Head-Fi.org › Forums › Equipment Forums › Sound Science › How far can EQ really go towards truly equalizing headphones?
New Posts  All Forums:Forum Nav:

How far can EQ really go towards truly equalizing headphones? - Page 11

post #151 of 204
The general idea of resonances in the high frequencies can be captured with a dummy head but it can't reproduce the exact frequencies and amplitudes of the resonances when the phones go on YOUR head.
post #152 of 204
Quote:
Originally Posted by Joe Bloggs View Post

The general idea of resonances in the high frequencies can be captured with a dummy head but it can't reproduce the exact frequencies and amplitudes of the resonances when the phones go on YOUR head.

 

Agree.

post #153 of 204
Quote:
Originally Posted by ultrabike View Post

Not sure if acoustic impedance effects can be captured in the measurements if a dummy head or appropriate measurement coupling is used.

 

I don't see why they wouldn't be reflected in the measurements when using a dummy head. It would mostly be reflected in the resonance and CSD waterfall data, I expect. 

post #154 of 204

I think so. Like Joe Bloggs said, it may an approximation given individual differences (i.e. the dummy head and my dummy head smile.gif.) But overall, I feel some (hopefully most) of it would be accounted for.

post #155 of 204
Quote:
Originally Posted by ultrabike View Post

Not sure if acoustic impedance effects can be captured in the measurements if a dummy head or appropriate measurement coupling is used.

It's pointless to use anything other than the head which will listen to these headphones for measurements, as ear canals have very high deviations. Even headphone position would produce high difference. While acoustic impedance plot can be obtained for a headphone out of context, it only shows how much flawed a HP will be in the 3-14kHz region in general, and won't allow one to see how it will interact with his own ears and hence compensate for that.

post #156 of 204

This may partially account for individual preferences. Perhaps this is one reason why it may be difficult to equalize an IEM to sound like a full size open. However, two same class (say full size open) headphones might not be too hard to equalize, provided the equalized headphone is well behaved, given they face similar coupling (by observing frequency response differences).

 

I also feel that when equalizing and/or evaluating performance (for purchase consideration), measurements should serve as a guide. One should also use one's ears to see if things sound right.


Edited by ultrabike - 7/4/12 at 3:17am
post #157 of 204

wandering off the rails again I see

 

for circumaural headphones personal ear cavity resonances are not the target of EQ - your brain already compensates for your personal physical differences, whatever your personal internal subjective experience it is the response to the external sound field that is the shared stimulus

 

Smyth SVS Realiser gives impressive "EQ" duplication of the sound of loudspeakers and room by "calibrating" with mics placed at your ear canal opening - its not trying to EQ at your ear drum or compensate for ear canal resonance - your brain already does that

 

only iem have to consider the effects of ear canal variations causing additional frequency response effects since they modify the cavity resonance by their physical intrusion

 

 

also people really should get to the 2nd page on the the innerfidelity repeatability article:

 

"Once we approach about 3.4kHz the variability does really begin to become a strong function of headphone placement. The resulting 2x standard deviation is now 1.8dB. Again, this would mean, should we measure the same pair of HD800s 1000 times (and place the headphones on the head in between each measurement) we could expect 950 measurements would be within 1.8dB.

However, once we reach higher frequencies (>8.4kHz), the variability really begins to become strongly affected. The average standard deviation in this region is now almost 5dB. So yet again, variations of +/-5dB are to be expected in this region. The maximum variability is approximately 18dB, higher still. So it appears that the slightest headphone placement on the head can significantly change the high frequency resonances in the dummy head and drastically reduce the measurement system's ability to precisely measure higher frequencies.


