[Currawong]

Ok, I started up Play, which can access the 31-band equaliser in Mac OS X and tortured myself for another 5 minutes with Sweet Jane. The HeadRoom graph verifies what I discovered, it's definitely the 6.3kHz (or so) peak that's the issue. If I drop the 6.3k slider down, the ssssshhhhhwwwweeeeeeet Jane is down to the same, less noticeable level it is with all my other headphones. John was right, the recording does have sibilance, just it's at the worst possible frequency point in that recording for the HD-800s.

Now I'm stuck at what to do. If this makes a great deal of my favourite music un-listenable, then it'll have to be a return, not an exchange I do. Then I can consider buying another pair if, in the future, they fix this.

 

[muz640]

More or less the opposite of the HD650

 

[pompon]

Quote:

Originally Posted by Skylab /img/forum/go_quote.gif flavor, which is a slightly overdamped bass, and a slightly hot treble. It's also possible that those two traits are related.

I 100% agree with your conclusions.

You need to match with proper gear to avoid to overbright them. It's why tube amp help that area.

 

[Skylab]

Interesting that the HD800 frequency response, in the Headroom graph, certainly doers correlate very strongly with what I seem to be hearing. That is a very noticeable peak of almost 8db at 6 kHz versus 3kHz. I have not seen that graph before, but to look at it now, I cannot see how anyone can argue that such a peak won't have SOME audible consequence.

 

[k3oxkjo]

Quote:

Originally Posted by muz640 /img/forum/go_quote.gif More or less the opposite of the HD650

Which, of course, many say is "dull" or "muffled". What's a designer to do?

In the case of treble balance, it seems there is a range of "correct answers" (just as treble balance depends on where you sit in the concert hall), the designer, though, has to pick just one.

Bottom line, if our preference for treble is in substantive agreement with the designer's, the phones treble is "perfect" and he's a genius, if not, the headphone is "broke" and he's a bum...

To me, the HD-800 is balanced pretty much like where I prefer to sit in the concert hall (around 1/3 of the way back from the orchestra), so nothing seems awry to me. But objectively, I recognize that if you like to sit around the 2/3 point you will find the HD-800 a bit bright and if you like a back-of-the hall or back-of-the-balcony perspective, just plain bright. And some folks like the sound from the lobby...

 

[Skylab]

Right - and this is, in the end, the only point I have been trying to make - that the HD800 *does* have a flavor. It might be the exactly perfect flavor for many listeners. But it is not perfect for all listeners, because it does have some sonic characteristics that depart from neutral. Kinda like every transducer that has ever gone before it

 

[k3oxkjo]

Quote:

Originally Posted by Skylab /img/forum/go_quote.gif Right - and this is, in the end, the only point I have been trying to make - that the HD800 *does* have a flavor. It might be the exactly perfect flavor for many listeners. But it is not perfect for all listeners, because it does have some sonic characteristics that depart from neutral. Kinda like every transducer that has ever gone before it

Yes, OK. But, since we are talking treble, let's make it clear that you are talking about your sense of "neutral treble" here. Not some universal, objective and authoritative treble balance that we would all recognize as "neutral".

Eg, if Sennheiser dropped the treble 2 or 3 db in future production, I would say the early phones had a more "neutral treble balance". But this would not be authoritative, for some people the "problem" would be fixed.

Here is an excerpt from something I wrote for a DT-48 thread that may shed a bit of light on people's reaction in general to the HD-800:

"So far, the HD-800, in my view combines a technical correctness with a dollop of the hard to define magic such as the DT-48S has. Some of this is the midrange, which is nip-and-tuck with the Beyer. Some of it is the freedom from distortion which seems even better than the DT-48S. Some of it is the HD-800's ability to unravel musical threads and its' more spacious soundfield (still very different than a speaker system).

But I don't doubt that all this technical competence can sometimes lead to a sense of listening to the phone rather than the music. The DT-48S doesn't allow this, it just disarms you. You know there are things wrong, but you don't really care.

The HD-800 can sometimes appear to be a victim of it's own excellence, it's residual flaws can catch your ear despite their relatively small presence. This is not to say the HD-800 are perfect or even for everyone. I do think the HD-800 overall have to be compaired with the best..."

 

[powertoold]

If the problem is at ~6khz, then everyone should be somewhat annoyed by it, but many people can't hear it, even with old rock recordings that are generally harsh. Wouldn't this likely be some sort of bad HD800 copy problem, or does everyone really hear that much differently?

 

[Skylab]

Quote:

Originally Posted by k3oxkjo /img/forum/go_quote.gif Yes, OK. But, since we are talking treble, let's make it clear that you are talking about your sense of "neutral treble" here. Not some universal, objective and authoritative treble balance that we would all recognize as "neutral""

Actually I WAS talking about a universal neutral - ruler-flat frequency response. So I disagree with you - I was NOT talking here about "my" view of neutral in this case.

 

[k3oxkjo]

Quote:

Originally Posted by Skylab /img/forum/go_quote.gif Actually I WAS talking about a universal neutral - ruler-flat frequency response. So I disagree with you - I was NOT talking here about "my" view of neutral in this case.

Literal "Ruler flat" frequency response in the treble would be so bright that almost no one would like 'em! Except, perhaps, folks like my late father whose hearing was trashed in WW2, they might make good hearing aids.

Look at just about any headphones' measured response and you will see treble roll off, not "ruler flat"! In fact, look at the HD-800 response. The treble is rolled off!

