CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS
Lobby => Headphone Measurements => Topic started by: transparent201 on June 05, 2013, 02:23:46 PM
-
Hello to everyone,
I'm an audiophile/headphile who believes in objective evaluations(measurements) and that's the main reason i've been watching this forum(and some other relative sites) for quite some time now. My congratulations to purrin and the rest for their efforts in revealing the truth. Please forgive any grammar or spelling mistakes because english is not my mother tongue. So, let's get to the point now.
Up to now I haven't understand why some headphones sound much more transparent than others, despite that they measure almost equally good. For example hd 650 produces more vivid and detailed sound than e.g hd 595 even at low volumes where distortion is not a problem. Both cans measure relatively flat. Dt 990 are also reportedly less transparent than T1 although the last has worst CSD and weird FR. So what is/are the measureable factor(s) that determine perceived transparency/vividness/detail retrieval. Im talking about real transparency and sharpness not treble spikes.
Last and very important question. I have noticed HUGE differences in relative frequence response, between headphones, among various sites. For example at headphoneinfo.com dt 990 pro seems a little bassier than hd 650, while here(http://www.geocities.jp/ryumatsuba/dt990pro.html (http://www.geocities.jp/ryumatsuba/dt990pro.html)) the difference seems larger. Also, notice that in the second site hd 650 seems to boost 3 Khz while at headphoneinfo.com there are absolutely no peaks. I mean the whole picture is quite deceiving.
Can you explain please?
Thank you in advance.
-
Up to now I haven't understand why some headphones sound much more transparent than others, despite that they measure almost equally good. For example hd 650 produces more vivid and detailed sound than e.g hd 595 even at low volumes where distortion is not a problem. Both cans measure relatively flat. Dt 990 are also reportedly less transparent than T1 although the last has worst CSD and weird FR. So what is/are the measureable factor(s) that determine perceived transparency/vividness/detail retrieval. Im talking about real transparency and sharpness not treble spikes.
- One possible / contributing factor which was discussed last year: http://www.changstar.com/index.php/topic,593.0.html
- Also, don't discount distortion. Although the non-linear distortion tests here were done at 90db SPL/A, there is still distortion a lower normal listening volumes. The distortion tests do correlate to subjective experience, although how they do is less obvious than the CSDs. Overall high non-linear distortion, even at the 90db SPL/A tests tend to correlate with muddy sound, especially in the bass and lower mids. Think "low-fidelity" or "stock audio systems in cars." The SRH1440 and SRH1940 have high distortion, and it certainly sounds like it. Although two or three people on the other site feel those two headphones sound great.
Last and very important question. I have noticed [size=78%]HUGE[/size] differences in [size=78%]relative[/size] frequence response, between headphones, among various sites. For example at headphoneinfo.com dt 990 pro seems a little bassier than hd 650, while here(http://www.geocities.jp/ryumatsuba/dt990pro.html) the difference seems larger. Also, notice that in the second site hd 650 seems to boost 3 Khz while at headphoneinfo.com there are absolutely no peaks. I mean the whole picture is quite deceiving.
- This has been discussed at length before. Different measuring methods can yield different results.
- Bass response can be difficult to measure because of quality of seal. The recent MadDog measurement (done twice) was a good example of this. To my knowledge, even Tyll has had to futz with seal quality on the measurements. I actually have used two methods (couplers) which I use to specifically measure bass. One seals a little bit better than the other and results in higher measured bass.
- The HD650s are a bad example. There are two versions of it which do sound very different from each other. Also, there is strong suspicion on the part of a few guys here, including me, that pad wear will affect HD650/600 results around 3-4kHz drastically. Also, I know pad wear will also affect DT990 measurements.
-
Thank you. Very nice information.
You are right about hd 650. I have the silver silk version and doesn't sound slow or boomy at all. Just a little dull.
But that is not the problem I'm reffering to. Take for example dt 990 2005 edition and the latest 990 pro(which I own).
Headphoneinfo measurements show a significant difference in their FR. Dt 990 2005 edition has way more treble above 10 Khz than the pro model they measured in 2008(which happens to measure almost the same as in goldenears.net and sound as the model I own). In fact the difference is day and night. I tried the 2005 edition (on my xonar stx) and felt like being batman, able to hear supersonic waves. Sound was very clear but totally fake and with greater, but not irritating, sibilance.
What's strange is that many people consider them sonically similar. Exactly as the ryumatsuba, doctorhead.ru and your tests show. I don't think that all of this has to do with worn pads etc. Nor do I know of any modifications in the pro series lately.
-
I'm not sure if the 250 ohm measurement I have here is the Pro or the 2005 Edition (evidently both are 250ohms)
It should be noted that the 600 ohm version has more bass and less treble compared to the 250ohm version here.
Beyer's nomenclature is extremely confusing. I do not doubt that the Pro and 2005 Edition sound very different. That's Beyer.
