Analysis of Head-FI HQ's Sony MDR-Z1R Measurements and Tech Talk

Discussion in 'Measurement Techniques Discussion' started by purr1n, Jun 19, 2017.

  1. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,623
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    OK, I've had some time to sleep and process a lot of technical stuff.
    1. ON FREQUENCY SHIFTS OF PEAKS
    2. FURTHER ANALYSIS OF HEAD-FI HQ MEASUREMENTS
    3. MISUNDERSTANDING MEASUREMENTS
    4. OTHER
    5. FALSE NARRATIVE AND COMPARING MEASUREMENTS (HOW NOT TO)

    ON FREQUENCY SHIFTS OF PEAKS
    I skimmed Jude's subsequent post on his measurement apparatus and it appears that it does simulate an ear canal. I would postulate that this canal might be responsible for the shift of the 10kHz peak to 8-9kHz. I've seen IEM couplers (tubes of differing lengths to simulate a canal) exhibiting similar peak shifting behavior.

    As you mentioned, the fact that my anechoic-like foam coupler measurement for the Z1R exhibits a 10kHz CSD ridge (see below) means that it's definitely there.

    [​IMG]
    Whether this peak shifts in frequency by the time the sound hits the eardrum is inconsequential if we are trying to obtain measurements which are the perceptual equivalent of speakers*, which is what I am trying to do. Speaker measurements are taken with a microphone after all, and if a 10kHz signal at the microphone appears as a 8kHz signal at the eardrum, it's still an 10kHz signal for all affects and purposes for sound engineering. Still, more exploration is warranted.

    *Note that this does not make Jude's measurement rig inherently wrong. It's just a different framework, which requires study of his pool of measurements and how they relate to our perception of sound.
     
    Last edited: Sep 28, 2017
  2. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,623
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    FURTHER ANALYSIS OF HEAD-FI HQ MEASUREMENTS
    CSDs and NINJA PEAKS

    Here are my MDR-Z1R measurements and Jude's. I've taken the liberty of cherry picking similar looking plots given the variability of the Z1R headphone. One of the reasons I like CSDs is that they sometimes show us peaks that don't seem exist in the frequency response plots. These peaks manifest as ridges in CSDs that evolve right next to dips in the frequency response plot. Yes, these are sneaky ninja peaks. They hide! It happens in headphone measurements with regularity because of internal standing waves inside the headphone cup causing a null at certain spots. It actually happens with speaker measurements too!

    Here I've numbered the peaks in both measurements. I would postulate that the sharp 10kHz is Tylls' frequency response measurements manifests as the dip number #5. Just because Jude's measurement doesn't show this peak / ridge (ringing), doesn't mean it's not there. Heck, my frequency response measurement of the Z1R doesn't show it well either.
    Sony MDR-Z1R CSD R.jpg
    FR-Sony-MDR-Z1R-Serial-0323-and-Serial-X009_LABELED.jpg

    Jude says that the 10kHz peak is "something that he has not heard (and I would definitely hear that)." But let's assume that I am wrong about the frequency shift, or even wrong about the null-to-ridge behavior. Then why does Jude not hear the smaller peaks at #3 and #4, 6.5kHz and 8kHz respectively? Granted Jude may have been truthful in his statement that he couldn't hear the 10kHz peak because his plot didn't show one. But I find it's the things that he doesn't say more amusing.

    Anyway, I firmly believe that the peak is there. Whether we want to call it 10kHz in Tyll's frequency response plot, or 10kHz CSD ridge in my CSDs, or 9kHz dip in Jude's measurements, it's there. If you ask me, personally, I think Jude's deaf. Well let me be politically correct. Let's say he's not as attuned to small deviations from non-linear as most of the folks here. At one time when I was deeper into speaker building, I trained my ears to hear small narrow peaks of 1db. Even today, I can hear deviations of 0.25db for broad peaks.
     
