CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS
Lobby => Headphone Measurements => Topic started by: Marvey on September 03, 2012, 01:03:09 AM
-
At the H3 meet, three of us, including Jason, noted that Jason's HD800 was pretty darn harsh, even with the Anax mods*. We compared threeunmodded HD800 with Jason's modded HD800 and still noted that Jason's HD800 was harsher and grainier in the treble. It wasn't like these HD800s sounded completely different. Jason's HD800 still sounded like an HD800. Just that it was a little bit worse than the other up in the treble.
To help unravel this mystery, Jason sent his HD800 (along with the review Mjolnir) for testing. It should be noted that the s/n is 16K. It's a newer HD800. Note that measurements are without the dust covers on unless indicated.
So do you guys see anything interesting? I will follow up with measurements from other HD800s.
* These were from the Tyll cutouts using one layer 2mm foam and thinner ring. Note the official Anax mod is different.
-
Here are Anax's. No mods. No dust cover.
-
8kHz adding to the 16kHz on Jason's pair making some not-niceness?
-
Here is HeadRoom's s/n 5xx. An early one. No mods. No dust cover.
-
Here is Max's. No mods. No dust cover.
Added March 24, 2013: Paradoxper's HD800 s/n/ ~20k. No Mods. No dust cover.
-
Gotta run to dinner with family. BBL.
-
The schiit one looks to have a mild case of the double-octave-penetration at ~7 and ~14khz.
-
Interesting that 6k on Jason's pair isn't a peak either since 8k is pushed up so the FR plateaus from 6k+. I would think this would make the treble sound more even, but as Mav said you do get that double octave thing going on.
-
Max's HD800 does sound the best - the smoothest.
-
Max's HD800 does sound the best - the smoothest.
Why? What's missing from the others? Is Max's HD800 an anomaly amongst the various HD800 samples you've heard?
-
They are not that much different. They all sound like HD800s. Just normal variability with production. Amazing that Senn matches the drivers so well. In contrast, my super LCD3 (which lasted less than 24 hours) and my veiled LCD3 were completely different headphones.
My HD800 (not shown) is actually similar to Anax's, but they are within 500 s/n.
Interesting to note that the HeadRoom unit was s/n 500 and Jason's was a very recent production. I highly doubt Sennheiser has changed the formula for the HD800.
Note the curves for Max's HD800 decay are the smoothest.
-
Interesting! Seems the extra 7.5k/8kish resonance might be the culprit here. Feel free to pull the foam out if you want to see how that affects things.
Also, those headphones are a bit green. We bought them after our long-term loaner went back to their owner, so they probably only have 60-80 hours on them. I thought these sounded brighter, but I put it down to mods/no mods/ears playing tricks on me.
-
I'll take the mods out and re-measure. Then burn them in with white noise at 80db with a 1 minute cool-off every five minutes for the next day or two.
-
Max's HD800 does sound the best - the smoothest.
Why? What's missing from the others? Is Max's HD800 an anomaly amongst the various HD800 samples you've heard?
Max's has a 1-1.5dB rise along 1Khz to 4khz. In other words, his mids are less recessed.
-
Man, this thread makes me curious to hear the Mjolnir with another set of HD800s now.
-
Took out the Anax mods*. This mod was from the Tyll cutout pattern with the skinnier rings. Only one layer of 2mm foam was used. Note that the official Anax mod uses thicker rings which extends the size of the "baffle" somewhat.
Interesting. So WTF...
-
Interesting. Does it sound better without the mods?
-
It does look like the resonance got shifted to 10kHz. I can't say for sure it sounds better this way. I do think it sounds better... Was playing the same tracks before I tore out the mods.
I need to work in the official Anax mods on this pair.
-
FYI - we eventually turned Jason's HD800 into this: http://www.changstar.com/index.php/topic,497.msg9368.html#msg9368 (http://www.changstar.com/index.php/topic,497.msg9368.html#msg9368)
-
I don't understand why the measurements from goldenears website, look
much worse. There must be a problem with their setup ?
(http://i50.tinypic.com/ba8i0.png)
-
I don't understand why the measurements from goldenears website, look
much worse. There must be a problem with their setup ?
