HD800 and HD800S distortion analysis

Discussion in 'Headphone Measurements' started by purr1n, Jan 17, 2018.

  1. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    I have a stock HD800 coming in. For now, I am using a modded HD800 (V version Bill-p) modded. This HD800V has a little bit less bass than the stock HD800 at the same voltage level, so distortion for the HD800V might be a little bit higher than stock (I drove the HD800V at higher levels to get the same volume at 39Hz). I'll redo these when the stock HD800 comes in.

    I ran these at two different levels, to so how the distortion characteristics changed. FWIW the HD800S has higher second order compared to the modded HD800. The modded HD800 has lower second order but higher third (and higher) order. Most people who have heard these headphones know that the HD800S does have a different sound in the lower registers compared to the HD800.

    hd800 1.gif

    HD800 2.gif
     
    Lyer25, ultrabike, Halaudio and 5 others like this.
  2. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    So I am going to continue on with this thread because evidently this is the one that peeved Jude off into demonstrating how his measurements can get a 45db lower noise floor than mine, which in the case of transducer distortion measurements, does not matter.

    This time around I took the same type of measurements above with the MiniDSP EARS. Again, this is with modded HD800s. The first set of results is cranked up to a high level, maybe over 100db. Here we see that the distortion characteristics are very similar to each other between HD800 and HD800S. The HD800S maybe has just a little bit less distortion, but the difference is not significant.

    HD800 modded
    HD800V 20.png

    HD800S
    HD800S 20.png
     
    ultrabike likes this.
  3. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    Let's move down to more reasonable volume level. Again, this is on the EARS

    HD800 modded
    HD800V 32.png

    HD800S
    HD800S 32.png

    The HD800S appears to have quite a bit more second order but less four order distortion compared to the modded HD800.
     
    Maxx134, ultrabike, Halaudio and 2 others like this.
  4. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    Now going back to the old flat coupler measurements, here are D2, D3, D4 broken down. Note that overall THD distortion is close (THD is not shown, but one can eyeball THD knowing the others), but the characteristics are different. Note that the higher third order with the HD800 (even though modded) is similar to past measurements.

    HD800 modded
    HD800V.png

    HD800S
    HD800S.png
     
  5. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    Now let's do the same on the EARS coupler. At these levels (same as the flat coupler above - I didn't calibrate the vertical axis for the EARs), the HD800S does appear to have higher second order distortion in the bass.

    HD800 modded
    HD800V EARS distortion.png

    HD800S
    HD800S EARS distortion.png
     
    Serious, ultrabike, MrTeaRex and 3 others like this.
  6. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    At the end of the day, no one cares about the measurements per se. The premise behind the distortion measurements from @Tyll Hertsens (and now me) was to explore why the bass of the HD800S sounds slightly different (warmer) than the HD800 - at least to my ears, @Tyll Hertsens ears, and maybe @sorrodje's ears. It's all about exploring why and how - finding relationships between what we hear and what is measured.

    I don't know what Jude is trying to prove with this 16M FFT points and artificially low noise floor from the FFT process gain.
     
  7. sorrodje

    sorrodje Carla Bruni's other lover - Friend

    Staff Member Friend MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    2,778
    Dislikes Received:
    0
    Trophy Points:
    113
    Location:
    Dijon / Burgundy / France / EU
    This ^ ..

    I compared HD800 stock and SDR modded against my HD800S and difference in bass area was small to my ears. Very small. and I think the small difference we can hear is more related to the less masking effect induced by the lifted 6 khz peak than to anything else. This or maybe a slight modification of the grey mesh which "close" the cup. I forgot to explore that option when I had HD800 and HD800S side by side.

    I still stick to my opinion expressed in a wall of text on HEad-fi two years ago.
     
  8. Merrick

    Merrick A lidless ear

    Friend
    Joined:
    Jan 6, 2016
    Likes Received:
    5,063
    Dislikes Received:
    5
    Trophy Points:
    113
    Location:
    Portland, OR
    I still think he's trying to position himself as the only reliable authority on measurements, both as a way of undercutting competitors, and also to bolster the numbers of sales he can claim to provide to his sponsors. We know he doesn't do well with dissenting opinions, especially if he thinks such opinions will cost him money. You and Tyll have been using measurements to complement your impressions for years, and I'm sure Jude felt like he needed to compete. But for him, competition means destroying all others. It doesn't matter if you and Tyll aren't actually trying to compete with him, he feels the need to compete with you and Tyll.

    I don't know how else to explain his multiple appeals to authority when asserting the superiority of his measurement system over yours and Tyll's--oops, I mean over DIY enthusiasts. He doesn't want anything but his megabuck system to even be considered legitimate.
     
    FlySweep likes this.
  9. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    Different rigs, headphones, operators, time of day = different results. Jude's results are close - too close to call. I don't think he's massaging the results; but I also don't have trust in his results because of the following: 1) he has no consistent history, e.g. visualizations are not consistent, axis and other parameters are all over the place; 2) he doesn't totally understand what he is doing, e.g. not understanding FFT process gain, or that nulls are sometimes peaks in disguise; 3) he makes false associations, e.g. a low noise floor equates to accuracy with gross error distortion from transducers; 4) cherry picks assertions from others (Tyll, and now myself) and uses his measurement gear to prove himself "right" - always in favor of a manufacturer, e.g. no increased distortion, no peak, etc.*

    Measurements won't save HF or make it better; and what makes SBAF work is not measurements: it's friendships.