    Last edited: Jun 20, 2017
  3. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,623
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    MISUNDERSTANDING MEASUREMENTS

    Wow, coming from Herb, a writer for Stereophile, I don't know what to say. This statement is a fundamental misunderstanding of measurements. Measurements DO NOT dictate whether a person will like a how a piece of gear sounds or not. What measurements do is show us non-linearities.*


    While I don't disagree that there are "measurement persons" or "objectivists" who refuse to acknowledge that measurements do not tell the whole story, Herb's stance seems a little bit too tight, too rigid in categorizing people who feel that measurements are useful. I may well as respond back to him in a similar fashion: "non-measurement persons" are too lazy to understand what they mean. You get the idea. Let's not go there. I'm short-tempered with people who are intellectually lazy.

    *Many of you know that I work in the entertainment industry. Why are minimizing non-linearities important? This is funny because I just had a conversation with @LFF about the decaying art of sound engineering for music, but how the good sound guys can still be found in TV and film work. Ever wonder why your experience at the movies is so consistent? From film to film. From theater to theater? It's because of measurements, and sound guys with good ears, that Darth Vader always sounds like Darth Vader, and not Kermit the frog.

    They don't use Sony MDR-Z1Rs to mix and master the sound. They use stuff like this: http://www.superbestaudiofriends.or...rop-jbl-lsr30x-5-active-studio-monitors.4570/


    Does this mean the MDR-Z1R suck? Hell no! That's really up to you, regardless of measurements. However, the entirety of the MDR-Z1R measurements, from me, Tyll, and Jude - they do paint a consistent picture of how it sounds despite differences among those measurements.
     
    Last edited: Sep 28, 2017
  4. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,623
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    OTHER
    First, this depends upon who's flat plate coupler? My $99 flat plate coupler gets results more similar to @Serious' head coupler than his plate coupler. And as you noticed before, also gets frequency response results eerily similar to Jude's KEMAR.

    The main areas of concern with accuracy is the area centered around 3kHz (1kHz to 5kHz) where the pinna interact with the inside of the headphone cups. Different types of cup designs will result in different responses that the flat plate coupler will not capture.

    But to answer your question, EQing a headphone on my flat plate coupler will result in a headphone that is too bright. The target curve for my flat plate coupler would be something simliar to the B&K curve. Flat out to three hundred Hz, following with a slow downward slope 6-8db down by the time we hit 20kHz. In addition, a slight depression at 3kHz of a few db might be warranted depending upon headphone.
     
    Last edited: Jun 19, 2017
  5. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    To avoid generalization (because I know full well there are those who offer EQ and DSP solutions to measured headphones), I think it's safer for me to say that my experience with EQ based on measurements have been less fruitful compared to acoustically modifying the FR. This applies not just to flat-plate measurements but also to measurements that take into account the ear canal. But I find the experience even worse with some kind of ear simulation due to the HRTF involved. (more on this later)

    And that's probably just me, but I have some theories regarding this.

    1. A headphone's FR looks a certain way due to conscious acoustic tuning decisions made by the engineers behind it. Admittedly, I started my headphone modification journey arrogantly thinking that some engineers have no clue. And after all these years, my conclusion is: I am wrong. Perhaps this is true of the "lazy" designs like those airplane headphones that we used to get back around a decade ago or so. But in the context of headphones, speakers, earbuds, earphones, etc... targeted toward the audiophile market, a lot of considerations go into these products for sure, and there is a certain "house sound" that every manufacturer strives to achieve. Not because they cannot make a headphone with "flat" FR, but because if everyone goes for the same thing, there would be no differentiation anymore. But ironically, in avoiding such a thing, all of the manufacturers in the world have also prevented a "reference house sound" that has flat FR.

    2. A headphone's FR is actually not the sole factor that determines its sound. Distortion, CSD, spectrogram, etc... play into that as well, but I do believe there are certain other factors that we have yet to figure out how to measure. No, it's not voodoo or magic or unicorn tears, or pixie dusts, but @Marvey said it best: headphone measurement is still in its infancy. It will probably take years before we master this technique. On that note, because it is in such a state, I believe the methods we are using, be it flat-plate or with ear simulation, will probably have a lot of improvements coming up, and also that none of the methods used thus far are truly more accurate than the others at anything. Case in point: we do not need to be put in anechoic chambers in an isolated environment (space??) in order to hear and recognize certain qualities to different headphones that do not show up in measurements.

