[floppy-ear ted]

I'm looking for a pair of headphones that will sound pleasant and non-fatiguing, especially in the midrange frequencies. I've DIYed a couple of loudspeakers, so I'm looking for info from a slightly different technical perspective. For me some of the most useful information that speaker manufacturers have often provided were waterfall plots, aka: cumulative spectral decay plots. When I first saw such graphs I wasn't sure how to interpret them, but it's quite easy: basically they show a combination of time and frequency response, so you can pick out any frequency bands that are resonant and take a long time to decay. Although these plots are always unique for each speaker, there are 2 common groups:

1)Smooth and fast decaying up to a certain frequency, with a big slow-decaying ridge at say: 5kHz, and at higher frequencies it goes completely out of control. This characterizes a speaker with a very stiff cone that moves like a piston and sounds great, as long as the unwanted stuff is filtered out.

2)Reasonably smooth decay across a wide range of frequencies, but has lots of smaller resonances that are unavoidable. This indicates that the cone/dome is soft and flexible. The movement is not supposed to be pistonic except at low frequencies, and instead the coil provides mechanical energy that is absorbed by internal damping and the surrounding air. Just about everyone will be familiar with this sound if they've ever heard paper-coned woofers and fabric-domed tweeters.

I haven't been able to find any waterfall plots for headphones, and the closest I've seen was at Headroom when they still had those square-wave response measurements up. Those provided similar information, but it was hard to interpret (hence waterfall plots were invented). I'm not too worried if the frequency response isn't outstandingly flat, as long as it's smooooth, and the small, sharp seismic ripples (tell-tale sign of resonances) only start at a high frequency. I know that some resonances are unavoidable because of the air cavity between the ear and headphone speaker, but it would be useful to know how the different brands and models compare in that regard too.

I'm mostly unfamiliar with high quality headphones, and I don't really want to start playing around with lengthy auditioning and repeated buying and selling. Here in New Zealand almost any headphone costing above appox. US$80 is only available from snobbish hifi shops who give themselves a 100% profit margin. Therefore, it's far cheaper for me buy directly from the US or Germany or elsewhere, even though I have to pay extra for transport and tax. This means that I have to do some research before purchasing.

I once listened to the Sennheiser HD-580 "Jubilee" with the intention of buying them (heard through a cd player with a built-in valve headphone amp). They were made with carbon-fibre, and were apparently identical to the up-and-coming HD-600 and made for "hyping up" purposes. I was not impressed with the sound. I didn't like the loud "warm and intimate" mid-bass, with a slight drop-off at very low frequencies. The treble was droopy, giving me a disconcerting feeling of having cotton in my ears. That could have all been fixed with a bit of EQ, but the midrange is what killed them for me. I hated them right from the word go, and I couldn't put my finger on the reason. At first I suspected that the valve amp was producing noticeable harmonic distortion, but it didn't sound that way, and the bass was quite clean.

Perhaps I was expecting too much improvement over my cheap closed-back Sony MDR-CD170, but (regardless of how morally corrupt this may sound) I felt that the Sonys did some things better than the Sennheisers. I read about the large flexible "Duofol" diaphragm design that Sennheiser uses, and it all fell into place. The old 100Hz square-wave measurements at Headroom showed that although the HD-600 headphones had the nicest frequency response, they had a lot of delayed energy, and there were small ripples that kept on ringing for a long time after the initial "step".

I realised that this all reminded me of paper-coned woofers, and was exactly the kind of sound I was trying to avoid. On the plus side, because the diaphragm is flexible, measurements show that Sennheiser does an outstanding job of absorbing resonances formed in the headphone cavity. Nevertheless, I haven't found those "cavity" resonances to be obtrusive on other headphones, and I'm still yearning for the type of clean sound that I can get from loudspeakers with hard (metal/composite/ceramic) cones. I've set my sights on the Beyerdynamic DT 880, but would like to know what other people would recommend for me, and are there waterfall / step-response plots around that could help me make an informed decision?

FET

 

[Eagle_Driver]

On the other hand, I hate the sound of most closed-back headphones regardless of price. The trouble with those is that they (as a group) resonate way too much within the headphone cavity, causing honky, echoey, tinny and ringing sound. And plotting the transient response curve of those closed headphones showed that there is a lot of delayed energy, and ripples that kept getting larger and larger for a while even after the initial step. The result is an overly boomy, muddy, uncontrolled bass (especially upper bass) sound.

You say that you dislike the sound of paper-cone woofers? Well, the sound from most closed headphones reminds me of a cheap, poorly-designed polypropylene-cone woofer that has been installed in too small of an enclosure - the non-reinforced plastic material itself "waffles", or resonates, even more uncontrollably than paper ever could.

 

[Geek]

Cool.

I don't hear any resonances in the HDxxx though.