That spike and resonance @ 2.5khz doesn't even exist on most if not all charts. I certainly don't hear it at all as I'm extremely prone to vocal sibilance w/ phones that highlight it. You have drop the floor below -50dB to get anything remotely similar for the HD800 CSD but the spike is still is a mystery. Some kind of artifact or error.
I want to say he measured a K701 and labelled the wrong phone. Who knows.
-
Marv can explain this better, but not only do they pickup resonances where there clearly are none, and not pick up resonances where they clearly exist, he also starts the measurement at random times, so the initial "plataeu" that represents transient response is completely innacurate. Makes some headphones look like the driver got stuck.
compare their M50 measurements to Marv's and you will never want to see another one of theirs again. No headphone could be that slow...
-
I'm getting sick of all this BS "My HD800 response graph shows I don't have the 6khz peak!" It's seems people have forgotten of don't know that the target curve of the the HD800 HAS the 6khz peak by design. So if your graph doesn't have one it means one of three things. Your non 6k headphone has a design flaw. They changed the target curve and never updated the marketing material. The newer measurements are rigged. Our tests of over 6 different HD800s (from new to old) seems to suggest the latter is the most likely.
This conclusion is based on two things. One, every new phone we measured that came with chart provided by Senn w/o the peak showed a peak on our gear just like all the rest. Two, we were able to replicate the newer Senn curves by altering placement of the headphones by a relatively few mm from reference placement.
Pic of the Target Reference curve:
(http://i.imgur.com/YtRgw7Ul.jpg)
General specs:
(http://i.imgur.com/8fwtpEvl.jpg)
_______
As for the theory of audible differences between new and old, it seems according to Mike at Headfonia, early bass light versions of the HD800 had earpads that were thinner with a larger internal volume compared to current pads. The pics of his old pads seem completely unlike any I have seen and have yet to see to confirm if this is the case. So it is another hypothetical as far as I'm concerned till proven otherwise. Be aware too, HD800 cups are large and it might be possible for a small head to not get a proper seal to ensure low end response. Upstream differences relative to audition instances are the other obvious variable.
-
Cool thanks for the info. It's good to be set straight by people who can measure several and show that the tolerances are low.
-
LOL, the HD800s I've measured were remarkably consistent. I don't know how Sennheiser gets their graphs. I assume that random technicians (who also moonlight as disgruntled euro youth protesting at G8 summits) haphazardly slap on the HD800s on the measurement heads and take a quick reading. I've measured the HD800 so much that I actually have precise guide marks on the measurement apparatus specially for the HD800.
-
I don't know how Sennheiser gets their graphs.
By the looks of it they use narrow-band noise, at least that would explain the stepped 'target' plot.
These measurement types 'remove' sharp peaks and dips in a 'smoothing' way and the filter bands are similar to the 'bands' our own hearing is 'divided' in.
They also use diffuse field compensation.
It is actually a good way to measure speakers as well as narrow resonances are equalized, better than sine-wave sweeps.
Deviations around the target are allowed, meaning they can measure differently a bit from sample to sample, so variances (not too big) are likely to exist.
Even Sennheiser is not able to make them all exactly the same as there are too many mechanical variables.
Sennheiser (like a lot of other brands) is also known to make small changes during production over the years.
It is clearly stated on the box you can download the actual measurements made by them of your individual HP (and they probably aren't measured after thousand hours of burn in and with fresh pads) and since you guys have the serial numbers perhaps download and compare...
That could be a fun exercise and since these are all measured on the same rig in the same way under the same conditions this could tell something about the spread in production.
Doesn't mean they aren't excellent headphones.
Of course Purrin has a good point about different people and perhaps different positions on the head.
I assume Sennheiser is aware of this though.
Note all of the above is speculation on my part and not based on actual facts.
-
I don't know how Sennheiser gets their graphs.
By the looks of it they use narrow-band noise, at least that would explain the stepped 'target' plot.
No it doesn't. Purrin can do the same thing by manipulating the graph to be a stepped visualization. He doesn't for obvious reasons.
These measurement types 'remove' sharp peaks and dips in a 'smoothing' way and the filter bands are similar to the 'bands' our own hearing is 'divided' in.
