Damping Mechanical Energy Distortion of STAX and other phones with SORBOTHANE and other materials.

eurobeat said:

  So you're saying the frequencies are being altered/lowered?  Wouldn't that affect the entire sound then? Wouldn't it be a lot diffrent/odd?

Click to expand...

I am no expert but here is my intuition:
 
When "damping" anything - be it a car or a pair of earphones, you are damping/absorbing vibrations / resonances that are unwanted. Hence in a car "lowering the noise floor" means that you are eliminating broadband noise that normally intrudes in the "listening space" via wind and road noise transferring through the metal / plastic / glass surfaces. You can defeat that noise floor by going even louder, but to some extent I think that the speaker / "room" interaction might lead to excited or dampened frequencies (peaks and dips in response) that you didn't want. Of course if the problems are with the speakers themselves, you still end up with problems, but that is what EQ is for too. A "perfect" speaker in a "perfect room" doesn't, to my knowledge, have a perfectly "flat" response, but would be close to it (maybe a slight downward tilt). I consider the Revels to be a benchmark in that regard.
 
Anyway, you can get above the noise floor, but you are not eliminating distortions that are present given the way the speaker and the room are interacting (e.g. boosting or cutting frequencies). Particularly with the HE-400s, "ringing" in the midrange seems to be a critical issue. In my opinion, that could relate to the plastics inside or the driver itself - is excess energy present because the materials naturally resonate near critical frequencies? Is it because of the "waffle" frame of the driver? We can't change that. But, it it would be ideal to have the ability to push the resonant frequencies of the "room" itself far away from human "audibility" (probably either below 20 hz or above 20 khz). If the "room" of the HE-400 driver can be pushed 40 dB "below" whatever it is currently at any frequency above 50 hz, perhaps it will "get out of the way" and allow the drivers to make their music with less distortion. This isn't changing the speakers, it is changing the room - we just don't know what the impact on the overall response then becomes. Ideally, we kill the ringing issues and any reflections that might contribute to frequency response peaks that are audible and undesirable contributing to fatigue. In short, I believe that what this does is create a "dead" room for the drivers that should, I would think, eliminate some of their existing distortion as measured. Or maybe lots - who knows?
 
In a car some of the noise occurs at low frequencies, others at high (you can sort of tell where your problem frequencies are particularly in the bass where there never seems to be good midbass response but often jacked up at 40-50 hz). Next time you are driving turn off the radio and just listen - that "white noise" is competing with your speakers to be heard - it still exists even when you drown it out. Hitting expansion joints etc "thumps" and usually has a lower frequency. The engine makes noises and so on. It all impacts the response of the car overall no matter the quality of the system. The speakers are affixed to panels that are treated - but you don't treat the speakers themselves so they are free to vibrate / resonate / create noise into the empty space - they just no longer do so with whatever additional resonances or reflections the car might add. What has changed is not the speakers output or response, but the characteristics of the "room" insofar as they relate to the overall response. There are always lots of unwanted reflections + less than ideal placement but killing broadband noise is doing wonders for headroom and allowing you to play speakers at levels where they probably "behave" better.
 
In short, outside of an anechoic chamber, the world is a noisy place and any speaker will behave differently whether in a yard, in a room, a closet, etc. Optimizing response goes beyond having a good driver - it includes having a good room. Hence, your bass response will be way out of whack if you have a massive subwoofer in a tiny room, or a tiny subwoofer in a large room. I just have no idea how this is actually working in a headphone space, since both the room and the ear introduce physical variations.
 
