Mechanical Transmission of Noise via Headbands

Discussion in 'Audio Science' started by Lyander, May 3, 2020.

?

How badly've shelter-at-home orders been affecting you?

  1. Hardly any change from my usual routine, really. If anything, this is pretty good.

    11.5%
  2. Mild inconvenience, but I'm working around it with minimal discomfort.

    65.4%
  3. This sucks, I can't wait until the pandemic dies down so I can get back to normal life.

    11.5%
  4. I'm literally making threads discussing some of the silliest stuff re: headphone design ever

    3.8%
  5. What the hell you're still locked up? I'm free to go ou whenever I please!

    7.7%
  1. Lyander

    Lyander Official SBAF Equitable Empathizer

    Pyrate Contributor
    Joined:
    Feb 25, 2017
    Likes Received:
    10,961
    Trophy Points:
    113
    Location:
    Philippines, The
    Starting a new thread because I'm actually really curious about the Sony R10 and want to keep that one on-topic, talk of CNCed steps in the cups and all :p

    LOL, didn't think about that! I'd always somehow assumed that the dummy heads would have something in them on account of how they're supposed to be representative of actual human heads; last I checked most people not deeply involved in politics or showbiz generally have stuff mass-loading the skull, haha.

    I dunno about headband resonances not transferring any bass, though. How easily do human heads transmit sound? Guessing that LFE might be more likely to pass through the hodgepodge of bone, flesh, and viscera that we are. Given the dangerous nature into research on infrasound, specifically its effects on the human body because of its potential as a WOMD, it's understandably hard to find literature on the topic, the closest I can get with a brief search being:

    https://phys.org/news/2016-10-scientists-effects-infrasonic-vibrations-humans.html

    .... which states that the human body produces infrasound via natural physiological processes e.g. cardiovascular circulation and brain activity. This is getting down to ridiculously minute detail and I honestly doubt people are gonna take this much care to look into something for recreational listening, but I've always been curious about the effect these body sounds have on perceived listening. Chances are remarkably high that it's just more stuff our brains shove out of our perception though; maybe it's no mistake that our bodies produce sounds that are below what is commonly accepted as the threshold of human hearing (20Hz).

    I think the problem here is that with few exceptions the people in this hobby don't have access to perfect simulacra of human heads (how close does ballistics gel get? MIGHT know where I can score some), so a lot of this is going to be difficult to determine unless you shove microphones through someone's sinus through their Eustachian tubes and just have mic diaphragms take the place of their eardrums.

    Good luck getting THAT past an ethics board.

    Just thinking about it, shouldn't headband coupling with skull (assuming proper fit) dampen most sounds, and any cross-channel contamination would then have more to do with sound transmitted via air surrounding the head? I shudder to think at the volumes one must listen to to get that much of an effect. Spent a bit of time looking for research but came up empty, maybe because no one's bothered with something this close to quartz purity nervosa.
     
  2. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    This. I'll post some of my measurements in a bit, but basically all the crosstalk I saw seemed to be airborne sound travelling from the left driver back into the right earcup (while also reflecting off off room walls, etc.). With closed headphones we may see more headband-conducted sound relative to the sound leaking from the cups, but I doubt it's that bad. I'd expect accelerometer testing to be more interesting (I still don't have one, I should change that).
     
  3. Lyander

    Lyander Official SBAF Equitable Empathizer

    Pyrate Contributor
    Joined:
    Feb 25, 2017
    Likes Received:
    10,961
    Trophy Points:
    113
    Location:
    Philippines, The
    I don't have any closed headphones now either, gave the Mezes away to sister (who lives with me so no biggie can probably steal back for a bit), but currently unable to assess since I doubt an EARS rig is up to the task.

    The stuff about accelerometers went over my head a bit, all I know is that they were a big deal in smartphones about a decade ago, but looking forward to seeing what you come up with!
     
  4. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    EARS rig should be a good indication. Might be better than a hollow dummy head. All I do is stick a WM-61A into my ear canal.