Edited by jcx - 7/4/12 at 5:37am
post #158 of 204

The first and second part of your rebuttal seem to contradict each other.  On the one hand you seem to say that circumaural headphones have effectively no impedance mismatch with the ear canal, on the other hand you are saying that the slightest change in positioning of the headphones can change the high frequency FR drastically, which seems to indicate the center frequencies of *some* kind of resonance being affected by positioning.

post #159 of 204

read the innerfidelity article in full yet? http://www.innerfidelity.com/content/expert-tests-innerfidelitys-headphone-measurement-repeatability-and-reproducibility-page-2

 

 

yes, the world is complicated, particularly wave propagation, coupling where wavelengths vary from <1/100 the cavity dimensions to 10s of wavelengths in some direction (varying by factors of >3)

 

especially when what we really want to emulate is far, free field propagation interactions, interference with the head, ears, pinnae, ear canal

 

its hard making simple statements that don't require participant's "cooperation" in learning the terms, construing the meaning to fit the domain - simple semantic analysis doesn't replace thinking about the physics

 

the 1/4 wave fundamental mode of the ear canal is ~ 3 kHz, even for the next few longitudinal modes the response is still mostly due to the averaged SPL at the opening

 

so one part of my argument is that whether one person's ear canal is 21 mm and another's is 23mm long (or even if your R,L ears differ) doesn't change what EQ is applied to a circumaural headphone to give average SPL across the opening at frequencies where the response is reproducible - our brain is already adapted to our personal differences here

 

but the other part is just how much weight to give to the higher frequency range where headphone placement frequency response reproducibilty is poor - what should EQ be doing above ~5 kHz?


Edited by jcx - 7/4/12 at 9:46am
post #160 of 204
Quote:
Originally Posted by jcx View Post

wandering off the rails again I see

 

for circumaural headphones personal ear cavity resonances are not the target of EQ - your brain already compensates for your personal physical differences, whatever your personal internal subjective experience it is the response to the external sound field that is the shared stimulus

 

Smyth SVS Realiser gives impressive "EQ" duplication of the sound of loudspeakers and room by "calibrating" with mics placed at your ear canal opening - its not trying to EQ at your ear drum or compensate for ear canal resonance - your brain already does that

 

only iem have to consider the effects of ear canal variations causing additional frequency response effects since they modify the cavity resonance by their physical intrusion

The problem is that ear canal resonances are changed even by circumaural headphones - so brain can't really compensate the difference. The differential FR of loaded resonances vs. unloaded seems to be quite jagged, with very narrow and high peaks and troughs, so they may mask each other to some extent, as ear's ability to distinguish nearby frequencies reduces at high frequencies, but that doesn't mean there won't be any side effect.

 

There is a criterion on wether a headphone has correct enough acoustic impedance as to not shift resonances too much, called Free-Air Equivalent Coupling criterion. I haven't found the precise description, but seems that the requirement is less than +-6dB SPL variations over 3kHz, caused by acoustic loading. Few headphones fit it, but open planars seem to do so.

post #161 of 204

"your brain already compensates for your personal physical differences, whatever your personal internal subjective experience it is the response to the external sound field that is the shared stimulus"

 

Okay, riddle me this, from your own link

 

Quote:
Even the smallest variances in placements alter the resonance artifacts (peaks and valleys) in the higher frequencies

 

(by up to 18dB, from elsewhere in the article)

 

1. Why would the brain compensate for such differences between headphone listening sessions when it was trained on natural stimuli presented from an outside sound field?

2. Why would you assume that the ear canal resonances (or "resonance artifacts" as written here, since I'm not sure at all that ear canal resonances are what we're actually discussing here) are the same for sound presented in an outside sound field and for the headphones case, when it is pointed out that "smallest variances in placements alter the resonance artifacts (peaks and valleys) in the higher frequencies"?  Obviously resonances that change so much with so little position change of the headphones themselves cannot possibly remain unchanged when the sound is coming from an outside source instead of headphones.

3. If, as per point 2, the resonances at the ear for an outside sound source are drastically different from the resonances presented by the headphones, what value is there to preserving the resonances peaks presented by the headphones?  At best they provide no extra value in positional cueing while degrading treble timbre and smoothness.