Use a Tact or some other EQ box to make your home speakers "ruler flat" in the treble at your listening spot (I have done this). WAY too bright. You end up having to put in a "target curve" that inevitably includes a treble roll off. This concept is basically universally agreed upon.

What the target curve should be in the treble is where the controversy is, there is no univerally agreed upon standard, just like what the "proper" treble balance in a headphone should be.

 

[Skylab]

I'm not sure that ruler-flat resonse in the treble alone would be a problem. Look at the frequency response just in the treble - most headphones have significantly elevated treble versus the 1 khz level, including much of the HD800's until 10kHz. The response above 10kHz has little musical cntent.

Most headphones DO feature a rising lower midrange and bass reponse versus the 1kHz level, and I agree that if they did noit then they would sound awfully bright. But this elevated lower response I agree is required.

 

[tyrion]

Quote:

Originally Posted by Currawong /img/forum/go_quote.gif Ok, I started up Play, which can access the 31-band equaliser in Mac OS X and tortured myself for another 5 minutes with Sweet Jane. The HeadRoom graph verifies what I discovered, it's definitely the 6.3kHz (or so) peak that's the issue. If I drop the 6.3k slider down, the ssssshhhhhwwwweeeeeeet Jane is down to the same, less noticeable level it is with all my other headphones. John was right, the recording does have sibilance, just it's at the worst possible frequency point in that recording for the HD-800s. Now I'm stuck at what to do. If this makes a great deal of my favourite music un-listenable, then it'll have to be a return, not an exchange I do. Then I can consider buying another pair if, in the future, they fix this.

I've listened to Sweet Jane on vinyl using my speakers, HF2 and HD800 and the part she sings the words Sweet Jane, each produces a similar almost sibilant "S". There are differences between all three but all three have about the same sibilant quality. To my ears, there is nothing unlistenable about this but to each his own.

 

[JaZZ]

Quote:

Originally Posted by Skylab /img/forum/go_quote.gif Interesting that the HD800 frequency response, in the Headroom graph, certainly doers correlate very strongly with what I seem to be hearing. That is a very noticeable peak of almost 8db at 6 kHz versus 3kHz. I have not seen that graph before, but to look at it now, I cannot see how anyone can argue that such a peak won't have SOME audible consequence.

The treble response of headphones can't be accurately measured. The distance between measuring microphone and driver is beginning to approach the wavelengths of the signal, so there's a lot of dips and spikes telling nothing about the sonic result (see also: HRTF). The 6-dB spike most likely doesn't represent the truth.

The Stereoplay graph e.g. shows a plateau starting at 4 kHz and culminating at 5 kHz, whereas at 6 kHz there's some sort uf dip within the plateau.

......

Now which graph is closer to the truth?
.

 

[gordie]

Great song, but the version on plain CD is a terrible recording, at least the one I have. It is mono, and recorded in 1957 and it is very low in level with very diminished bass.

That's definitely piano at the end (starting around 3:45 on the recording I have, your mileage may vary since I don't have the SACD). Through 580s, Stax, and 800s it all sounds like piano to me, not anything like glass breaking with the 800s.

I wouldn't dream of using that recording to judge any audio equipment, but different strokes I guess.

And suing the Doors because Light my Fire sounded too much like this? Which part? I don't see that at all.... Actually if anyone should sue anyone Thelonious Monk should sue Gil Evans (or the other way around, but I think Monk started recording/playing out first).

BTW, my 800s are anything but lean (through a Benchmark), listening to good recordings of classical, jazz, rock, techno - it all sounds great. Maybe lean is a relative term if compared to severe basshead cans, but I'm getting all the bass I want out of them.

Quote:

Originally Posted by greggf /img/forum/go_quote.gif Good analysis, skylab. I have a Gil Evans SACD, "Gil Evans + 10", which was released on LP in the '50's. Toward the end of the last cut, "Jambangle" (for which Evans considered suing The Doors because it sounds so much like "Light My Fire"!), there's a bit of high-end piano tinkling that sounds like a piano on the HD650, HD600, and Stax O2. On the HD800, it sounds like fragile glass breaking. Which is correct? Piano or glass? Perhaps the piano SHOULD sound like glass in this instance, but I doubt it. I assume that this is the tipped-up treble, or the treble difficulties in general. Mr. Willett, please listen carefully to the HD800 when you finally actually get production samples. Keep in mind that your credibility here will be greatly enhanced if you hear at least some of the problems or issues discussed by others, and you will have no credibility whatsoever if you hear none, because there are no perfect headphones. The HD800 is good, but rather thin, lean in the bass, and tipped up (and weird at times) in the treble.

 

[Skylab]

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif The treble response of headphones can't be accurately measured. The distance between measuring microphone and driver is beginning to approach the wavelengths of the signal, so there's a lot of dips and spikes telling nothing about the sonic result (see also: HRTF). The 6-dB spike most likely doesn't represent the truth.

Perhaps you are unaware that the Headroom measurements factor out the HRTF?

From their website (the bold emphasis is mine):

Quote:

To perform this test we drive the headphones with a series of 200 tones at the same voltage and of ever increasing in frequency. We then measure the output at each frequency through the ears of the highly-specialized (and pricey!) Head Acoustics microphone. After that we apply an audio correction curve that removes the head-related transfer function and accurately produces the data for display.

I am not trying to say that the Headroom data are 100% accurate. But their measurements are consistent across different headphones, which provides a useful point of comparison.

And FWIW, I worked for a major microphone manufacturer for 8 years, and have taught classes on audiology at the University level, so I am pretty familiar with things like the HRTF.