    Here are the results I get for HD800 and HD600. I averaged both channels and took two measurements of the crosstalk. One as is, sitting normally in my room (green) and another one with a large pillow pressed against the earcup (red) producing the sound to try to isolate more the effects of the headband and of the sound that is reflected around the room. All measurements uncompensated with a WM-61A mic in a silicone tip in my ear. Both HD800 and HD600 are modded, but the HD600 could really use new pads. My glasses impact the seal and limit bass extension here, but that's not too relevant for crosstalk.
    HD600 crosstalk.jpg HD800 crosstalk.jpg

    A couple observations:
    1. Crosstalk as is is extremely low with these open headphones, better than phono cartridges for example. I'd say the results are similar to the ones you mentioned in your other thread here.
    2. There's going to be a larger uncertainty in the lower bass region, but both of these headphones seem to fare better than the closed headphones Keith Howard measured in the upper bass "headband resonance" region, if that is really what it is.
    3. It's interesting how the HD800 is so affected by putting a pillow on the earcup in the midrange, while the HD600 is not. Either way, it seems the pillow was not very effective at blocking sound leakage, this is despite it being more than 10cm thick and covering the entire capsule. I used a similar pillow to block the Voxativ driver's sound when measuring the PAP Trio 15 design here, where it seemed very effective.
    4. Headband resonances might be slightly lower with the HD800. Marginally better crosstalk between 30Hz and 300Hz I'd say. Considering how much effort Sennheiser put into the HD800 headband I expected to see more of a difference, though.
    5. We are hitting the noise floor past the treble range (smooth slope from 10kHz on or so). It's not entirely consistent from run to run for various reasons, but try to ignore the area, where the line is smooth. I ran the measurements at around 95dB to prevent it. That is really loud, I wore earplugs in the ear not covered by the microphone. Sidenote: Somehow the foam earplugs seem to isolate a tiny bit less than a mic inside a silicone plug.
    6. I underestimated how much crosstalk there is in the bass region and overestimated how much there is in the upper midrange. I think it's interesting how there is this distinct pattern for the measurements where there are essentially two peaks, one lower in frequency presumably from vibrations and mechanical crosstalk and a second one from the sound emitted by the driver leaking back into the other cup. The excess group delay plots are much more even than I expected, though. I expected the upper midrange to be more delayed compared to the upper bass and lower midrange, but this doesn't seem to be the case.
    7. The HD800 with its larger driver leaks more sound at lower frequencies than the HD600. Nothing new here.
     
    Last edited: May 3, 2020
  5. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    @Lyander: I still don't understand Ron Howard's L to R crosstalk measurements...
     
  6. k3oxkjo

    k3oxkjo Friend

    Pyrate
    Joined:
    Sep 30, 2015
    Likes Received:
    64
    Trophy Points:
    18
    Location:
    N. J.
    Home Page:
    Just for fun, I took a 5 foot piece of shielded, twisted pair like you might find on a cheap headphone and put a signal generator on one end between shield and red wire, then measured with my oscilloscope between shield
    and first the red wire, then the unconnected white wire on the other end. The results:

    Red wire=40v p-p
    White wire=4v p-p

    That's 20 db difference.

    I then redid the test with a 100 ohm termination rather than open circuit at the measurement end:

    Red wire=20v p-p
    White wire=.5v p-p

    With termination, 26db difference.

    I thought there would be more crosstalk than you would think, but actually measured more crosstalk than I would have expected.

    Best, Kevin
     
  7. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Can do redo this experiment with properly shielded headphone cable? I don’t get anything remotely close to this in crosstalk measurements on my dummy head where I only play a signal on one channel and record on the other.
     
  8. Lyander

    Lyander Official SBAF Equitable Empathizer

    Pyrate Contributor
    Joined:
    Feb 25, 2017
    Likes Received:
    10,961
    Trophy Points:
    113
    Location:
    Philippines, The
    I'm tempted to say that while he's bringing attention to something that may be worth addressing, his manner of going about it is... odd?

    Dude's got decades on me in measurements and audio geekery so maybe not, I'm just happy to see others give it a shot :)

    @Serious EARS might work but I'm not kidding when I say that my place is NEVER quiet. If it's not my dad's aviary just outside my window with its constant chirping and twittering it's the dogs or cats or cars grumbling by.

    At least there are fewer people out and the cantina/karinderia next door's closed.
     