The HE-400s are, stock, still one of the better headphones for the money particularly if you want to get some Planars (my new choice would be Oppos). Slight mods addressing damping have improved the sound so far - and I do think there is still performance on the table. Certainly planar speakers are placed differently than dynamics in a room, because dynamics either have a "dead" backwave trapped in an enclosure ideally of non-parallel surfaces or are vented to boost bass response (notice vents NEVER get tuned to midrange / treble frequencies). When I get my sorbathane, I will try to do some listening before and after (including with grill on and grill off to see if that is the real "difference"). I will try to "measure the backwave" / frequency response with totally non-professional equipment. Just basic freq response and SPL if I can get it. If sound overall is improved, I'll let everyone know. This is a pretty neat little experiment - and I would tend to think there is some real potential benefit to it. Just know that "overdamping" can be bad - "dead" or "lifeless" sound, probably with little "air" or "excitement". This is partly a balancing act, and my personal preferences and biases will apply. But I would fully expect this stuff to help with ringing somewhat, and lower THD even, but without getting a "properly damped" measurement, we won't know. Maybe we send a modded pair to Tyll!

Just taking a trip over to Innerfidelity to re-examine the measurements, I would add that the HE-400s definitely have significant ringing issues.
 
Also, THD seems to be high from 200hz-600hz, 1 to 1.2 khz, and about 2.5 khz. Possibly an issue starting around 7 khz - 10 khz, but the measurements don't reach that far. That is with pleather pads on, presumably, totally unmodded. So I would assume these are the offending frequencies, and of course the overall measurements show lots of rippling probably masked by smoothing the data.
 
No matter how you slice it, I still think these things need a fair amount of creative EQ'ing to balance out the sound a bit. That will be my last step.

mrmateohead said:

  I am no expert but here is my intuition:
 
When "damping" anything - be it a car or a pair of earphones, you are damping/absorbing vibrations / resonances that are unwanted. Hence in a car "lowering the noise floor" means that you are eliminating broadband noise that normally intrudes in the "listening space" via wind and road noise transferring through the metal / plastic / glass surfaces. You can defeat that noise floor by going even louder, but to some extent I think that the speaker / "room" interaction might lead to excited or dampened frequencies (peaks and dips in response) that you didn't want. Of course if the problems are with the speakers themselves, you still end up with problems, but that is what EQ is for too.

A "perfect" speaker in a "perfect room" doesn't, to my knowledge, have a perfectly "flat" response, but would be close to it (maybe a slight downward tilt). I consider the Revels to be a benchmark in that regard.
 
Anyway, you can get above the noise floor, but you are not eliminating distortions that are present given the way the speaker and the room are interacting (e.g. boosting or cutting frequencies). Particularly with the HE-400s, "ringing" in the midrange seems to be a critical issue. In my opinion, that could relate to the plastics inside or the driver itself - is excess energy present because the materials naturally resonate near critical frequencies? Is it because of the "waffle" frame of the driver? We can't change that. But, it it would be ideal to have the ability to push the resonant frequencies of the "room" itself far away from human "audibility" (probably either below 20 hz or above 20 khz). If the "room" of the HE-400 driver can be pushed 40 dB "below" whatever it is currently at any frequency above 50 hz, perhaps it will "get out of the way" and allow the drivers to make their music with less distortion. This isn't changing the speakers, it is changing the room - we just don't know what the impact on the overall response then becomes. Ideally, we kill the ringing issues and any reflections that might contribute to frequency response peaks that are audible and undesirable contributing to fatigue. In short, I believe that what this does is create a "dead" room for the drivers that should, I would think, eliminate some of their existing distortion as measured. Or maybe lots - who knows?
 
In a car some of the noise occurs at low frequencies, others at high (you can sort of tell where your problem frequencies are particularly in the bass where there never seems to be good midbass response but often jacked up at 40-50 hz). Next time you are driving turn off the radio and just listen - that "white noise" is competing with your speakers to be heard - it still exists even when you drown it out. Hitting expansion joints etc "thumps" and usually has a lower frequency. The engine makes noises and so on. It all impacts the response of the car overall no matter the quality of the system. The speakers are affixed to panels that are treated - but you don't treat the speakers themselves so they are free to vibrate / resonate / create noise into the empty space - they just no longer do so with whatever additional resonances or reflections the car might add. What has changed is not the speakers output or response, but the characteristics of the "room" insofar as they relate to the overall response. There are always lots of unwanted reflections + less than ideal placement but killing broadband noise is doing wonders for headroom and allowing you to play speakers at levels where they probably "behave" better.
 