They also accentuate contrasts that are not as severe as they actually are. But yes, they can 'disguise' issues but it's nothing as bad as severe smoothing that some ToTL phones use. Here is an original LCD2 r.1 according to their old standards:
(https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcS2stdUecY8XiN8IjUaJIzMYCnMOOjLkutFDSf030a3hyvjXZpPJw)
They also use diffuse field compensation.
The charts we have correlated to our measures very well w/ the exception of the 6khz peak disappearing on some newer ones.
Even Sennheiser is not able to make them all exactly the same as there are too many mechanical variables.
Sennheiser (like a lot of other brands) is also known to make small changes during production over the years.
All the HD800s we've tested measure incredibly similar and are far more consistent than most any other phone I can recall off the top of my head. Channel matching is typically superb as well.
It is clearly stated on the box you can download the actual measurements made by them of your individual HP (and they probably aren't measured after thousand hours of burn in and with fresh pads) and since you guys have the serial numbers perhaps download and compare...
You can't 'download' them unless this is a recent thing. You have to request the charts be sent via snail mail in hard copy. Which we have correlated for most of our samples.
That could be a fun exercise and since these are all measured on the same rig in the same way under the same conditions this could tell something about the spread in production.
Exercise complete. http://www.head-fi.org/t/433059/sennheiser-hd800-certificate-for-frequency-response-arrived/330#post_8662946 (http://www.head-fi.org/t/433059/sennheiser-hd800-certificate-for-frequency-response-arrived/330#post_8662946)
Of course Purrin has a good point about different people and perhaps different positions on the head.
Ahem. Now I know how MF feels. ;)
Note all of the above is speculation on my part and not based on actual facts.
:)p5
-
Here is the Response Graph for my HD800's. Ser # 20521
(http://i987.photobucket.com/albums/ae352/longbowbbs/HD80020521ResponseGraph-page-001.jpg)
-
Thanks. So we can confirm that all new graphs have the 6khz peak neutered even though they all measure having the peak when tested independently. Yes, we haven't measured yours but all the new batches we have are basically the same at 6khz. Nobody is special except for the luck of having .5-1.0 dB less at the peak or .5-1.0 dB more in the mids. That's it.
I dare anyone to send us a HD800 without a 6khz peak. If your HD800 doesn't have one, we'll cover shipping both ways. If it does, you cover shipping both ways.
Sorry, the new graphs are deliberately misrepresentative relative to their old measurement standard. When you don't like the score, you change the rules of the game.
Sennheiser should do what Audeze did and start printing randomized serial numbers to stop this nonsense.
-
Remind me to post that on HF
-
Finally... Nobody seems to notice the Anax-mods stuff I've posted in the HD800 appreciation thread. However. Discussing cables for 300 USD. No worries.
"I don't want to modify my perfect headphone..."
Thx guys for doing this... I'm loving my Rabid Dog-modded one. Pure bliss. :)p6
-
A recap: Paradoxper was kind enough to send his HD800 via my way.
Paradoxper's s/n ~20k
(http://www.changstar.com/index.php?action=dlattach;topic=488.0;attach=3110;image)
Jason's (don't remember s/n, but fairly high)
(http://www.changstar.com/index.php?action=dlattach;topic=488.0;attach=2204;image)
Maxvla's
(http://www.changstar.com/index.php?action=dlattach;topic=488.0;attach=2212;image)
Anax's s/n 13k
(http://www.changstar.com/index.php?action=dlattach;topic=488.0;attach=2208;image)
HR's s/n 500
(http://www.changstar.com/index.php?action=dlattach;topic=488.0;attach=2210;image)
-
CONCLUSION:
While there are minor variances among the HD800s, the HD800 have the most consistent measurement results to date of any headphone. This is despite the variances in the Sennheiser provided graphs, which were probably performed by disgruntled Euro youth (the type who also moonlight as rebel-without-a-clue protesters at G8 Summits) who simply slapped the HD800s on the dummy measurement heads without regard for precise alignment (yes, precise alignment is necessary to get repeatable and consistent results.)
-
Looks pretty darn similar meaning no 'silent changes' in "newer" production units. Maybe Cory's has a hair faster decay? I'm no good at reading these graphs.
6khz peak is still there, I wonder if it's just my ears that aren't very sensitive in that region.
And it goes to show how many things affect perception and opinion, from false memory, to mood at the time, gear, tastes, bias, novelty factor, etc.