  9. k3oxkjo

    k3oxkjo Friend

    Pyrate
    Joined:
    Sep 30, 2015
    Likes Received:
    64
    Trophy Points:
    18
    Location:
    N. J.
    Home Page:
    I have a 10 foot piece of twin shielded cable, that is each side has its own shield. With the shields connected at the source end and the signal generator fed to one of the sides the measurements I got were:

    20 v pk-pk on fed side terminated into 100 ohms
    .2 v pk-pk on non fed side terminated into 100 ohms

    This is 40 db, closer to what I expected. I suspect better results with twin twisted pair shielded cable with shields connected only at the source but don't know if I have any of that to test. I will also try to find a 600/600 ohm audio transformer (in my garage somewhere) to attach to the signal generator output to remove any possible loop effect of the signal generator AC ground.

    Best, Kevin
     
  10. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    300-ohm dummy load. AverLAB -> Magni Heresy (so there could be some crosstalk that the amp is adding). Output just below 0dBu which is ~ 0.775Vrms which is typical for headphones.

    Left channel driven. Right is -57db lower.
    upload_2020-5-3_21-1-4.png


    Left channel driven, but ground on L is lifted. Right is -85db lower.
    upload_2020-5-3_21-2-45.png

    Here is the photo of the cable with the R channel lifted. It's one of those detachable ones with 1/4" jack to x2 mono 1/8" deals. Nothing special.
    IMG_20200503_210330.jpg
     
  11. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    IMG_20200503_214213 (1).jpg

    Signal played back on L channel only.

    ORANGE = recording from L microphone on dummy head
    MAGENTA = several recordings from R microphone on dummy head
    GREEN = several from R microphone on dummy head - R connector pulled from driver (note that this have have slightly shifted the headphone)
    upload_2020-5-3_21-45-4.png

    @Lyander: I am wondering if Ron Howard had something else that was introducing crosstalk where his R capsule recording was tracking the L almost exactly (except for those swiggles here and there). The fact that this tracking behavior disappeared when he lifted the R ground only serves to support this possibility further. Note that my results are fairly consistent with @Serious's except I am getting more pickup in the higher frequencies. The latter could be the result of the dummy head on a table next to a wall.

    Note that there are a lot of dummies in the measurement game, especially when it comes to publications. I remember some dude who posted distortion plots of a headphone and announced "looky, no distortion!" when he plotted the distortion % without log scale.
     
    Last edited: May 3, 2020
  12. k3oxkjo

    k3oxkjo Friend

    Pyrate
    Joined:
    Sep 30, 2015
    Likes Received:
    64
    Trophy Points:
    18
    Location:
    N. J.
    Home Page:
    I believe the gentleman's name is Keith Howard. He has been around a long time and he does have a good reputation.

    But that doesn't mean he isn't wrong in this case.

    Does your test setup have common ground between left and right generator output?

    My measurements were quick and dirty as I don't currently have a permanent test setup errr… setup. It may be fun to set something up and revisit my youth! I have a Sound Technology test set in my garage, I wonder if it still works?

    Best, Kevin
     
  13. Lyander

    Lyander Official SBAF Equitable Empathizer

    Pyrate Contributor
    Joined:
    Feb 25, 2017
    Likes Received:
    10,961
    Trophy Points:
    113
    Location:
    Philippines, The
    This is his testing rig, I believe, GRAS fake pinnae on a stacked MDF "head" with an ear canal simulator in a box:

    https://www.hifinews.com/content/better-headphone-testing

    Once again feel compelled to say I know jack doodoo about materials, but from what I do understand MDF is pretty damn efficient at damping sounds in turntable applications. Whether it's nearly as effective with the presumably more disruptive vibrations borne of running sine sweeps with headphones directly coupled to that stack of MDF though, I have no idea.

    There are ports and cavities there that I presume aren't comparable to how sinuses in the head perform.

    P.S.
    @k3oxkjo I mentioned his name earlier over in the R10 thread before this split off so I presume the L. Ron-Hubbarding is deliberate :p
     
    Last edited: May 3, 2020
  14. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    The Magni 3 Heresy is used as the headphone amplifier for the dummy head setup. It does have common ground - obviously in this case. The gain is set to x1 for lowest distortion.

    I suspect misconfiguration of the GRAS system. I've used one before. It's a lot more complex than it needs to be for the measurements that we typically see.
     
