Head-Fi.org › Forums › Equipment Forums › Headphones (full-size) › Headphone CSD waterfall plots
New Posts  All Forums:Forum Nav:

Headphone CSD waterfall plots - Page 52

post #766 of 937

You can't have your cake and eat it too Nevod.  You mention ear canal resonances which can not possibly be compensated for by headphone FR and must be EQed out with a parametric EQ tailored to the person...

post #767 of 937
AFAIK, dummy heads can model outer ear at best and small in ear mics on a real person can also only get the spl at the entrance of the ear canal, but at least the ear canal is not necesserily blocked.

I don't see where the problem is with this though as the goal here is not to equalize out the personnalized ear canal response.
post #768 of 937
Originally Posted by arnaud View Post

AFAIK, dummy heads can model outer ear at best and small in ear mics on a real person can also only get the spl at the entrance of the ear canal, but at least the ear canal is not necesserily blocked.
I don't see where the problem is with this though as the goal here is not to equalize out the personnalized ear canal response.

Seems like some extension tubes are used for measurement at eardrum,mics are coupled via these tubes.


Seems like some people would though want to equalise even that, and it would be problematic, as the FRs gathered from all the measurements do not represent the shift of the ear canal responce due to added impedance. Some headphones though have more or less flat PDR and hence they can be safely equalised by blocked-canal measurements.


However, I'm personally not so interested in EQing as it has obvious drawbacks compared to fixing the real cause. I have thought of using Helmholtz resonators at the earside chamber to reduce radiation impedance at the important frequencies (PDR seems to manifest at about 5-14 kHz, some of these could be tamed.) I intended to use resonators made out of matchboxes, but haven't yet thought them out and calculated. Maybe I'd have to construct them some other way.

post #769 of 937
Originally Posted by Joe Bloggs View Post
Originally Posted by purrin View Post

Beyer T5P




No comment.


This looks kinda like what might happen if someone tried to equalize the phones to cancel out one particular person's ear canal resonances?  I tried to do it with a cheapo headphone the Panasonic HTF295 and this is the EQ curve I came up with:



"Looks like someone at Beyer trust his ears more than his his dummy head"... indeed?




Mountains = flat ?



Originally Posted by ultrabike View Post

Based on Purrin's work, the T5P present quite a few notches that would likely be very challenging to even equalize. A shame Beyer is putting up this lackluster product on the market at any price (let alone > $1000)


Well, I heard Beyer is actually releasing an equalizer preset especially designed for the T5p, to make live recordings of rock and classical music sound more realistic, like near field studio montior speakers.

post #770 of 937
Originally Posted by kiteki View Post




Mountains = flat ?


Well, if you look at PiccoloNamek's EQ thread, a lot of the final EQs people posted look as jagged as that Beyer response plot--especially big circumaural headphones, those seem to have especially complicated interactions with the ear.  Now imagine that some of those headphones probably tested quite flat at the test bench, and imagine that someone at Beyer were to tune the headphones to HIS ears to sound flat (after compensating for the equal loudness curve) to his EARS instead of to instruments, it would be like taking one of those jagged EQ curves posted in the EQ thread and "burning" them into the headphones' actual frequency response, and the result might measure something like what purrin got.  I'm not saying it's a good idea (since such phones would sonically only fit his ears), I don't even think it's possible to tune headphone response in such a detailed manner (note that even the custom custorm UERM only provides tuning for bass, mids and treble...), but it could be one explanation for what we're seeing.

post #771 of 937
Originally Posted by rhythmdevils View Post

Beyer T5p





Fostex TH900




They both look pretty bad to my untrained eye, compared with some of the better ones posted here, and the near perfect result xnor obtained by convolving (I still don't know which phone's) impulse response with its minimum phase inverse filter.  In fact after seeing that result I reckon all the CSD plots here should be convolved with a minimum phase inverse filter first, to see how much of the shapes in the waterfall result directly from the magnitude response (e.g. an FR with as many peaks and sips as the T5p can be expected to have a ringing CSD.  But why doesn't the big dip of the Fostex near 10kHz or the big hump before that ring nearly as much?) and how much is extra non-minimum-phase resonance that can't be eliminated by correcting the magnitude response.

post #772 of 937
Thread Starter 

The TH900 is a colored headphone. Personally, I have issues with its treble spike at 7k (I cheat and use EQ). But it's much more tolerable to me than the Beyer spike. I'm not sure why. It could be that the TH900 ridge is broad, not well formed, and dies to dark dark blue by 1ms.