In short, outside of an anechoic chamber, the world is a noisy place and any speaker will behave differently whether in a yard, in a room, a closet, etc. Optimizing response goes beyond having a good driver - it includes having a good room. Hence, your bass response will be way out of whack if you have a massive subwoofer in a tiny room, or a tiny subwoofer in a large room. I just have no idea how this is actually working in a headphone space, since both the room and the ear introduce physical variations.
 
The HE-400s are, stock, still one of the better headphones for the money particularly if you want to get some Planars (my new choice would be Oppos). Slight mods addressing damping have improved the sound so far - and I do think there is still performance on the table. Certainly planar speakers are placed differently than dynamics in a room, because dynamics either have a "dead" backwave trapped in an enclosure ideally of non-parallel surfaces or are vented to boost bass response (notice vents NEVER get tuned to midrange / treble frequencies). When I get my sorbathane, I will try to do some listening before and after (including with grill on and grill off to see if that is the real "difference"). I will try to "measure the backwave" / frequency response with totally non-professional equipment. Just basic freq response and SPL if I can get it. If sound overall is improved, I'll let everyone know. This is a pretty neat little experiment - and I would tend to think there is some real potential benefit to it. Just know that "overdamping" can be bad - "dead" or "lifeless" sound, probably with little "air" or "excitement". This is partly a balancing act, and my personal preferences and biases will apply. But I would fully expect this stuff to help with ringing somewhat, and lower THD even, but without getting a "properly damped" measurement, we won't know. Maybe we send a modded pair to Tyll!

Click to expand...

Noise is generally defined as "unwanted sound."  It can apply to both sound from internal and external sources, i.e. background sounds in a listening room are noise, electrical disturbances which cause sound are noise in a stereo system, and I guess you can define all manner of distortion in stereo systems as noise but at some point you have to bear in.mind the different origins of these different kinds of noise because the methods of defeating them are going to be quite different.
 
 Damping is the process of reducing the amplitude of oscillations in an " oscillatory system. " What we seem to be doing with sorbothane is reducing the oscillations of portions of the earcups of headphones which seem to be coming from the drivers in accordance with the Newtonian principle that for every action there is an  equal and opposite  reaction. I.e. sound is going by direct mechanical coupling from the drivers to the surrounding structures. 
 
Ideally there should be no contribution to the sound you hear in a headphone from oscillations of the structures surrounding the drivers anymore than you want to hear sound coming from the surface  of your loudspeakers and good speakers go to considerable effort to reduce this. To reduce these sounds, speakers are often heavy, rigid and coupled to even larger structure by things like floor spikes.
 
With headphones, there appears to have been a general assumption that such direct mechanical oscillations are not a problem. However,  the sorbothane effects show there is a quite noticeable problem.  Why this has not been noticed before, I am not sure. Possibly just that  there are few thing you can do with a headphone compared to a loudspeaker to reduce the problem. You can't just keep adding mass and you are not going to add spikes. Even increasing rigidty may not do.much because everything resonates no.matter how stiff. It may very well be that adding something like sorbothane is just about all you can do.
 
The term "resonance" is also used with different meanings in audio. Often we are talking about the resonances within chambers, such as rooms or the open spaces of headphone earcups. However here I am talking about  what I call "mechanical resonance" to refer to vibrations within the physical structures of the earcups, not the spaces around them.
 
Ayway, this is my understanding of what is happening here. Obviously some serious research is needed to confirm this or provide an alternate explanation. 
 
As regards measurement, what I think is really needed is measurements of the vibrations of the earcup structures themselves. This would need some kind of sensor on the surface of the earcup, not a microphone in the air.

Edstrelow - great additional comments, and a good idea to distinguish what we mean by noise, damping, and resonance.
 
Polycarbonate plastics have a density that approaches that of metal, so I would assume natural resonances would be similar. 500 hz? 600 hz? I don't know. Either way there is definitely mechanical coupling taking place, and these are powerful drivers. Amazing to think these could behave like big speakers in a big room (where you can hear the windows rattling! Just the scale is so much smaller one can't really perceive the issues by "ear".