-
The hair cells in your inner ear which respond to 6k may have been destroyed by other headphones or occupational hazards.
-
My HD800 is 16021 fyi.
-
Oh no. :'( The only thing I've been doing for the last 17 years was sit infront of the desk studying. (21 years old) And the only can I've owned for a long period of time is LCD2. facepalm
-
Then it's possible you may grow to hear the 6k effect of the HD800 and start to hate it over the next 10 years. Many folks here tolerated the Grados for years.
-
Let's certainly hope that doesn't happen. I don't find snares hurt at all, they just sound... normal. The only thing that hurts my ears really bad is the beginning part of Sexy Back - Justin Timberlake. Yes I listen to pop; young fella here forgive me. :spank:
-
LFF and I love pop.
-
I like pop too. Its amazing we live in a day and age where we get to listen to this infinite amount of music. I love classical, pop, jazz, rock and so on. Its not weird for Schubert, Taylor Swift and Jimi Hendrix to be played on the same night here.
-
I like to listen to sentimental korean pop with female vocals. I didn't expect HD800 to perform this well, and with Anax mod, it's pretty damn good.
-
Eh, the 6k peak in HD800 causes some synth and trumpets to sound ear-piercing, didn't do one thing to vocals and is compensated by the 3k dip when it comes to cymbals and guitars, just causing a coloration change. Makes for a flatter soundstage than expected given its sense of space.
Granted, I had only a 30 minute listen of those...
-
Currently having 2 HD800s with me, 1 my own and another my friend's which is on a long term loaner. S/N019XX for my friend's with thin cloth pads. As you guys have noted, these are bass light but I like them better since they seem a little more neutraish and the 6k resonance does not sound as bad.
Current pair I own is S/N14k~ they have more treble and also moar bass. A little more v-shaped than the older pair.
Shoot me, I will be using these to listen to jpop and watch cartoon porn.
-
Currently having 2 HD800s with me, 1 my own and another my friend's which is on a long term loaner. S/N019XX for my friend's with thin cloth pads. As you guys have noted, these are bass light but I like them better since they seem a little more neutraish and the 6k resonance does not sound as bad.
Current pair I own is S/N14k~ they have more treble and also moar bass. A little more v-shaped than the older pair.
Shoot me, I will be using these to listen to jpop and watch cartoon porn.
Swap pads and report back if u can.
-
Swapped the pads once and on brief listen, seems that the older pads do take out some of the treble of the HD800s and I do think the stage presented is different with the newer pads sounding wider than the old ones.
As for product variances, the 2 HD800s I have are virtually identical in sound with the newer one I THINK sounding to have slightly(~1db?) more 6khz than the older one.
-
Goldenears.net has retested HD 800(objectively and subjectively).
http://en.goldenears.net/index.php?mid=GR_Headphones&category=275&document_srl=22602
Very strange results. It seems as though the elevated treble and midrange resonance of the previously tested model gave the false impression of excellent transparency.
Apart from this, some people say that HD 800 is mild/soft(e.g. avguide review) while others find them very sharp and transparent.
-
Goldenears.net has retested HD 800(objectively and subjectively).
http://en.goldenears.net/index.php?mid=GR_Headphones&category=275&document_srl=22602 (http://en.goldenears.net/index.php?mid=GR_Headphones&category=275&document_srl=22602)
Very strange results. It seems as though the elevated treble and midrange resonance of the previously tested model gave the false impression of excellent transparency.
Apart from this, some people say that HD 800 is mild/soft(e.g. avguide review) while others find them very sharp and transparent.
Another goldenears fiasco...
That's the most mismatched HD800 I've ever seen. And we've measured more than 10 units by now. I'm going to chalk it up to user error based on his inconsistent results and subjective impressions. I've also heard more than five HD650s and never have I heard them as being more 'transparent' than any HD800. That's just absurd. Probably what happens when he pairs the HD800 w/ his mysteriously unknown 'Goldenears' amp. It's been well known from tons of impressions the HD800 will react with and display whatever is upstream more than any other phone. His own inconsistent measurements and impressions point this discrepancy out. Our consistent ones indicate it's not the headphone that's the problem.
Just so people know if they don't already, all the dynamic headphones measured on this site use my O2 amp. Mostly to satiate the 'objective' crowd for a 'reference' amp (lol).