 

--

4. The Smyth Realizer test setup measures at the ear opening for both loudspeakers and headphones.  The mic would also be in the reverberant sound field when headphones are put on.  I don't understand reverberant fields very well, but it would seem to me that if you measure impulse response at any point between the drivers and the eardrum and convolved it to match the impulse response from loudspeakers measured at the same point, you would cancel out the reverberant field of the headphones resonance acoustics and replace it with the reverberant field of the ear canal resonance acoustics for the loudspeaker presentation case.  There would be some error caused by the measurement point not being at the ear canal, but surely the effect on those resonance peaks cannot be zero.

post #162 of 204

have to think about your points more, some cross posting going on and I like to edit after I post  as along as I don't mess with the line of reasoning

 

my main motivation was to say we can put some limits on the ~ "we all differ so we can't say anything" position that seems to creep in

 

 

 

I certainly like my ESP/950, Smyth offers a Stax 2020 Lambda system as their reference headphone

 

so I can see/agree there may be additional problems as cup volume, driver radiating surface area go down and has more modes, these could become bigger issues in placement frequency response accuracy

 

it may be that stronger ear canal-cup cavity coupling could make personal EQ more ear canal dimension dependent than I 1st thought as smaller headphones are considered - my experience of "audiophile" circumaural headphone is that they are generally open back and on the larger end of the available range of headphones


Edited by jcx - 7/4/12 at 10:13am
post #163 of 204

I can understand that two very different class of headphones would have to account different interactions with the body differently. Some of these differences my explain why some IEMs sound bass lean even when the FR show them well extended in that region relative to other types of headphones.

 

However, I also feel that if the interactions between headphone and head/ear are properly accounted for by the dummy head, one should get useful measurements over a frequency range (maybe 100Hz to 8kHz):

 

http://www.innerfidelity.com/content/expert-tests-innerfidelitys-headphone-measurement-repeatability-and-reproducibility

http://www.bksv.com/doc/bo0455.pdf

 

Variation from person to person is expected, and things will not be perfect, but this shouldn't not make measurements irrelevant.

 

Moreover, I wouldn't dismiss the measurement information in the 8kHz to 20kHz range, there are ways to smooth out the data and get something meaningful at those frequency ranges:

 

http://www.innerfidelity.com/content/expert-tests-innerfidelitys-headphone-measurement-repeatability-and-rexeproducibility#comment-482802 

 

Someone with experience making measurements would likely set the record straight.

post #164 of 204
Quote:
Originally Posted by jcx View Post

have to think about your points more, some cross posting going on and I like to edit after I post  as along as I don't mess with the line of reasoning

 

my main motivation was to say we can put some limits on the ~ "we all differ so we can't say anything" position that seems to creep in

 

I don't think that "we all differ" neccessitates that we can't equalize headphones with EQ, if that's what you mean.

 

In this thread

www.head-fi.org/t/615417/how-to-equalize-your-headphones-advanced-tutorial-in-progress

I'm trying to detail my method to, in one fell swoop, identify your personal ear canal resonances for each pair of headphones with great accuracy, and equalize headphones with such ease that you can churn out FR comparison graphs like these for fun like one per hour

CKM50-vs-SHE3580.jpg

(red: Philips SHE3580, black: AT CKM50)

 

SHE3580-SHH3580.jpg

(red: SHE3580, black: SHH3580)

 

SHE3580-vs-SHE8005.jpg

(red: SHE3580, black: SHE8005)

 

SHE3580-vs-etys.jpg

(red: SHE3580, black: ER-4P)

 

SHE3580-vs-HJE120.jpg

(red: SHE3580, black: Panasonic HJE120)

 

HTF295-upright.jpg

(Panasonic HTF295 (only circumaural headphones I measured...) upside down FR chart, higher is quieter (only for this chart))

 

UE100-vs-SHE3580.jpg

(red: SHE3580, black: Ultimate Ears UE100)

 

...and tell people that you can make all your headphones sound 95% the same, to your ears, by dialing in these curves, and watch everyone stare in disbelief.

 

tongue.gif

 

(sorry for the long long list of graphs, I've never posted them all together before, thought this thread might be a semi-good excuse to do so.  I can point people back to this post if they want to see my gallery... tongue.gif)

post #165 of 204

and Unique Melody will go out of business.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › How far can EQ really go towards truly equalizing headphones?