-
There's many other factors that don't show up in our typical current staple of widely used metrics. One that comes to mind is wavefront propogation. It's no wonder that the most resolving phones on the market (HD800, 007/009, perfect specimen of LCD3) present a large and typically unified wavefront where timing to ear remains consistent. They are also quite a bit larger than typical dynamic drivers. With planars, the soundwave looks like a wall as it travels to your ear. Most dynamics throw a bulbous wave front out which is more problematic in headphones than speakers because of their immediate proximity to the ear, whereas speakers allow time for diffusion and use other tricks to maintain coherency (or make it worse). The 800 uses a ring radiator design to emulate a planar wavefront in a cone based dynamic design. The other factor is housing design and driver positioning (angle, distance, etc.) which assists in soundstage and imaging. The better a phone images, the better it allows the micro detail from a recording to come through unabated. Impulse response/driver speed is also an important factor as well but not the whole story as shown above. Obviously the faster the better, but I personally feel too much mass reduction for the sake of speed affects timbre (what you called 'vividness' but that's a whole other story. The other aspect is power delivery. This is an interesting one and segues to my next point. You can measure the same phone using two different amps that also measure similarly and get the same objective measures we typically rely on. However, listening to the same phone on two different amps can yield a boring, undynamic, unresolving presentation while the other amp simply grips the driver better and pushes it around like it's bitch causing everything to open up and sing. Yet the measurements we typically look at show both systems to be largely identical. The HD800 is a great example of this and I can tell you from vast experience w/ this phone and many, many amps, that high output impedance is the least likely explanation for the HD800 sounding fuller. The best amps I've heard w/ the HD800 run from 0-3ohms output Z, the worst from 0-150ohms. Same for some orthos like the HE500/6.
Of course, your question is even more perplexing if we apply it to amplification. DACs and amps are much more similar than different compared to headphones yet some are amazingly more transparent than others even though their measurements are much more similar than different relatively speaking. So that's another big mystery unless you either have not heard any differences between amps or have convinced yourself there are no differences that are audible based on a philosophical choice to adhere to a specific null.
Personally I think we are often only measuring a fraction of the metrics we should be and doing so in a static fashion unrelated to the dynamic system which listening to music represents. I have no doubt that their are well known speaker and phone makers out there who apply much more extensive, comprehensive and likely proprietary measurement methods for their purposes than what we normally see on the Internetz.
-
I'm not sure if the 250 ohm measurement I have here is the Pro or the 2005 Edition (evidently both are 250ohms)
It should be noted that the 600 ohm version has more bass and less treble compared to the 250ohm version here.
Beyer's nomenclature is extremely confusing. I do not doubt that the Pro and 2005 Edition sound very different. That's Beyer.
DT990 250 Pro.
Thank you. Very nice information.
You are right about hd 650. I have the silver silk version and doesn't sound slow or boomy at all. Just a little dull.
But that is not the problem I'm reffering to. Take for example dt 990 2005 edition and the latest 990 pro(which I own).
Headphoneinfo measurements show a significant difference in their FR. Dt 990 2005 edition has way more treble above 10 Khz than the pro model they measured in 2008(which happens to measure almost the same as in goldenears.net and sound as the model I own). In fact the difference is day and night. I tried the 2005 edition (on my xonar stx) and felt like being batman, able to hear supersonic waves. Sound was very clear but totally fake and with greater, but not irritating, sibilance.
What's strange is that many people consider them sonically similar. Exactly as the ryumatsuba, doctorhead.ru and your tests show. I don't think that all of this has to do with worn pads etc. Nor do I know of any modifications in the pro series lately.
To me the DT990 250 Pro sounded very bright, and my experience with them seem to correlate well with Purrin's measurements. I know about the measurements at Headphoneinfo and Goldenears.net.
As fars as Goldenears, they seem to apply quite a bit of smoothing to their FR plots (1/3 oct is quite a bit IMO.) Even so, the tremble is quite pronounced in their measurements. They somehow drop quite a bit from 10 kHz to 20 kHz which may or may not be due to the smoothing.
Headphoneinfo gave quite a bit of praise to the DT990 Pro. Note however that their measurements seem to roll of significatly from 10 kHz to 20 kHz. Also, their measurements seem to show more bass than tremble for the DT990 Pro. That is not what I heard.
Maybe there is Beyer variation here, but Beyer classic line seems more consistent so far.
-
Personally I think we are often only measuring a fraction of the metrics we should be and doing so in a static fashion unrelated to the dynamic system which listening to music represents.
I think a lot of the discrepancy comes down to the above statement. I have noticed in scientific inquiry a tendency towards missing the forest for the trees. It is important to recognize that although we can measure and understand certain aspects of sound as it relates to audio equipment, this does not preclude other important attributes from existing. I think it is safe to assume that the human ear is a much more complex system than we can fully understand at this time. In my mind any question of transparency has to be related to the ability of our ears to interpret sound. The question is not what aspect of headphones make them transparent. It is what aspect of the sound produced by headphones allows us to recognize transparency.