-
At least they did a better job than the previous time they measured it. 8)
The CSD seems 'better' now. ;)
FR is accurate (I think).
At least the few times I heard the HD800 I perceived it that way.
Strange that the 'subjective' score now is further away from 'ideal' as it was described in an earlier (and clearly faulty) measurement (this one http://en.goldenears.net/4326 I mean)
Impulse response is now excellent, would dare to say exemplary.
a little EQ improves it considerable.
-
FR is accurate (I think).
Which? The right or left channel? Perhaps he forgot to check for hairs on the left side... ::)
-
Which? The right or left channel? Perhaps he forgot to check for hairs on the left side... ::)
;D ;D ;D ;D
-
Driver matching seems excellent. We all know that high frequencies are prone to large variations.
As far as the subjective impression is concerned I haven't understand how exactly he perceives transparency. Transparency generally means fidelity. Mainly, lack of distortion and resonances(and no serious FR deviations).
Regardless of this, audiophiles, generally, are more divided on these cans than e.g Hd 650.
Personally, I can't accept the pseuso-scientific notion that amplifiers have their own magical properties and can improve the sound of headphones. We all know that correctly designed electronics are light years ahead of any imaginable transducer in terms of fidelity. And the main reason is feedback, the ability to correct mistakes almost instantaneously. So, to claim that a headphone can reveal the-orders of magnitude lower-flaws of your electronics is not very persuasive. Especially when we talk about equipment with infinitesimal current requirements.
-
Why does an electrical device interacting w/ another electrical device 'X' differently than electrical device 'Y' always come down to magic and pixie dust w. you people?
I also take your argument as non persuasive. The usual overly simplistic view of comparing apples to oranges because the real world is just too complex to deal with beyond the security blanket of a couple of basic graphs. It's like saying two cars are equally fast around a race track because they dyno the same. Nonsensical logic.
You statement that amps/dacs are 'light years' ahead of transducers in 'fidelty' is non sequitar especially when fidelity is still a moving target based on evolution of compensation curves. Also presumptious since DACs are in fact less developed than the transducer historically (they are relatively new to the audio world). Again, bad logic, apples and oranges. If people think as sloppy as this, they can't produce or analyze good data.
Based on your argument, the HD800 or other proficient transducers can't tell the difference between any modern Dacs or SS amps. Yeah, good luck w/ that fantasy. There's another site for that BS.
-
Amps have to interact with voice coils to produce sound. That's true.
The amp applies some voltage and fluctuates it to produce sound. The coil demands specific current at specific moments according to its characteristics. If the amplifier has the potential to deliver the demanded current, then voltage will remain unaltered and sound will come through undistorted. As you can understand, I don't adhere to the "synergy" theory. I believe only in the ability of amp A to drive headphone B by delivering what is needed.
Let's just say that we view things differently and stop our dispute.
Regarding our topic the most bizarre thing I have found is this:
http://www.head-fi.org/t/482386/building-a-headphone-measurement-lab/150
I am sure that purrin remembers it. The strange thing is, even if the methodology is wrong Tyll applied it on all headphones he had at the moment. The relative measurements, so to speak, indicate that HD 600 is clearer than HD 800. He never gave an explanation.
After that and this(look at bass distortion): http://headphones.reviewed.com/content/beyerdynamic-dt-880-review/science-page
I'm starting to question the validity of headphone measurements. Some differences are to be expected but this is just ridiculous.
-
Depends on how you define 'clarity'. If distortion was the answer, the 600/650>800 theory would be correct, especially if it was correlative to subjective impressions. The fact that a listener can easily point out specific instances of increased resolution and transparency to the original recording using the HD800 over the 600/650 tells you distortion is not the sole answer (as we have seen it is not for amps). However, it may certainly be part of the whole equation. If you want to actually pursue the argument, start by actually defining your terms first in plain English, then we can convert those definitions into electrical theory, math and science. Just posting a few sets of select graphs w/o strong correlations that are also in dispute with each other only serves to confirm the difficulty in measuring headphones, and that we need to take all objective data in relative contexts rather than universal absolutes. So far, I agree with you that measuring headphones and drawing strong conclusions from their data sets is difficult.