The 7.5-8k spike on the Beyer is well defined, much sharper, and longer lived, dying to dark blue by 2.5ms. The Beyer also has a sharp well defined ridge exactly one octave higher at 15-16kHz. I have never found sympathetic ringing one octave apart to ever be benign.

post #773 of 937

The test result of TH900 meshes very well with my impressions at last headphone festival in Tokyo: <<I listened to the high-end Fostex amp (HD-A8) and phones (TH900) with my own music (iPhone feeding the a fostex DD then optical in to the amp). It was very nice sounding overall, nothing wrong jumping at me on the short listening. The only thing I did not like with that slight harshness / sibilance when playing Miles Davis "Someday my prince will come". I know the trumpet is not supposed to sound muffled on that recording but at least I have no issue with any of my stax gear and with other gears I sampled on that day.>>


I never bothered to check the frequency content of this recording but wouldn't be surprised it aligns with the heapdhone peak at 7kHz.

post #774 of 937

Any measurements of the K400-501 series on the way, Purrin?

post #775 of 937
Send him a pair.
post #776 of 937
Originally Posted by MomijiTMO View Post

Send him a pair.


I believe tdogzthmn are doing this already.

post #777 of 937
Thread Starter 





Just heard both the SRH1840 and SRH1440.


I laughed so hard when I heard the SRH1840 because it sounded so low/mid-fi given it's price. The tonal qualities were actually OK. I know some people had some issues with its treble, but I was fine with it. A little shifted up with not enough bottom end, but no ringing. But ZOMG! Massive distortion as if Shure picked up from random drivers from Radio Shack or Mouser, and stuck them in there. The 1840 sound like slightly more refined full sized ibuds


The SRH1440 response is even more screwy than the 1840. Very unbalanced and shifted to the upper mids and treble. No ringing though.





1840 CSD


1840 Distortion



1440 CSD


1440 Distortion



By far, the worse distortion measurements I have seen yet. Not only is the distortion high, its mostly composed of odd order distortion (D3/D5). The 1840/1440 get the DISTORTION OF THE YEAR AWARD. At least CSDs are clean. But distortion of the SRH1840 is hard for me to ignore. The SRH1440 is too unbalanced for me to even worry about distortion.

Edited by purrin - 8/26/12 at 8:39pm
post #778 of 937

I was about to bash consumer headphones, but then I realized they are probably marketed as pro headphones. redface.giffrown.gif


Idk in any case, hi end audio is such a circus.  "Hey guys lets release new headphones every year that are no better than the ones last year and change the marketing and shape of the plastic rolleyes.gif.  Add some placebo golden fleece earpads polished with snake oil for extra bass.  90% profit...problem consumer? *trollface*"



lol anyway rant aside, thanks for measuring all these headphones to save us money biggrin.gif.

Edited by ZeNmAc - 8/27/12 at 12:18am
post #779 of 937
Thread Starter 

Other than the HE400 or maybe a few others, none of the new crop of headphones has been particularly impressive. A decent modest headphone like the ATH-M50 still beats the crap out of the SRH1840/1440, K550, D7100, etc. See http://www.head-fi.org/t/583950/shure-srh1840-and-srh1440-unveiled/1650#post_8647821 for a distortion analysis of some these compared to an old standard, the HD650.


It's unbelievable the amount of crap which is being released.

post #780 of 937

It seems they have slightly less bass distortion than the ksc-75/35/porta pro.  How would you compare the distortion in the rest of the spectrum vs ksc-75?

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Headphones (full-size)
Head-Fi.org › Forums › Equipment Forums › Headphones (full-size) › Headphone CSD waterfall plots