    3. In combination, the "house sound" that I mentioned in #1 with the unmeasured factors in #2 together make a "flat" FR sometimes inappropriate for certain headphones. I think it is pretty obvious to see why: if you just EQ it, the factors that make up the "house sound" that have not been measured may be excited past a certain point that you cannot predict. This will cause certain headphones to sound very "off." Some actually will react quite well to EQ this way, but in general, I have not found more than a handful of headphones that will sound "decent" using this method. In fact, acoustically modifying a headphone can only get so far, because the effect of the enclosure will persist either way. This is why some have opted to take the drivers and put them into new enclosures. That helps but then you'll still have a part of the "house sound" due to the drivers. It's a never-ending cycle.

    4. We are completely at the mercy of the manufacturers' tuning whims here. No doubt. This is why whenever the good folks at Massdrop request my assistance in determining whether or not something sounds good to me, I am very very happy to help. It's also why I try to talk to manufacturers whenever I go to trade shows. Sometimes a manufacturer has already gotten "so close" to that sound that I want, and so a push in certain directions may actually make their products better for me. And seriously, manufacturers' tweaks are almost always better than what we can come up with in the DIY community. There may be some exceptions, of course, when the manufacturer gets lazy, but otherwise, they would always be capable of pulling this off. It's only when a manufacturer is stubborn about wanting to change their "house sound" (like in the case of Sony or Sennheiser) that I think the "last resort" is to basically try and perfect the tuning yourself somehow. Acoustic modifications, EQ, amps, DACs, etc... seriously, I think all of those are very much equal in effect and will all contribute toward the one goal of making the headphone more listenable to you (or me), and that's all that matters.

    5. Which brings me to this one thing: I think "listenable" does not mean "flat FR" to different people. Some may like it "flat", some may not. I know I do like certain headphones better when they are "flat"-ish, but not all of them like that. My friends who know me well and understand my preference know full well what to expect from me when I post new modifications to some things, and then they adjust accordingly. I also make sure to note these things all the time because I realize that what works for me won't work for everybody else. It's a moving target. With that said, my conclusion is that: just do whatever works, be it EQ, or a new amp, or a new DAC, or hell, new cable, and don't worry about measurements too much. Take measurements if you can, but use them only as a "point of reference" to see if and when you go and measure something else and compare it to your "reference", if you'll find that something else sounding like how its measurement would suggest. If not, then you may have to make more considerations in order to generalize the results.

    6. To touch upon HRTF and ear simulation that I briefly mentioned above: to me, it complicates the measurement process, and the HRTF becomes that one extra thing that I need to tweak in order to make the final measurement more representative of what I hear. That's on top of the compensation curve that's applied after HRTF as well, and sometimes doing HRTF takes away certain things that I think should be there because I do hear them. Sure, sometimes HRTF is closer to what I hear, and then sometimes a flat-plate coupler is actually closer to what I hear. Perhaps a common ground is the perfect solution, but honestly, the two methods are so different that I don't know what a common ground would even look like. I'll leave it up to the more experienced engineers to figure it out, but as of currently, I'd use different methods for different use cases. Flat-plate still gives me better hit rate compared to ear simulation when it comes to full-size open-back headphones. For closed-back headphones and speakers, though, I have found ear simulation to be the superior method, but... not all the time. Overall hit rate still wee bit higher with flat-plate, and flat-plate is also simpler for me to consider to. Just gotta deal with the compensation curve.

    TL;DR: I typically like a flat FR myself, but I'd prefer it when it's due to acoustic tuning, and not due to EQ.
     
    Last edited: Jun 19, 2017
  6. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,959
    Trophy Points:
    113
    Location:
    Irvine CA
    Indeed. I tried that before. And I remember you recommended targeting the B&K curve. It indeed yielded better results IMO.

    @Lurker, here is a thread with some discussion about it:
    http://www.superbestaudiofriends.org/index.php?threads/headphone-fir-equalization-using-foobar.2772/

    Finding the absolute frequency response magnitude flat line in a headphone measurement system is IMO difficult. I kept my POS setup as it is for consistency. But I'm sure there are ways to improve it in the absolute sense. I use the measurements more on a relative sense though.
     
