Burst Response! HD800, SR-207, HD650

Discussion in 'Measurement Techniques Discussion' started by purr1n, Jan 8, 2017.

  1. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Pyrate MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    12,284
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    Er, wait, you found the Clear muted in the upper mids compared to the Elex? I would have expected the Clear to look, um, "wilder" than the RS2e in some of these results, and to think the Elex might sound less muted in the upper-mids than the Clear...I'm already cringing a bit.

    Agreed, the Utopia would probably be an interesting one to test. It had some peculiar characteristics that might shed more light on things.

    It's curious, you find the Elex can be difficult with its attack, and yet these visualizations make it look pretty well behaved compared to the RS2e...At least in the last two frequency results. But it seems we'd both agree the RS2e might be easier to listen to, if nothing else because it's slightly dark in some ways.

    Which, speaking of, I don't hear a lot of crunch or other weird effects on the RS2e. It sound surprisingly subdued, polite, possibly even dull in some regards. I'm not sure I hear it as unmistakably Grado in the highs, though my relatively limited exposure to Grado compared to some means I'm less qualified to make this statement. Nonetheless, it makes the HD650 sound sprightly in the highs, at least. (May be an extreme case of fat head syndrome.)

    But, yeah, I totally understand the goal here is the opposite of "one measurement to rule them all." More a matter of seeing if we can fill in those mythical subjective and objective gaps, including all the confusion, doubt, excitement, and sometimes seemingly contradictory data that comes with the territory.

    And no measurement of any kind will be a total match for all listeners. It takes time to process and integrate what these might mean, as you said, including what you, as an individual listener, might have to decide doesn't really follow what you hear. Even fat head syndrome can make for some interesting discrepancies in select cases.
     
  2. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Pyrate MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    12,284
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    This is why I've always recommended folks consider keeping other possible front damping material on hand to play with. I think a lot of DCA headphones are of the sort where the included filters might get you 80% of the way there, but the best solution is probably one you pick yourself.

    Now, while I am not suggesting people should want to have to deal with that, the DCA headphones are nonetheless worthy of consideration if one isn't afraid to try a bit of filter rolling. Like on the HD650 mods, one simple option is cutting a small hole in the center of a filter that's just a bit too much otherwise. Maybe put a really light secondary filter material in there as well, as a whole piece or just to fill that hole.

    You mentioned you felt the ear pads were too thick on these. I don't disagree, but I also found it less of an issue probably because of my FHS (fat head syndrome). Given enough time for the memory foam to warm up, I found they compressed quite a bit and sounded better for it. May be one reason they didn't sound dulled or dead to me even with the thickest filter. I'm curious if the 50Hz results would differ much on my head, as I heard the bass in general as quite strong.

    It might also be that I'm more forgiving of closed back stuff, because it's slim, slim pickings in that field, and always full of compromises.

    (Sorry for that being a bit off topic, but tried to circle back to something related to this discussion.)
     
  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
    Latest stuff. Note db scale down to -40db on the decay changes things.

    B1696 HD800.wav_atk.jpg
    B1696 Verite.wav_atk.jpg
     
    Last edited: Apr 23, 2020
  4. Martigane

    Martigane Acquaintance

    Joined:
    Mar 5, 2019
    Likes Received:
    97
    Trophy Points:
    18
    Location:
    Belgium
    Thank you Purr1n for working on this measurement method and sharing those with us.

    You are going deeper than ASR in the "how can we measure/visualize what we subjectively hear". Something I am very interested about.

    Did you measure the hd58x? I d be curious to see how much of an improvement we would see from a simple disk mod (4mm hole) in the back of the magnet (solved 5khz ringing and flatenned midbass hump according to my measurements).

    Applying the same method on speaker would yield interesting results, but decay would really be compromised by room interactions. Even attack for sub 2k frequencies I guess. (10periods are 5ms already). Did you try it?
     
  5. CEE TEE

    CEE TEE MOT: NITSCH

    Pyrate IEMW
    Joined:
    Sep 25, 2015
    Likes Received:
    3,715
    Trophy Points:
    93
    Location:
    California Bay Area
    Home Page:
    ^Wow, HD800 and Verite share many similarities on those measurements above!
    Have not heard them yet...does Verite sound like a modded or different flavor of HD800?
     
  6. 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
    It could be possible for speakers if done nearfield - to some extent. This is new, so really we need to get more data rolling in and seeing is there is any relationship between what we hear and what's in the graphs. What matters and what doesn't? Important questions to answer.

    Lots of ideas, but not enough time to explore. Maybe one day when I am a semi-retired 1%er like Amir. I got young ones.
     
  7. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Pyrate MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    12,284
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    They share a similar-ish upper-mid/lower-treble depression that can make rock sound a bit softer than usual. Lots of nasally crunch in that area that affects distorted guitars.

    Playing with the HD800 driver in a different enclosure, there have been times they've reminded me of the Verite tonally. But I'm not sure it quite matches the unique sense of speed and clarity on the Verite.
     
  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
    The devil is in the details. The following is what I would say based purely on the measurements and current datasets (not taking into account what the headphone is). Sometimes this is right, sometimes not so right, keeping in mind that frequency response will also have an effect.

    upload_2020-9-29_9-36-59.png
    1. Grayness, indistinctness, or murkiness in lower bass
    2. Slowish decay for lower bass
    3. No oddities in high-end - no sandiness, no excess grain, jumpiness, spikey, etc.

    upload_2020-9-29_9-44-20.png
    1. Fast decay in lows
    2. Sharp leading edge at first tick for 500 and 1500Hz = perceived fast or crisp transients
    3. Mellow high-end timbre
     
    • Like Like x 9
    • Epic Epic x 2
    • List
  9. Biodegraded

    Biodegraded Friend

    Pyrate Contributor
    Joined:
    May 28, 2017
    Likes Received:
    7,952
    Trophy Points:
    113
    Location:
    Vancouver BC
    For the decays you're showing just the first 5 cycles, right? Maybe change the legend to reflect that?
     