A headphone's frequency response in and of itself is a good judge of resonances and flatness at the same time, I think. Both accuracy and resonance characteristics can be seen in these graphs. This is a valid statement because as a test tone sweep is produced, those frequencies which cause resonances will show up as bumps on the chart. Simple as that. It's impossible to tell if a "bump" is due to resonance or just boost at that point, however. I'm just saying that what you are referring to is already noticeable to some extent on standard graphs.

I'd take my ears over information any day for judging phones, but it doesn't hurt to have both.

Cheers,
Geek

 

[Geek]

Also it would be GREAT to make several measurements of a really high end phone (such as the HD650) using a series of amps, stepping up from the low-end to the ultra high-end. It woudl be neat to see how the characteristics of both frequency response AND decay would change as you moved from equipment that barely keeps up with the phones to stuff that's actually acceptable.

Final edit:

Another thing to do is have a live reference. In reference to what do you refer to your music as "wooly" through HDxxx phones? Do you even have a reference? Remember, without a reference to UNAMPED, ANALOG REALITY then this whole hobby is MEANINGLESS.

I prefer violin myself.

Cheers,
Geek

 

[floppy-ear ted]

Quote:

Originally Posted by Geek Also it would be GREAT to make several measurements of a really high end phone (such as the HD650) using a series of amps, stepping up from the low-end to the ultra high-end. It woudl be neat to see how the characteristics of both frequency response AND decay would change as you moved from equipment that barely keeps up with the phones to stuff that's actually acceptable. Final edit: Another thing to do is have a live reference. In reference to what do you refer to your music as "wooly" through HDxxx phones? Do you even have a reference? Remember, without a reference to UNAMPED, ANALOG REALITY then this whole hobby is MEANINGLESS. I prefer violin myself. Cheers, Geek

Yeah, I know that the Senns were generally a lot better than the Sonys. But I just felt that the midrange was a disappointment with the Senns. I didn't mean to say that the actual quality of the midrange was better in the Sonys than in the Senns, just that for the Sonys it was a relative strength compared to the rest of its bad performance, and for the Senns it was a relative weakness. I've slightly modded my Sonys by adding some badly needed cotton wool into the back chamber, and de-squeaked them with some oil. So they're not as bad as they would be straight from the shop.

For the home hifi I've got an AMC cdp with 6x60W Rotel power amp, home-brew pre-amp + active subwoofer XO, and DIY loudspeakers using Accuton speakers with ceramic inverted domes. The speakers are the real gems in an otherwise average system. All the paper/polypropylene speakers and metal tweeters I've ever heard sound like cr@p in comparison. But yeah, nothing surpasses a live chamber emsemble as a reference.

I know I can't expect the same level of stereo sound from loudspeakers as I can get from headphones, but I was hoping for headphones to be able to give me a similarly smooth tonal quality. My loudspeakers are a real blessing for female vocals, and difficult instruments such as oboes and saxophones sound great. Violins sound good too, but I think the jittery DAC in my cd player isn't quite up to that particular task.

FET

 

[JaZZ]

As a (former) speaker builder I'm an advocate of hard (especially metal) cones myself -- combined with steep low-pass filters, of course. At the same time I've been happy with the HD 600, among others for the reasons you mention (low cavity effects), but I agree that it has some of the sonic properties of paper cones, which give it a warm, but also slightly mat character. BTW I like its midrange anyway. It's a matter of synergy with the associated electronics.

To my ears what comes closest to the hard-cone sound is the HD 650. It has exactly that quickness, smoothness and purity. And it preserves the typical Sennheiser care for low reflections and resonances. I think you won't find any other headphone which comes closer (maybe the AKG K 1000 is another candidate, but I haven't heard it). Oh... and if you want the treble (which is absolutely great in stock configuration) to match the clarity of your ceramic tweeters (Thiel?) you need to attach one of the top aftermarket cables. I use the Zu Mobius and am awaiting the Moon Audio Silver Dragon.

And BTW: Welcome to Head-Fi!

 

[Takashi]

just a thought of an amatuer....

headphone's drivers are smaller and lighter, they should have faster decay than any speakers. i guess that's in fact the problem why headphones don't sound like speakers, but on the contrary they give more details. i just speculate hi-end headphones are meant to have more slower decay or resonance to come closer to speakers.
why don't you try some "fast" phones such as grado sr series, or monitor headphones such as akg k271studio. i don't know but you may like it.

sorry, but i just love butting in.

 

[myself, aka me]

Grado + NZ = expensive

Try Alessandro-Grado. Perhaps the MS-Pro although it's more expensive than the DT880 by about 2x.

If you get a DT880 you'll need to amp it nicely. I sold mine because I just couldn't like the sound with my META and i'd rather not buy another amp for possible improvement.

Aren't electrostatics 'fast' or am I thinking of another term?