  7. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,589
    Trophy Points:
    113
    Location:
    near Munich, Germany
    @Marvey it would seem odd to me if a 10kHz peak gets shifted down to 8kHz somehow due to the ear canal, at least if the 10kHz peak is actually an inherent driver resonance. If it's related to earcup acoustics (which I bet it isn't), then it could make sense. I still think it's just that the dummy head has some sort of inherent dip around 9-10kHz that the Z1R fills in perfectly (see Jude's Utopia and Ether measurements).

    Is that a coupler I don't know about :p ?
    Anyway, I wouldn't necessarily say that. It depends on the headphones. It might work with the Z1R and maybe also other headphones like the HD600 or the Utopia, but it certainly doesn't work for headphones like the HD800 and HE1000. At least with the couplers I've seen and measured. Some headphones are made to use the ear-gain to get to a certain target response and some (like the HD6X0 series) use less of it, so they will always measure very flat.* Or does your coupler show a flat response from 100Hz to 5kHz for a modded HD800? I definitely haven't seen one that did, but I do think the HD800 can be modded to be more neutral than the HD600.
    Essentially what you said about the pinna gain. This is a very important region. Saying that the coupler is mostly accurate, except it's not super accurate around 1-5kHz essentially means that it's not too accurate after all. As it stands I like my way of using two totally different coupling rigs to give me a better picture of the overall sound, but then again I don't really measure anymore and I've almost completely sworn off headphones anyway. Maybe at one point I'll try to improve my coupler. I do know that my material gave me CSD results closer to the skin on my head than a dead chicken did with the HD800. The dead chicken was similarly ringy as the creatology foam, but my skin wasn't for some reason.

    *I actually think this is at least part of what @Bill-P talks about in 3 and 5.

    Again, depends on the headphone and the coupler. The results can range from pretty good to horrible. Generally bright and with too much energy around 3-4kHz. I posted my target curves for my coupler and a few different headphones somewhere. I think they're in the "2016 NY area measurement" thread somewhere and maybe also in the thread ultrabike linked. For my coupler I like to see flat to 500Hz with around a 5db downward slope to 2kHz, then a dip that varies with headphones and up to the -5db level for 5kHz and up.

    This is the main reason why I use the in-ear measurements. No compensation is needed for the most important part of the midrange. I use no compensation up to 4kHz. I lower a 5kHz peak that often shows up in my measurements a bit and compensate for an overly rolled off treble, but the rest is left flat. I think Hands actually doesn't use any compensation at all, except for an inherent bass rolloff with his measurement setup.
    Speakers as measured with the in-ear method all measure way off, even if they measure the same with an omni mic. Some people think it's a good idea to use a cardioid mic to measure speakers in-room, but I haven't tried. I generally wouldn't worry too much since speakers can still sound right, even if they sound different. I know it doesn't seem to make sense.

    The headphone measurements don't look similar to the speaker measurements with the in-ear method at all. I think it has something to do with how close the microphone is to the drivers in the headphones, where a lot of the signal actually bypasses the pinna and hits the microphone directly or something. Headphones that sound flat will measure flat, not with a 3-4kHz peak like in the dummy head measurements.
     
    Last edited: Jun 19, 2017
  8. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,623
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Haha, I prefer to think that it works for everything except the HD800 and HE1000! Do you happen to have on head HE1000 measurements BTW? Just curious. Trying to decide if I want to spend SBAF funds on a head simulator. Might not be worth the trouble since most people don't seem to care about the art of measurements.
     
  9. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,589
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I only got one channel at the meet. The Code-X actually didn't measure too well, either. Generally planars with their huge drivers seem to interact more with the ear than dynamic driver phones, with the exception of the HD800.
    Here are the HE1000 measurements: http://www.superbestaudiofriends.or...area-measurements-by-serious.2518/#post-67058
    Note that these are uncompensated measurements which show more 5kHz peak (about 3db) and less >10kHz (about 5-10db).

    Head simulators are serious money. Might be a better idea to try to DIY something, but getting the materials right is probably not easy.
     
  10. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,623
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I dunno, the Code-X, even with 5kHz compensation you recommend, doesn't reflect anything at all what I just heard with @brencho's pair. He was using thicker pads though; but such thick pads won't increase 5kHz that much.
     