Thanks for the info...  Still no reply from them, and I decided that I'm probably not buying the Yamahas MT220's anymore.


Got the Pioneeer HRM-7's instead.  They might need sdome rework, but dont see any comments like that.

the he 400 shell earcup is a bell.... full on unwanted resonance partly annihilated by the sorbothane.... When i read this thread some months ago it was intuitively evident.. i dont regret it a bit.... the most important upgrade for minimal cost..... thanks to Edstrelow.....

How easily does the adhesive stick?  Does it come off fairly easily if we put too much on/want to remove it?

the self adhesive sorbothane 
stick very well and is easy to remove without almost any trace........ 

Ordered some 1/8 40D self adhesive sorb.
 
Will most likely end up messing around with some sort of hybrid fuzzor/sorbzor(?) modifications + some on the inner outside housing which has been shown previously on the HE-400's.

jackharm said:

  Ordered some 1/8 40D self adhesive sorb.
 
Will most likely end up messing around with some sort of hybrid fuzzor/sorbzor(?) modifications + some on the inner outside housing which has been shown previously on the HE-400's.

Click to expand...

remember to report here... 40 duro its interesting....

As I drove down the highway today, I noticed how bad the reflection from the windshield / dashboard was. I could literally see the whole dashboard as if it was grafted onto the road in front of me like a hologram. Then I thought about how I wouldn't be seeing it if I were wearing a pair of polarized sunglasses. That sort of "visual glare" must be similar to audible "glare" - that sense that the image of the sound you are trying to see is both present but accompanied by reflections that are mostly unwanted. Probably common sense to the lot of you, but a needed "insight" for me before I went about the business of modding my HE-400s (again).
 
Particularly when the recording itself makes ample use of reverb / chamber-like "echo" and space, glaring resonances and reflections only make the listening experience a bit glaring / difficult. One of the first songs I queued up was Joss Stone "Love Me", a song I've been listening to a lot lately. I am not 100% positive, but immediately it seemed the treble peak was subdued quite a bit (even missing?) and a lot of midrange glare was absent - her voice and that of her backup singers seemed to be presented nearly as clearly as it is on my PSB M4Us. I am going to try out more recordings that usually assault my ears and report back.
 
I am doing some listening right now but will post pics and some pretty unscientific data later.  

interesting remarks indeed ...

Sorbathane Mod:
 
You can lookup the "grill mod" to see how the stock grills are removed - its surprisingly easy but I did use a razor as I have useless fingernails. The rings seem to not be brittle, which was a relief. That said, they are tiny. Be careful. Once two tabs are popped, you can safely pull out the ring and continue.
 
Here you are looking at 1/10" thick 30 Duro Sorbathane crudely cut by hand with a razor. The stuff is gummy and I was using a 6x6 sheet. You have about 1/4" depth in the HE-400 cups so I cut to about 1/4" or so and then lined the inside with varying lengths of the stuff. You can pinch this stuff hard and it deforms readily, but goes right back to shape:

In the process of doing this I did notice that there is no "empty cavity" - there is either some sort of liner/membrane, or plastic obstruction that seems to separate the main chamber from the back chamber. Even so, I decided to cut additional strips of 1/8" width and added an "L" shaped treatment to the driver housing itself - on the side of the driver that would NOT be the back of my ear, but the front. I guess this might be like putting a bit of carpeting in front of your loudspeakers as opposed to doing it behind them "absorbing first reflections". I am thinking about surrounding the whole driver with these thin strips, but am now curious as to why doing the back, as opposed to the front, seems to have made a desired impact - reducing "glare" and "resonance" from the mid-range response in particular and "cleaning up" the treble response insofar as I can claim to actually hear a difference (and I am quite sure I can). I am quite stunned. Yes, the treble is still dialed up a little too hot, but I swear it is not as pronounced / distracting as it was on some tracks before the mod (I note that I am NOT EQ'ing presently, but will probably EQ after playing around). In short, the "polarized sunglasses" are on, and it's a nicer ride. I think one could easily use thicker sorbathane as well - 2/10" would seem to fit fine with a little cramming.
 