    * LOL, when was the last time Jude identified irregular measurement behavior that Tyll or I did not catch? Hahaha. Surely with his super high-end equipment, he would have found at least something?
     
    Last edited: Jan 18, 2018
    TazerMonkey, Cos, Vtory and 3 others like this.
  10. james444

    james444 Mad IEM modding wizard level 99

    Friend IEMW
    Joined:
    Jun 3, 2016
    Likes Received:
    1,088
    Dislikes Received:
    0
    Trophy Points:
    93
    Location:
    Vienna, Austria
    Thanks for your efforts! I'm looking forward to your measurements of the stock HD800 vs. HD800S.

    Meanwhile, I'm wondering why there's a noticeable difference at lower volume level, but not at high level?
     
  11. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    The high level measurements I think might actually be around 110db! I haven't calibrated my EARS, so no idea on exact SPL. The very high SPL 39Hz tone might be pushing the driver so hard to clipping that it doesn't near matter. Clipping as in squaring of the sine wave, which does push up odd harmonics - hence the HD800S, which favored the even harmonics at lower levels, exhibited behavior that looked more similar the HD800 which already emphasized the third harmonic.
     
  12. james444

    james444 Mad IEM modding wizard level 99

    Friend IEMW
    Joined:
    Jun 3, 2016
    Likes Received:
    1,088
    Dislikes Received:
    0
    Trophy Points:
    93
    Location:
    Vienna, Austria
    I see, thanks! So, as long as you're below clipping level, the difference should stay pretty much consistent, regardless of whether you're down by -10dB, -15dB or -20dB?

    Also: what's the output impedance of your amp, and have you done any research on how OI might affect distortion?
     
  13. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    The amp is a Magni 2. Stock HD800 incoming, so more data this weekend. I'll try it on the EARS, V1, and V2 couplers, at different levels.
     
    james444 likes this.
  14. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    OK. Here are the stock HD800 vs HD800S results on the EARS. I did two different frequencies, 39 and 78Hz, at two different SPLs, ~90 and ~100db.

    39Hz at ~100db RED = HD800S YELLOW = HD800
    HD800S v HD800 39Hz 100db.png

    78Hz at ~100db RED = HD800S YELLOW = HD800
    HD800S v HD800 78Hz 100db.png
     
    Lyer25, Serious, Priidik and 9 others like this.
  15. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    39Hz at ~90db RED = HD800S YELLOW = HD800
    HD800S vs HD800 39Hz 90db.png

    78Hz at ~90db RED = HD800S YELLOW = HD800
    HD800S vs HD800 78Hz 90db.png
     
    Lyer25, Serious, Garns and 6 others like this.
  16. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    Let's add one more for good measure at around 80db.

    HD800S vs HD800 78Hz 80db.png
     
    Serious, Garns, Klasse and 1 other person like this.
  17. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    46,281
    Dislikes Received:
    72
    Trophy Points:
    113
    Location:
    Antarctica
    CONCLUSION

    I'm not going to say Jude's results are wrong or that second order distortion was intentionally added to the HD800S. I will say that the HD800S results here taken with the EARS are consistent with Tyll's (professional measurement equipment) and others (DIY measurement equipment); and that the HD800 measurements are consistent with all my past measurements showing the pattern of higher third order to second order. The HD800S does exhibit considerable more second order than the HD800.

    I will also say that the quality of the HD800S's bass does sound slightly different from the HD800, a bit warmer and syrupy. I do wonder if a tiny bit more damping or mass was added to the HD800S's diaphragm in order to lessen the 5-6kHz peak on the HD800. The reason I think this might be the case is because the SDR resonator mods still didn't quite push down the peak of the HD800 to HD800S levels, so I suspect there might be more in play to the HD800S than just the resonator.

    Jude's seems to imply that he knows what's really going on with Sennheiser the HD800S, and that the HD800S doesn't have any more distortion than the HD800. My ears tell me differently. The effect is not necessarily a bad one as the HD800 is often criticized for being dry sounding, especially from solid-state amps. Either way, and despite the Jude selfie with Axel Grell, I'd rather hear from the man himself to comment on these results than than Jude. My bet is that Axel will never respond publicly. He's not obligated to and he shouldn't.
     
    Last edited: Jan 22, 2018
  18. sorrodje

    sorrodje Carla Bruni's other lover - Friend

    Staff Member Friend MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    2,778
    Dislikes Received:
    0
    Trophy Points:
    113
    Location:
    Dijon / Burgundy / France / EU
    Kinda balanced conclusion indeed.

    Besides that measurement war, the interesting point is how a lot of people suddenly claimed how HD800S bass was bloated , shitty and all that kind of overstatement after Tyll and a few others showed Distorsion measurements. A lot of people should learn to use their ears and build personal opinion... Dishonest people would have then much less power in this Hobby.
     
    Last edited: Jan 23, 2018
    aamefford and Priidik like this.
  19. Priidik

    Priidik Friend

    Friend BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    1,340
    Dislikes Received:
    2
    Trophy Points:
    93
    Location:
    Estonia
    The third harmonic content is quite odd. In speaker drivers this would indicate compromised motor design or too stiff spider.
    The HD800S impedance plot suggests slightly higher mech q, perhaps the peripheral membrane is softer.
     
    sorrodje likes this.
  20. sorrodje

    sorrodje Carla Bruni's other lover - Friend

    Staff Member Friend MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    2,778
    Dislikes Received:
    0
    Trophy Points:
    113
    Location:
    Dijon / Burgundy / France / EU
    Yea. I suspect they worked on the mesh permeability for the HD800S.
     

Share This Page