    Last edited: Jun 19, 2017
  11. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,589
    Trophy Points:
    113
    Location:
    near Munich, Germany
    It could be that the pair was weird. Audeze Vegan pads. It did sound like that to me. From distant memory it was too bright for me. Brighter than my HD800 actually, along with a 2kHz dip that annoyed me. Not quite what I expected. It sounded U-shaped in a way, with warmth to the lower midrange, but also brightness and an etch to the upper midrange and treble. (I listened to them in @Maxx134's home some time after the first meet.)

    EDIT: I just saw your Code-X measurements and I wouldn't say that they're inconsistent with them, especially considering that there were probably significant differences between the two pairs. My coupler measurements also looked a bit different, but I think that this is also because the pads (and possibly damping) were different. Looking at my coupler measurements I think what happened is that the pads decreased the 2kHz region a whole lot compared to the right pads, which is why they sounded U-shaped to me. My coupler should actually show more 2-4kHz than the creatology foam coupler or at least it does with the HD600 and HD800.
     
    Last edited: Jun 20, 2017
  12. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,623
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    FALSE NARRATIVE AND COMPARING MEASUREMENTS (HOW NOT TO)

    https://www.head-fi.org/f/threads/t...e-from-ifa-2016.818846/page-691#post-13547603

    Tyll's distortion measurements of the Sony MDR-Z1R shows distortion around 1-2% in the area under 80Hz. Jude in his thread presents his own distortion plot seen here.

    [​IMG]
    Jude actually calls out Tyll directly by quoting Tyll:

    ...Distortion is fairly low, but bass distortion looks problematic. I suspect this may be excursion limiting due to the small magnet? Who knows...

    Jude's then points out that his measurements show half the distortion that Tyll's plot does. After a few more words, he goes on to conclude that perhaps Tyll has an outlier Sony MDR-Z1R (because Jude's measurements are better, and "we" don't hear the same thing that Tyll does).

    In this case, Jude has commited the cardinal sin (at least with headphone measurements, and distortion measurements at that) of comparing across different measurement systems. It's likely Jude got lower distortion numbers than Tyll, not because Tyll had an outlier sample, but because Jude's measurement system has lower inherent noise and distortion.

    To get better sense of how the MDR-Z1R compares, Jude should have asked readers to compare to his other data. The problem with this is that Jude does not offer an easy way for people to compare to his own other data, whatever little of it exists publicly. Jude should have known better if he wants to fancy himself as a measurement authority.

    But here, allow me to provide comparative distortion measurements using Jude's own plots. This took a little bit of effort in Photoshop to align the plots and scale the Y-axis to produce a meaningful visualization.

    These plots compares the THD of three headphones: The Z1R (two samples), the Elear, and the Utopia.
    [​IMG]
    TOP SET OF RED/BLUE: MDR-Z1R
    BROWN: ELEAR
    BOTTOM BLUE: UTOPIA

    Now, we still can't say the Z1R's distortion is bad. But we can say it's worse than the Elear's and Utopia's. Everything is relative. It's where you want to set the bar.
     
    Last edited: Jun 20, 2017
  13. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    I want to set the bar right here. :p

    [​IMG]

    Haha, kidding aside, I think this is reasonable to assume from "mapping" this Utopia distortion plot to distortion plots measured on other rigs at about the same SPL.

    Past that line, things get crazy.

    Subjectively, I do hear Utopia as being a bit blurry/soft/indistinct in the bass compared to the rest of its range, which is quite excellent. So to me, anything above that "bar" per se will most likely mean something is "wrong."

    And I did hear this with the Z1R as well: excellent midrange clarity and actually decent speed, but bass so "slow" and "boomy" that it's not worth it at all. Also interesting to note that relatively, Z1R distortion in the bass compared to Utopia is a factor of 500% in this plot! Z1R has 5 times more distortion than Utopia as measured here, and damn, Utopia is already borderlining on "too soft" in the bass for me.

    That's not to mention treble distortion in Z1R in this plot is "only" 10,000% (10 TIMES!!!!) more so than Utopia's treble distortion at 5-6KHz. At 10KHz, it's "only" 33,333% (33 TIMES)!!