 

[floppy-ear ted]

Quote:

Originally Posted by JaZZ As a (former) speaker builder I'm an advocate of hard (especially metal) cones myself -- combined with steep low-pass filters, of course. At the same time I've been happy with the HD 600, among others for the reasons you mention (low cavity effects), but I agree that it has some of the sonic properties of paper cones, which give it a warm, but also slightly mat character. BTW I like its midrange anyway. It's a matter of synergy with the associated electronics.

Thanks JaZZ. I haven't discounted the possibility of the tube amp (through which I heard the HD-580 Jubilee) from stuffing things up. However, what I heard seemed to match the frequency response measurements done by Headroom.

I'm not 100% sure what to look for in the measurements from Headroom, but I've sure found lots of interesting stuff. The Etymotics had insanely high distortion, yet people always seem to give them good marks for sound quality. So perhaps it doesn't matter so much if (mechanical) harmonic distortion isn't as low as that of other headphones? Anyway, Etys are not for me. The only thing I want directly connected to my eardrums is air.

The HD-580, 600, and 650 frequency graphs look very similar, yet people say there are big sonic differences. I've wondered about this, and suspect that one important aspect of sound quality is having precisely matched sound on both sides. The measurements show that above 2kHz the 650s have the best amplitude matching of the three. Considering that phase shifts are linked to the frequency response, this would explain people's comments about the 650s having a better soundstage. I wonder if Sennheiser just make 1000s of those speakers, measure them, and mate the ones with the most identical graphs for the 650s, and the "seconds" go to the 600s? All the subtle differences on the graphs at high frequencies seem a bit too random for there to be any predictable trend. Someone could have accidentally swapped the frequency response of the Left HD-600 with the Right HD-650, and no-one would be the wiser.

When looking at speaker measurements, the main reason that I've ever found impedance graphs of any use is when they were accurate enough to show little glitches at mid to high frequencies. That indicated where any major resonances were, and again I've found this useful when looking at headphone measurements. The top 3 Senns all have a tell-tale impedance peak at around 3-4kHz, and it's damped much more effectively in the 650 than in the other 2 (600 & 580).

I don't fully understand why headphone manufacturers can't make headphones that match an ordinary tweeter in the top end. They make huge (30mm - 50mm) shallow diaphragms instead. Couldn't they do something like 10mm diameter metal dome speakers, and recommend that people use current-controlling amps instead of voltage-controlling amps? That would actively absorb resonances created in the ear cavity. It would however, only work with open-backed headphones, and even then it would slightly decrease the isolation of external noises. I actually intend to build a DIY headphone amp that does this, but first I need suitable headphones.

Btw: yeah Accuton's just another name for Thiel & Partner.

FET

 

[Eagle_Driver]

Headphone design is a compromise. Very small drivers (10mm) have never worked well in large-cup headphones - in such a large enclosure, tiny drivers provide no bass at all whatsoever (regardless of the material), nor can they provide any response at all whatsoever below the midrange. What's worse, very tiny drivers placed in a large enclosure have a severe peak in the lower treble - in fact, so severe that your ears would be left ringing even at extremely low listening volumes! And 10mm drivers are even smaller than the drivers that are used even in really cheap earbuds! (Sorry about that ranting, but that's just the law of physics that cannot be circumvented.)

On the other hand, if you're talking about using tiny drivers to provide high-frequency response in addition to the larger driver used for low-frequency response, well, such two-way designs have never been very successful in headphones. Many of the bigger flops in headphone design have been two-way designs, with their harsh, incoherent sound - worse than even the Sennheiser HD-590 that I have been criticizing recently.

 

[floppy-ear ted]

Quote:

Originally Posted by Eagle_Driver Headphone design is a compromise. Very small drivers (10mm) have never worked well in large-cup headphones - in such a large enclosure, tiny drivers provide no bass at all whatsoever (regardless of the material), nor can they provide any response at all whatsoever below the midrange. What's worse, very tiny drivers placed in a large enclosure have a severe peak in the lower treble - in fact, so severe that your ears would be left ringing even at extremely low listening volumes! And 10mm drivers are even smaller than the drivers that are used even in really cheap earbuds! (Sorry about that ranting, but that's just the law of physics that cannot be circumvented.) On the other hand, if you're talking about using tiny drivers to provide high-frequency response in addition to the larger driver used for low-frequency response, well, such two-way designs have never been very successful in headphones. Many of the bigger flops in headphone design have been two-way designs, with their harsh, incoherent sound - worse than even the Sennheiser HD-590 that I have been criticizing recently.

