pressed. Equipment swap, software swap, format enhancements.

iPhone dongle (rated at 0.1ohm output impedance) -> TRRS splitter -> high efficiency capsule microphone -> IEC60318-4 compliant inner ear simulator

App used is AudioTools. Expensive, but much more customisable than FFTPlot. Due to the octave scaling, I'll have to use pink noise as opposed to the white noise that's used on the decade scaling of FFTPlot. Those who require the Periodic Pink Noise file that I've created, just holler.

Measurements will be taken at roughly 3mm away from reference plane. This will create a resonance point at around 8000Hz that should generally be ignored.

Some new sample data to show y'all what's about to go down from now on.

Hyla CE-5
NCM Bella
Alpha & Delta D6
AAW W900 Universal
TFZ King Pro
Rose Mini2
Rose BR5 Mk2

Yes, I am aware that the bass looks spiky. Live with it, just extrapolate a smooth line in your mind, I can't fix that myself.

Good thing now is that I also have text data that can be saved so I can recall any previous graphs to be overlaid on command. That's something pretty sweet. Those who want the raw data just feel free to ping me.

Future measurements under the new format and equipment will be consolidated under this imgur album.

I much prefer the look/format of your new graphs What's the new sotfware you're using?

I hate to keep banging on this same old point, but why should we ignore the 8 kHz peak? If your distances are anything like those in an actual ear canal, it'll be there. And if not, it'll still be there, but at a slightly different frequency. What am I missing?