    "Low distortion" is fine and all, but... "low compared to what" is the question, I think. Definitely not "low" at all in comparison to Utopia.
     
    Last edited: Jun 20, 2017
  14. Ringingears

    Ringingears Honorary BFF

    Pyrate
    Joined:
    Sep 26, 2015
    Likes Received:
    3,659
    Trophy Points:
    113
    Location:
    Northern Californium Valley
  15. SSL

    SSL Friend

    Pyrate
    Joined:
    Nov 12, 2015
    Likes Received:
    1,239
    Trophy Points:
    93
  16. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,623
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    The more likely explanations are follows: 1) ear training, of lack thereof; 2) sufficient regular exposure to extremely accurate transducers, loudspeakers; 3) and lack of direct comparison to better treble behaved headphones.
     
  17. Ringingears

    Ringingears Honorary BFF

    Pyrate
    Joined:
    Sep 26, 2015
    Likes Received:
    3,659
    Trophy Points:
    113
    Location:
    Northern Californium Valley
    And perhaps 4) At age 47, listening to lots of headphones ( he has almost all in existence ) at unhealthy volumes perhaps some high frequency hearing loss . Or explanations 1 through 3 and 4.

    Or.......$$$$? I have heard it can influence one's hearing amongst over things.

    Oh and 5. He his gotten too big for his britches.

    I don't know Jude other than a personal conversation at a meet. Just feel he is hurting the community.
     
    Last edited: Jun 20, 2017
  18. Thad E Ginathom

    Thad E Ginathom Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    14,090
    Trophy Points:
    113
    Location:
    India
    Whilst there are whole things that drop off the edge, like mosquitoes more than two inches away, baby birds outside the window, the kitchen timer.... hearing loss is something one doesn't actually notice, because mostly the world sounds the same as it did yesterday, the day before, and as we think it did last week or last month.

    But for those who have not worked in a noisy environment (which very much includes music), 47, according to me, is young. I wish!
     
  19. JimL

    JimL Tongues KG's hairy starfish for fun

    Joined:
    Nov 21, 2016
    Likes Received:
    191
    Trophy Points:
    33
    Location:
    Albuquerque, NM
    I'm not convinced that the difference is due to Tyll's measuring system having higher inherent noise/distortion. If that were the case, I would expect that all his low frequency THD measurements would be around 1% or so, but in fact, there are several headphones that he measured that run around 0.3% THD in the 20-100 Hz range (e.g. LCD4, Stax SR007and 009, Mr. Speakers Aeon pre-production, ATH MRS7, even the defective Focal Utopias).

    Incidentally, has anyone noticed that Jude's underlying assumption that his measurements are more reliable because he has all this expensive equipment is a bit like a 20 year old guy with the new Ferrari assuming he can out-drive anybody in a lesser car.
     
    Last edited: Jun 21, 2017
  20. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    Actually, I have a theory on this that perhaps it doesn't matter what the distortion level is (%-wise at least)... but it just matters whether or not the distortion graph looks relatively flat/constant. Basically, a headphone with a distortion graph that closely follows its FR in form or is perfectly flat would probably sound "good" or at least "better" than something that measures relatively flat but distortion is all over the place, and perhaps the reverse case (bad FR but flat distortion) would be bad as well. I have noticed that at least the former (good FR, bad distortion) is true when it comes to higher frequencies.

    And to me, sometimes a high bass distortion is not very indicative of "quality" per se (Utopia and HD800 come to mind), though it does seem to affect what I'd call "bass imaging" from past experience. Yeah, maybe I'm just spewing BS (like does bass actually have "imaging" per se? Like can you even tell in a track if bass is supposed to sound like it's coming from a certain direction and attached to a certain instrument?) here, but it's a theory.

    This is also how I have been looking at Tyll's distortion plots as of recently: Look at things comparatively to themselves and/or to other headphones that I know well, rather than judge the measurements as being absolute. That's why I did the "analysis" thing above with regards to Jude's measurements. It was meant to be a gag more so than anything, but ironically, it also highlighted the problems inherently with the way Jude is trying to present the Z1R here.
     

Share This Page