  15. Lyander

    Lyander Official SBAF Equitable Empathizer

    Pyrate Contributor
    Joined:
    Feb 25, 2017
    Likes Received:
    10,961
    Trophy Points:
    113
    Location:
    Philippines, The
    I looked it up, Aurorus Borealis has a nickel-titanium wire thingy for headband. Nitinol? Might make a poor candidate since it's anecdotally effective at damping vibrations, though it definitely has a resonant frequency of its own.
     
  16. k3oxkjo

    k3oxkjo Friend

    Pyrate
    Joined:
    Sep 30, 2015
    Likes Received:
    64
    Trophy Points:
    18
    Location:
    N. J.
    Home Page:
    I don't know what I did with my Sound Tech test set, but I did find my Potomac Industry test gear. I suspect I had ground loops or something going on before.

    But with relatively proper OG test gear:

    The 5 foot piece of single shielded, twisted pair like you might find on a cheap headphone measured 32 db crosstalk rather than 26 db.

    The 10 foot piece of twin shielded cable, (that is each side has its own shield) measured 88 db crosstalk rather than about 40 db.

    Best, Kevin
     
  17. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    The best way is to use accelerometers or laser inferometers with the headphone on a subject's head. A HPF can be applied to filter out subject's movement. The L to R crosstalk shit or hanging a headphone band on or off a dummy head to infer vibrations seems asinine at worst or convoluted at best.

    Electrical crosstalk or ground contamination is not going to play a significant role with open headphones where the sound leakage will be much greater.
     
  18. Lyander

    Lyander Official SBAF Equitable Empathizer

    Pyrate Contributor
    Joined:
    Feb 25, 2017
    Likes Received:
    10,961
    Trophy Points:
    113
    Location:
    Philippines, The
    I'm reviving this thread to submit a good case for headband design possibly being a cause for concern: the Meze 99 Classic, haha. Gave them away to one of the sisters but they returned em a few days back cuz of disuse; I've been running them out of my mobile and laptop pending the return of other components in my rig.

    Holy crap, just running my fingers lightly along the headband and cables makes for easily audible noise in the earcups, and tapping the one earcup with a fingernail seems to actually have an effect on the other channel. That said, this is just one headphone among many; it's not that headphones often have poor design, as proposed by K. Howard, it just seems that the 99 Classic has... a peculiar flaw, haha.

    Attached is a recording I made using MiniDSP EARS. The LEFT channel is the only one currently functioning because I kinda killed the right microphone while messing around with the rig (it's USD15 plus shipping for a matched set of microphones and cables from MiniDSP, if anyone's curious— I look forward to getting that replaced eventually :p), but the fact that right channel is dead kinda serves to validate things here, much as it's entirely possible that noise generated is just passing through the surrounding air or even through the test rig. Yes this is hardly definitive proof, but again this could just be a problem particular to these headphones, haha.

    0:03 is me tapping the middle of the headband with my fingertip
    0:09 is me lightly ticking the same spot on the headband with my fingernail
    0:15 is me tapping the RIGHT (distal) cup with my fingernail
    0:21 is me tapping the LEFT cup with my fingernail, same amount of force as above
    0:26 is me softly rubbing the LEFT cup surface with a fingertip
    0:32 is me tapping the LEFT cable about 2in from the termination with the same amount of force as on cups

    https://drive.google.com/open?id=1_TQKIpYi89Z340KKCyGQ1DI_LeRK_XfS
     
    Last edited: May 27, 2020
  19. Jinxy245

    Jinxy245 Vegan Puss

    Pyrate
    Joined:
    Nov 17, 2016
    Likes Received:
    5,733
    Trophy Points:
    113
    Location:
    Montgomery, New York
    The Meze are known for that, I remember that vividly. The headband transmitted all sorts of noise, IIRC even in the wind. Definitely an indoor stationary headphone IMO.
     
  20. Lyander

    Lyander Official SBAF Equitable Empathizer

    Pyrate Contributor
    Joined:
    Feb 25, 2017
    Likes Received:
    10,961
    Trophy Points:
    113
    Location:
    Philippines, The
    Not gonna be able to test "outside" use any time soon I think but that about sums it up, haha. I'm tempted to eff up the headband to try and reduce the amount of noise it picks up but I burnt through my stock of damping material so it'll have to wait. Thanks for verification though :))
     

Share This Page