[Arve]

Out of curiosity, how do you diyers think can reduce the 6khz peak? Microfiber cloth? Resistors?

Which peak?
 
They are a tad recessed above 2 kHz, and slightly hot at 9-10 kHz. It might be possible to change some of this by changing the ventilation at the rear of the driver - that'd come after any other mod I'm planning, though.

 

[squallkiercosa]

9khz peak. Drilling a small hole?

 

[Hawaiiancerveza]

The change is subtle at best, you'd do much better using taller earpads to enhance the soundstage, I'm still using a pair of fostex earpads stacked over the stock ones : /

Which Fostex pads are you using?  
Speaking of Fostex, not trying to derail but, has anyone heard the new models?  Th5BB and Th7?

 

[c61746961]

The stock T50RP ones, though I would not recommend anyone to get a pair to stack since the extra air volume will decrease the midrange energy and boost the midbass. They need taller and squishier pads, something similar to the hm5 pads would be ideal, I reckon.

 

[Jay_WJ]

I really do not want to pour water to excitement with the mod ideas. But here is my two cents based on my experience in speaker driver measurements (all dynamic headphone drivers are basically the same transducers relying on the same physics).
 
I have seen countless mod examples of headphones here at head-fi, MOST of which were done by ear, not by measurements. And all of them I've seen were supposed to have an effect on either damping or reflection, NOT on performance due to the driver's diaphragm resonance pattern or motor design. Tyll's measurements of the Tascam/Teac basically tells that there is not much to modify with these headphones. Possibly you may change their bass/midbass response a bit by adjusting damping, but you will not be able to improve their response behavior above 2 kHz to any audibly significant level, unless you can redesign/resource the drive unit itself. If you believe you hear a substantial difference, it will be only in your head.  EQ is your friend in this case.
 
-Jay

 

[xnor]

  I really do not pour water to excitement with the mod ideas. But here is my two cents based on my experience in speaker driver measurements (all dynamic headphone drivers are basically the same transducers relying on the same physics).
 
I have seen countless mod examples of headphones here at head-fi, MOST of which were done by ear, not by measurements. And all of them I've seen were supposed to have an effect on either damping or reflection, NOT on performance due to the driver's diaphragm resonance pattern or motor design. Tyll's measurements of the Tascam/Teac basically tells that there is not much to modify with these headphones. Possibly you may change their bass/midbass response a bit by adjusting damping, but you will not be able to improve their response behavior above 2 kHz to any audibly significant level, unless you can redesign/resource the drive unit itself. If you believe you hear a substantial difference, it will be only in your head.  EQ is your friend in this case.
 
-Jay

 
Why not? Cup resonances in the treble area can easily be changed with mods, even with nondestructive ones.

 

[lei111]

I'd agree that it's mostly the plastic and I could hear the ringing as well.  More breathable pads would definitely help.  I've tried some cloth and taller ones and it's better.  But they are not good fit and lack the cloth shield.

 

[Jay_WJ]

   
Why not? Cup resonances in the treble area can easily be changed with mods, even with nondestructive ones.

 
I highly doubt the dip and peak combination on Tyll's measured FR is due to the housing resonances. This dip and peak combo is VERY common with dynamic drivers, which cannot be remedied unless you design the diaphragm differently. You could tone down the treble and thus the tonal balance by inserting, say, some acoustic screen between ear and driver, but the dip and peak will still be there. Again, EQ is the best solution in this case. Sigh, a lot of snake oil, a waste of time and effort ... ... I understand if one says it's just for fun. But I wish them to quit this nonsense. I do not have an objection to all kinds of mod. A mod based on objective measurements for diagnosis and treatment will be very interesting, but there are not many, in fact, very few.

 

[Arve]

  I really do not pour water to excitement with the mod ideas. But here is my two cents based on my experience in speaker driver measurements (all dynamic headphone drivers are basically the same transducers relying on the same physics).
 

Just so we're clear here. Most of where this headphone fails does not have to do with the driver - I expect that if you want to make significant changes to the response curve, you will have to modify the motor or the damping of the driver itself.  That said - the driver on this thing _is_ pretty accessible, and I expect that it's possible to do some of the tuning by changing the membrane at the port holes of the driver itself.  Will need to be played with (and here I agree: Not by ear).
 