Subjectively, Tyll did just get a new amp from Doug which apparently takes it to a new level he did not appreciate back in 2010, perhaps even well into 2013. Might want to ask him for his impressions of clarity and transparency between the 600/650/800 w/ the new Doug amp. Or he could chime in if he sees this.
Rant inc. (warning)
My problem w/ the religious objectifiers is their utter ignorance and hypocrisy. They remind me of the fanatics that pop out of the woodwork regularly holding up their signs and screaming "the world is ending, repent now!" This claim has been going on for ages and is nothing unique, original or special about it. It's as if they have suddenly uncovered the golden tablets or Moroni or some shit. Nothing new under the sun to see here if you know the reference. Some of the folks I find to be the most fanatic advocates of 'science' are nothing but religious zealots disguised. The minute you stop questioning your results and beliefs, you are no longer a 'scientist'. Perhaps a glorified lab assistant at best. :-\ The hypocrisy also comes from their utter disregard for hearing as a sensory tool that can be honed and trained, then applied to a rigid methodology. A tool that has evolved as a necessary one to provide for our survival over millions of years btw. Yet, they go home every night, flip a light switch and trust their eyes! How do they objectively verify that a light bulb is actually working? Break a bulb, grab a meter and check for continuity across the filaments? No. They just look at it, trust their eyes and the repeatability of the subjective experience. So in eyes we trust, but ears we do not? Yet, they don't run around and say stupid things like, "how do you know the bulb is actually on, did you measure it?" "Maybe you're just biased towards seeing the light because the bulbs came in a prettier box." Just because bias exists doesn't mean you throw the baby out w/ the bath water. This type of trust/belief that the light bulb is in fact on is based in inductive logic, every bit as valuable as deductive logic. Perhaps even more so since it can be applied to wider ranges of conditions more quickly and easily. Yet ears become totally worthless in audio, especially for anything over $100 w/ these people. What a bunch of hypocritical garbage. And all because of a couple of pseudo experiments that get thrown around the web like the 10 Commandments despite their inability to mitigate the introduction of errors at various points along the way. Errors need to be mitigated in objective electrical testing just as they do in subjective listening.
Anyway, if you want to get to the bottom of things, look at our DAC comparison chart and use whatever tools you have at your disposal to determine why they are perceived to sound differently as has been independently verified via subjective listening. As you say, they should be light years ahead wrt fidelity so they should sound the same. Once you get to the bottom of that, then we can likely apply that new knowledge to examining transducers.
I would also say you are looking the transducer/dac-amp relationship backwards or incorrectly. The fact that 'fidelity' wrt headphones is more varied than w/ DACs/amps has nothing to do with their intended functions. The job of the transducer is to reproduce those tiny, miniscule differences occuring upstream. In essence, it is designed to be a looking glass (or magnifying glass depending on your phone/perspective) and let us hear how well that upstream gear is performing at reproducing dynamic and complex musical passages. That is their job. Scopes let us look at snap shots of very specific periods and samples of specific parameters. Equating one with the other, or discarding one set in favor of the other is simply incorrect. Good science does NOT ignore or disregard data, whatever it is. In the professional world you can do that, if you want to get fired, sued or go to jail. This practice is of course acceptable for internet warriors the world over however.
Just a reminder to those who have never realized this sites mission statement or may have forgotten. This site is not merely designed to regurgitate objective data. It is to correlate the subjective w/ the objective and explore the discrepancies that have yet to be resolved in that relationship. However, well articulated disagreement is welcome and rewarded with copious amounts of Grog. :money: :)p5 Ass talk regurgitated from certain other sites w/o the benefit of independent thought or analysis is severely punished. :spank: walk the plank2
-
You really should look at the total decay length in addition to distortion. Shorter, more even decay means more detail. Dominates over harmonic distortion, but not over ringing in my (not limitless) experience.
(Ringing is actually a long spike in decay, so it is related. Long even decay will smear over details, ringing typically also means emphasis in frequency response.)
HD6x0 has >2 ms decay below 2kHz, much longer in bass; its "smoothed minimum" is worse than HD800's. HD800 is almost evenly 1ms (max 1.5ms in bass), slightly decreasing in highs.
This is the objective measure of "it sounds faster". Though there are some people preferring slow bass - it tends to feel more hefty. That's what e.g. Audeze's do.