  10. crenca

    crenca Friend

    Pyrate
    Joined:
    May 26, 2017
    Likes Received:
    3,798
    Trophy Points:
    113
    Location:
    Southern New Mexico
    Tis Interesting. I don't perceive the the Verite (closed in my case) as having particularly fast/crisp transients in that 500/1500 range when compared to Focal Clears/Utopia (I am assuming that the Elex measurements are indicative of what the Clear/Utopia would measure). I wonder if I am not adequately distinguishing the difference between microdynamics and transient speed?
     
  11. Superexchanger

    Superexchanger Friend

    Pyrate
    Joined:
    Jul 11, 2019
    Likes Received:
    1,309
    Trophy Points:
    93
    Location:
    Flavortown, USA
    Newer time-domain techniques are fascinating. The burst-response behavior is exciting to see for some of these reference headphones.

    I had a question in passing about this. My (partial) understanding here is that the burst response highlights greater detail regarding dynamic mechanical response of the driver to a stimulus, coupled to a measuring device via a medium (air), with the cup/baffle/pad behavior potentially mediating the response. This seems sound (pardon the pun), because these factors are potentially audible/perceivable. Right?

    Is it in any way complementary or interesting to know the purely mechanical responses here, with potentially reduced mediating factors? I'm imagining what you would learn with perfect driver position vs time information, in addition to the "coupled device" measurements above. Would it further the understanding of what we hear, or is it simply redundant to the above data?

    I ask this because I design high speed, high sensitivity (MHz-GHz, μm resolution) magnetic sensors that could retrieve this type of information. It sounds like a fun application if it provides anything of value.
     
  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
    I wouldn't necessarily think of it as crisp transients in the 500/1500 range. Remove any association with any particular frequency. There could be association, there may not be. We need to get use to some vagueness, because none of these measurements alone are the one ring that rules them all and binds us to darkness. And then there is decay. And frequency response. And distortion. All these things need to be considered to give us the while picture.

    These measurements must never be taken as absolutes. Only the Sith and worshippers of SINAD do that. These measurements are best used in comparison. I think most with Verite would say that from a transient response and "speed" POV, they in general sound "faster" than HD800s. Let's Verite compare to Utopia.
    1. Marginally faster decay in lows (I sort of drew over it, but by the first mark it is a bit lower - this calls into question whether the visualization should be modified, maybe 8db per division instead of 10)
    2. In addition to the rising edge at 500Hz and 1500, Utopia also has one at 4800.
    3. The second and third marks are a bit elevated at 4800Hz - the "pricky" effect or sometimes @Hands has said grainy.
    upload_2020-9-29_11-21-56.png
     
    • Like Like x 3
    • Epic Epic x 1
    • List
    Last edited: Sep 29, 2020
  13. 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
    Most of the measurements we typically see: FR, harmonic distortion, IME, SINAD are steady state. I believe that, no, I've shown that there is even greater distortion when it comes to transients.

    The best way to measure this is with lasers. The distortion inherent in the capsule mic is reflected in these measurements. Lasers would do away with this. Maybe someone who is interested in audio science, like ASR, Harmon employed doctorates in sound, or an enterprising university graduate student can pick up from here. This stuff I am doing is just partially baked. The reason I haven't gone farther is because none of this stuff pays any money. I'm keeping it hobby level.
     
    Last edited: Sep 29, 2020
  14. Superexchanger

    Superexchanger Friend

    Pyrate
    Joined:
    Jul 11, 2019
    Likes Received:
    1,309
    Trophy Points:
    93
    Location:
    Flavortown, USA
    I've got the spare parts for a laser doppler vibrometer lying around. More to come [/s].
     
  15. Philimon

    Philimon Friend

    Pyrate Contributor Banned
    Joined:
    May 6, 2016
    Likes Received:
    2,091
    Trophy Points:
    113
    Location:
    Portland, OR
    Why cant burst response be shown like CSD measurements? But instead from a flat plane (no signal) to ON which would show the rise and speed to reach full amplitude. And since formatted like a CSD we can see the entire frequency range instead of these select 4-5 points. Sometimes these 4-5 points happen to fall right at a headphones ringing or resonance issue which might paint a different picture vs if we could see the entire frequency range like in a CSD.

    @purr1n etc
     
  16. Philimon

    Philimon Friend

    Pyrate Contributor Banned
    Joined:
    May 6, 2016
    Likes Received:
    2,091
    Trophy Points:
    113
    Location:
    Portland, OR
    Related: Transient Response Wiki. Seems pretty congruent to how we might interpret burst responses, see: intro, damping, and properties.

    Related: Testing Loudspeakers on Audio Express. See: dynamics (towards bottom of page) -> short-term dynamic compression and “spectrum dynamic response test.”
     
  17. 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
    This is basically a steady-state linearity test using 5 tones. Not gonna see much of a difference as long as the transducer is playing within its limits.
     
  18. Martigane

    Martigane Acquaintance

    Joined:
    Mar 5, 2019
    Likes Received:
    97
    Trophy Points:
    18
    Location:
    Belgium
    I get your point. Basically, you'd like the equivalent of a CSD but for the onset instead of the decay, keeping the time axis as number of cycles.

    @purr1n, i expect that you saw close correlation between a CSD and your decay characteristics? If not, was it a matter of stimuli? (Constant freq burst vs sweep)

    I am still very much interested in this unique behavior characterization.
     

Share This Page