I am reminded of the talk from the engineer at ESS where he discussed how even though the Sabre chip measured perfectly in all aspects that they could test, it still did not sound right to trained ears. It was only upon further research that they were able to see a flaw in their measurement methodology, and to further understand how our ears are able to interpret sound. Even though I am not familiar with any of the research that has been alluded to, I am always hesitant to accept the types of "truths" that have been mentioned. This is the equivalent to ESS saying our measurements are perfect and anyone who hears a difference is wrong. Instead they pursued the obvious line of inquiry...WHY don't our measurements match up with what our test subjects hear? In doing so they were able to contribute something far more valuable to the development of audio reproduction. This is the essence of what scientific inquiry is and I believe is the correct path in further developing our measurement devices and testing methodologies.
-
sorry for OT but Anax what do you not like about the B22? I was really close to building one because of never hearing a single bad thing about it. (this thread just seemed dead enough already. delete if not cool!)
-
sorry for OT but Anax what do you not like about the B22? I was really close to building one because of never hearing a single bad thing about it. (this thread just seemed dead enough already. delete if not cool!)
Oh we had quite the chatbox going on a few times full of people not in love w/ the esteemed B22. The consensus was that it was good for its time based on the offering available.
For me, it warms up the HD800 too much and simply fails to have the crystal clarity, precise imaging, uber resolution, punchy dynamics and sharp transients of other competitors. It's one of those amps that gets recommended as being the ultimate HD800 amp simple because it's a bright phone that gets warmed over by a dull/ warm amp. Not unlike a Luxman p1u or a PoS Leben whose syrupy signature is a disgrace to tube amps. B22 is fine for what it is, it's not a bad amp, but not the last word in anything. A job of an amp is to be a transparent window and get out of the way, not to darken it nor be bright, thin and harsh like a Benchmark or thin and bass light like the CEntrance amps are w/ the HD800. Both sides represent extremes whereby the amps in question are coloring the phone.
What I find humorous, is that pseudo objectivists will say the HD800 sounds bass light out of the Benchmark (and many SS amps) because that's how it sounds in reality. Yet these same folks fail to explain or realize that there is not one, NOT ONE single measurement of the HD800 indicating that it should sound bass light. So somehow they blame the headphone rather than the amp when their own evidence is even working against them!! I love it. :))
-
1) If measurements cannot capture all the [size=78%]important[/size] sound properties, how the hell all these high end experts(e.g Woo audio, ray samuels) design their machinery and manage to achieve higher quality? If you want to design you must measure something to be sure everything is ok. Because if you like distortion your clients might disagree with you.
And if you design by ear and there is no science at all why do you charge enormous amounts to your clients. For "playing" with circuits?
I don't know about RSA or Woo, but I know Craig Uthus of Eddie Current and Jason of Schiit rely on both measurements and subjective listening. Craig and Jason are both really anal on the math (theory) and measurements (beyond FR and delving into electrical properties) of the components, even before they are assembled into a final design. In the end, these two guys use their ears (since not all things are measurable.) But Craig's been a around for a while, working for UREI and then JBL. He's very old school. Jason been the high-end industry for a while too - with Sumo, etc.
I am more than willing to accept that there are important sound quality parameters that we haven't been able to quantify yet. From this sentence two questions arise immediately.
Second, if multinationals with billions of dollars for RnD don't know these secrets of high end audio how am I supposed to believe that some small "elite" manufacturers do and keep it secret for so many years. Are they some kind of crazy scientists light years ahead of the others?Last, and most obvious observation. Let's say that "X" amplifier has unlocked the mysteries of human hearing and sounds best(because of "no feedback" for example). Then someone would have copied it or studied its properties(if it was patented) and further improve it.
One thing which is critical for us to understand is that all audio circuits have pretty much already been invented. Generations ago. One generation for digital. Two generations for transistors. And three or four generations for tubes. A lot of stuff was patented in the past! (The circlotron / CrossFET thing in the Mjolnir amp - that circuit was actually patented.) Good thing they weren't such assholes about patents back then.
The state of audio equipment now is more a matter of execution and the specifics of implementation of a particular design. Also, you forget that large multi-nationals are not necessarily in the business of making great stuff (unless you think GM makes great stuff - IMO, they make good/OK stuff at good/affordable prices.) They are in the business of making money. Why do you think the Corvette still uses a push-rod motor, leaf springs (WTF), a front-engine, and POS LED panels from 1984? What GM has done is taken old-technology parts (cheap) and tweaked it (this kind of R&D is cheaper for large corporations) to make it work well - very well.
Do you think Monsanto is in the business of making great tasting vegetables?
Do you think Bose (arguably one of the most successful audio companies) makes great sounding stuff, despite all their patents?
This is not to say that all boutique manufacturers make great stuff. LOL, I think most don't.
-
^ Not to mention Mike Moffat who is one of the fathers of the DAC.
-
That patent stuff really hurts, especially when its already been around and then some guy decides to patent it and everyone else has to find workarounds. These days you get sued for selling patent leather.
Sucks when guys come in here with an agenda. I'd love to see this site be agenda neutral.