 
However, by far, the most prominent "issue" this headphone has is time-domain smearing due to enclosure resonances - when you compare to better headphones, there is some lack of finesse that you should be able to bring back by not having the entire headphone sing along with the music.

 

[Jay_WJ]

  Most of where this headphone fails does not have to do with the driver.
 
However, by far, the most prominent "issue" this headphone has is time-domain smearing due to enclosure resonances

Have you measured to confirm this enclosure ringing idea? You may hear the enclosure resonate at a very high SPL, but I would not ever wear them at that level. At the level the housing resonate that much, the driver should bottom out with a lot of low-freq distortion as well---there's no point dealing with the housing resonance in this case. I still doubt the mod that dampens the enclosure is worth your effort (i.e., significantly improves the anomaly shown in the FR). Again, I am 99% sure that the dip and peak combo in the measured FR of these headphones is an inherent property of the driver's diaphragm design. You will not be able to change that by making holes on the back, tuning the damping, etc.
 
Okay, I will not post on this matter again. If you doubt, go ahead perform the mod and measure.
 
Oh, by the way, enclosure resonances, if exist, should also show up clearly in a harmonic distortion sweep at integer fractions of the resonance frequency. Tyll's measurements do not show an indication of this phenomenon even at 100 dB---what I see is normal, driver-induced harmonic products at 4.5 - 5.0 kHz, the level of which is that I usually see due to cone resonance. By the way, their HD sweep performance is impressive for $30 headphones!

 

[xnor]

Indeed, and I agree with most of what you wrote Jay_WJ but even something as simple as swapping pads can cause audible differences in frequency response. There indeed may not be a resonance problem at ~10k and I didn't say there is.
I use EQ myself, because even a perfectly clean headphone usually still does not have a perfectly balanced and smooth FR.

 

[c61746961]

Besides, it is not the FR that is problematic, it is the ringing that we are trying to address (which does show in the measurements), and the current hypothesis is that the enclosure is the culprit. Yeah, unjustified so far, so what?

 

[xnor]

Besides, it is not the FR that is problematic, it is the ringing that we are trying to address (which does show in the measurements), and the current hypothesis is that the enclosure is the culprit. Yeah, unjustified so far, so what?

 
Nonflat FR necessitates ringing in the time domain and the ringing you can see in the impulse response is at 10 kHz, where the FR peak is.
 
It may or may not be the enclosure resonating at that frequency.
 
There also has to be ringing at 5 kHz due to the dip.

 

[Jay_WJ]

  but even something as simple as swapping pads can cause audible differences in frequency response.

 
I can't help but break the promise I made to myself ^^ Of course, replace the pads with socks. You will hear a BIG difference in tonal balance. What I meant was that by doing so you cannot change an FR anomaly due to diaphragm resonance which I am 99.9% sure causes the dip and peak above 2 kHz of these headphones. Did you notice that I added .9 to 99? Where do you think this confidence come from? Measure drivers yourself, or at least take a look at as many measurements of raw drivers as you can. This pre-ringing and ringing combination is typical of cone/diaphragm resonance. It is governed by physical laws---all drivers are affected. The driver designers can only control it using suitable tricks that alleviate the phenomenon. This is why high-end drivers take effort in designing a well-dampened diaphragm. But even in this case, the anomaly is just made less prominent. It cannot go away completely in theory.

 

[Arve]

Jay: Have a look at this frequency response comparison:
 
http://graphs.headphone.com/graphCompare.php?graphType=0&graphID[]=2861&graphID[]=2851&scale=20
 
That is the Sennheiser HD 598 and 558.  The driver on those two headphones are identical (quite literally the same part number - http://www.custom-cable.co.uk/sennheiser-replacement-capsules-hd518-hd558-hd598-each-left-drive-unit.html
 
Also: http://graphs.headphone.com/graphCompare.php?graphType=1&graphID[]=2861&graphID[]=2851&graphID[]=2821&scale=20 - that is the harmonic distortion spectrum of the HD 518, 558 and 598 in one graph.
 
(The difference between those three headphones is to be found in the ear cup, which is different between the three models.