Here is an attempt to measure any changes. This is a Nexus 7 + RTA app with NO calibrated microphone measurements, so they cannot be taken in any way as comparable to anything published out there. The mic clearly rolls off at around 150 hz and misses the 1 khz "hump" documented elsewhere, so I am sure its response curve might be odd (but appropriate for OK recording of human voices / phone call duty). But I am also certain that despite this, it is better than my ear at actually quantifying possible differences in amplitude. Given that the MIC reports high frequencies, it could be full-range and the lack of bass could be from a lack of seal - I am not sure which. Maybe I can record a "sine sweep" with it or something and figure out where its cutoffs actually are. Or maybe I'll actually get a measurement mic sometime.
 
HE-400 (Ear Side), Pink Noise, Raw response

HE-400 (Grill Side), Pink Noise, Raw Response

HE-400 Sorbathane Treated (Ear Side), Pink Noise, Raw Response

HE-400 Sorbathane Treated (Grill Side), Pink Noise, Raw Response

 
I should note that I did not discern, immediately, any real difference from having the stock grills on or off. I expected a sense of "air" or otherwise some enhanced clarity but in the few moments I bothered to check, found that inconclusive. I also alternated holding up the grills and not, and again, NBD. So, for all intensive purposes, they are probably quite efficient / transparent already. Having spent time with the PSB M4U1's lately, I also consider those to sound more "open" than the HE-400s anyhow with a more resolving and authentic treble (heresy I know - the PSBs are closed back!).
 
The Horizontal Bars you can see represent +- 12 dB - a SIGNIFICANT difference in amplitude. That makes interpreting this raw data a lot harder. Unfortunately, I don't have a program that scales more closely, so I can get a better window into response. Roughly half that is 6 dB, and half of that is a range (+- 3 dB) where I think we approach "flat" response. The yellow line is an averaged response (32 samples a second or whatever) that smooths things out a lot. If I had to guess, I'd say a change is, on the grill side, some smoothing of response between 2-3 kHz, and again from 3 kHz - 6 kHz. There seems to be some tightening of response in the 8-20 kHz range, and a smoother roll-off to 20 kHz. Mind you, I am not referring to the actual shape of the raw data line (which looks little like real measurements done elsewhere with good equipment) - I am focused on whether the rippling of the averaged line looks significantly different (for example, an absence of ripples that fit within approximately +- 3 dB that did not before). That would seem to indicate to me that the Nexus 7 is "hearing" something different.
 
Measuring Pink Noise on the ear side, I notice in both cases a response that steps down about 12 dB by 3 kHz, and 18 dB by 6 kHz. It then rises from 6 kHz up to a peak around 10 kHz, if my microphone is to be believed. The most significant difference appears to be that response tightens up after 6 kHz, and the treble roll-off after the peak is more severe. But it is still there.
 
All in all the measured response looks pretty darn similar - with closer inspecting revealing what look to me to be differences in slope and some clear changes in overall responses (pretend you can draw straight lines through the yellow line particularly in the treble regions).
 
It would be better to see waterfall plots to look for signs that ringing has been reduced at different frequencies. Maybe someone else can do this with a modded pair.
 
But i can say that listening comfort is up - particularly at high volumes - and that is a very good sign. I am quite satisfied with the change. As a next step, I am tempted to see if "framing" the driver on the ear side might provide a little more damping, or just ruin the sound. Or, would more damping in the rear be a good idea? I could follow the plastic around the outer "circle" to further drape the back wave. I am not sure what I might try first, particularly as the overall change has already been pretty good. I am open to all opinions!

And just to wrap up, here is an idea for really "overdoing" the back of the driver - by sticking more sorb on all the orange parts:
 

I could either double up on the cup itself, or focus on the white plastic "driver housing". Might make for a very "dead" backwave.

thank you very much....very interesting experiment ... i will read all your posts and that will inspire my future mods also