What you say can just as easily be applied to diaphragms with a larger diameter. A 50mm speaker is just a small midrange that wouldn't normally be expected to go far below 1kHz. But we're talking about a tiny volume of air of something like <200mL, and it seems that most headphones produce resonances from that anyway. I roughly worked out that a 10mm diaphragm playing into a volume of about 200mL could easily produce 120dB at low frequencies if it moves 1mm peak to peak, which is far too loud anyway. A current-controlled amp would be useful here because it would allow sound pressure peaks to escape through the diaphragm, so the air cavity would not be like the inside of a sealed speaker box.

Edit: I forgot something else which is: it looks like most headphones tend to use piece-of-junk speakers anyway, and most of the money paid goes towards ergonomics.

FET

 

[JaZZ]

Quote:

Originally Posted by floppy-ear ted Thanks JaZZ. I haven't discounted the possibility of the tube amp (through which I heard the HD-580 Jubilee) from stuffing things up. However, what I heard seemed to match the frequency response measurements done by Headroom.

I'm not sure about which graphs you're referring to -- however, it's hard to draw final conclusions from headphone frequency responses other than in comparison between different headphones, and even then they don't absolutely match the sonic impression. The reason is that you'll always have apparently arbitrary peaks and valleys in the upper frequencies due to comb-filter effects caused by reflections between driver and ear. Apart from that the closeness makes a slight treble roll-off necessary to make it sound flat.



Quote:

The Etymotics had insanely high distortion, yet people always seem to give them good marks for sound quality. So perhaps it doesn't matter so much if (mechanical) harmonic distortion isn't as low as that of other headphones?

I've also noticed the high distortion with the ER-4, and I have troubles to believe these measurements -- compared to to what I hear. I speculate that there could have been problems with keeping the drivers mechanically calm during the measurements -- in view of their tiny mass. I really think they have very low distortion, at least with mids and highs.



Quote:

The HD-580, 600, and 650 frequency graphs look very similar, yet people say there are big sonic differences. I've wondered about this, and suspect that one important aspect of sound quality is having precisely matched sound on both sides. The measurements show that above 2kHz the 650s have the best amplitude matching of the three. Considering that phase shifts are linked to the frequency response, this would explain people's comments about the 650s having a better soundstage. I wonder if Sennheiser just make 1000s of those speakers, measure them, and mate the ones with the most identical graphs for the 650s, and the "seconds" go to the 600s? All the subtle differences on the graphs at high frequencies seem a bit too random for there to be any predictable trend. Someone could have accidentally swapped the frequency response of the Left HD-600 with the Right HD-650, and no-one would be the wiser.

Well, the HD-650 drivers are said to be matched to 1 dB deviation. But I heavily doubt this plays any important part. The main sonic difference is the result of a new membrane geometry and improved membrane damping. It's confirmed by the distortion measurements: of all the headphones measured the HD 650 has by far the lowest distortion. Which is very obvious once you hear it.

You should know that generally the frequency response doesn't tell you all about the sound. And the difference between hard and paper cones has nothing to do with it -- commonly they aren't very different if at all in their operating range. Where the differences are is -- there you have it -- in the decay spectra. The sonic difference between HD 600 and HD 650 reminds me very much of the one between hard and paper cones. Balance-wise (read: as to the perceived FR) the difference is not so pronounced.



Quote:

I don't fully understand why headphone manufacturers can't make headphones that match an ordinary tweeter in the top end. They make huge (30mm - 50mm) shallow diaphragms instead. Couldn't they do something like 10mm diameter metal dome speakers...?

Headphone membranes are -- with very few exceptions -- partial vibrators, comparable to (very) soft polymer cones or even Manger transducers. This implies that when it comes to high-frequency reproduction they don't have to move the whole moving mass of the whole membrane, but only part of it. Like e.g. a cone tweeter. Besides the short distance to the ears with the consequential low power loss allows very low excursions, a further advantage in view of transient response. The larger radiation area compared to 1" tweeters is needed for the adequate reproduction of the whole audio range with reasonable dynamics. I guess their very thin membranes in relation to their surface -- necessary for low moving mass/wide frequency spectra -- don't allow very large travel, otherwise harmonic distortion would be unhealthily increased. All in all I'd say headphone drivers have no disadvantage compared to dome tweeters. Nevertheless I've once been tempted to design a headphone dome driver. But meanwhile my constructing enthusiasm has left me.

 

[JaZZ]

Quote:

Originally Posted by floppy-ear ted ...it looks like most headphones tend to use piece-of-junk speakers anyway, and most of the money paid goes towards ergonomics.

No, no, that's absolutely wrong. There's a lot of development work behind headphone drivers. (I'm not speaking of cheap earbuds and the like.) Don't let you mislead by the looks!

 

[wallijonn]

Have you looked into electrostatics?

 

[Geek]

Good headphones plugged into good associated equipment have one hell of a great transient response. The sound of HD600s through a bluenote turntable and my MOH amp is the fastest and most detailed I've ever heard something reproduce sound before, short of real life.

Cheers,
Geek