ZMF Aeolus Impression and Measurements

Discussion in 'Headphone Measurements' started by Hands, Oct 28, 2018.

  1. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Pyrate MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    12,285
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    Ah, does my unit have the Auteur baffles? It has a lot of ports. Nonetheless, with all the ports open and with flat pads in place, the sound is rather neutral and balanced overall.

    Maybe sell a neutral variant of the Aeolus like this? ;)
     
  2. zach915m

    zach915m MOT: ZMF Headphones

    Pyrate
    Joined:
    Oct 5, 2015
    Likes Received:
    2,517
    Trophy Points:
    93
    Yeah if someone wants the totally neutral version we would just use the auteur baffles. Not sure I'm ready to make a variant "official" but I'd be happy to accommodate any requests!
     
  3. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Pyrate MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    12,285
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    I'll have to get some measurements up then and see if I can ruin your inbox with custom Aeolus requests!
     
  4. Taverius

    Taverius Smells like sausages

    Pyrate
    Joined:
    Dec 27, 2017
    Likes Received:
    3,026
    Trophy Points:
    113
    Location:
    Rapallo, Italy
    What if I want mine double strength extra warmpoo? XD

    Jokes aside making it official seems a slippery slope, obviously hard to compare but look at Jason and Mike trying hard to decrease how many different things they're selling.

    But yeah I'll keep my order magnesium warmpoo thnx.
     
  5. famish99

    famish99 Friend

    Pyrate
    Joined:
    Dec 26, 2017
    Likes Received:
    1,714
    Trophy Points:
    93
    Location:
    Austin, TX
    The Aeolus was my last hope for a ZMF headphone: I loved the timbre of the Auteur but couldn't deal with its tonality, while the Atticus had a mostly enjoyable tonality (a touch too dark for my tastes) but lacked the technicalities of the Auteur. The Aeolus had me hopeful that it would be a mix of what I enjoyed from both, and to some extent it has succeeded in that. Most of my comparisons will be made against my newly purchased Clear and modded 6xx.

    My time with the loaner has been somewhat of a journey since it didn't ship with an XLR cable, which made me use the LCX for a day until I could DIY an adapter cable. Through the LCX, the treble peak/ringing on the Aeolus made the treble etch of the amp painfully noticeable, but otherwise I liked the rest of the tonality and enjoyed the midbass bump that wasn't overpowered by the subbass like on the Atticus. However after fashioning an incredibly crude XLR4 -> 1/4" TRS adapter, the treble peak was far less noticeable on the Af and was a lot less fatiguing than it was on the Auteur as well. There's a slight lower mid recession when paired with the Af as well.

    DISCLAIMER: I hesitate to make a final judgement call of the Aeolus's technicalities because it was most likely somewhat gimped by my crap adapter job. The clarity and layering seemed to be in the ballpark of the Auteur and Clear and well ahead of the Senns. Bass was much tighter and more detailed than my modded 6xx, but less so than the Clear. Mids and plankton are better on the 6xx than the Aeolus (or Clear). Headstage also seems smaller than 6xx by a smidge and has an ever so slight 3-blob effect especially compared to the Clear's stage precision (I admittedly suck at judging headstage though, mostly a speaker guy). Aeolus' transients were a touch better than 6xx but easily bested by the Clear.

    During the loaner, I also met up with @Joshvar and @ColtMrFire and spent some time listening to the Aeolus out of Colt's Gungnir Multibit A1/T3 (GE 12AZ7) combo, which I felt may be a better pairing for the Aeolus than the Af since the higher Z output helped fill in the bass and lower mids a little more and the T3's tube pairing made the treble peak much less noticeable as well. The combo also helped wetten the sound a bit, which I found a bit more enjoyable compared to the dryness of my setup (mostly due to my DAC).

    As always, build quality and ergonomics are fantastic. The cable split ended up being less of an issue than I expected, and the headphone feels very light for its weight as well as the pads are incredibly comfortable.

    In conclusion, if I hadn't recently bought a Clear due to the Source AV's recent deal, I probably would be ordering my own Aeolus. I would say for those looking for a super HD650, the Aeolus is close but sort of misses the mark a bit, especially with regards to the treble. It's still a bit fatiguing for long listening sessions, but I greatly prefer it to the Auteur and it's a great headphone if you just want to have fun and enjoy your music.
     
  6. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    [​IMG]

    ZMF Aeolus (loaner unit) impressions and measurements

    I like small boutique headphone makers. Compared to big boys, personal or small-group manufacturers seem to continuously improve themselves by actively reflecting feedback from users and/or audio community. The two of them I've kept my eyes on are MrSpeakers and ZMF. It's interesting these two follow slightly different paths: Dan seems to learn from failures (imo) and Zach is likely to explore his own various tastes with very solid level of perfections. Anyway, for this reason, ZMF products, particularly their dynamic ones, have been the safest when recommending to others, at least from my previous experience of Eikon, Atticus, and Auteur. And yes, now Aeolus is added to ZMF success for sure.

    Aeolus (abbreviated as aeo for convenience) shares most of zmf wood aesthetics and ergonomics. They fit most of human skulls and suspension head band does the justice. The loaner unit (Aeolus Zircote) weighs 16.7 oz (473g) with my kitchen scale, which is a bit heavier than the loaner auteur teak (16.5 oz or 467g). AFAIK teak is the lightest option among auteur wood, zircote may be on heavier side of available wood. @zach915m or anybody please correct me if I am wrong. Whatever Aeo's ergonomics is in the same ballpark of auteur.

    Regarding the sound, I think Zach was right in describing the following:
    Overall perceived performance must be a combination among personal preference, genre tastes, and upstream gears. I'd like emphasize the last. Zach's dynamic offerings are commonly very transparent and do not commit much sin (submission/omission). Thus, how they are amped (less important) and how they are sourced (more important imo) really matter.

    That being said, extrapolating my assessment beyond the associated source/amp should be cautious.

    Associated components
    • Dac: Dangerous music Convert-2 -- connected to PC via usb, master clock mode used
    • Amp: Eddie Current Black Widow
    • Other: Radial ISO twins isolator (Jensen 1:1 transformer used), PSA PPP AC conditioner
    • Headphones: Sennheiser HD600/800, MrSpeakers Ether 2

    Aeo's sonic character and philosophy are more in line with Eikon rather than Atticus. This is interesting because material-wise aeo is open atticus and auteur is open eikon. Very entertaining and involving. But not as much colored as Atticus. I'd put zmf dynamics (except verite which I haven't heard to date) in the following order: Atticus > Aeolus >= (or =) Eikon > Auteur (from the most colored to the least colored). The order between aeo and eikon can change depending on how different coloration is weighted.

    Bass is pretty deep and neutral. Well down to the lowest notes (20-25hz) of pipe organs or movie soundtracks. Sub and mid-bass are almost perfectly balanced for dynamic headphones, maintaining good traits of dynamic drivers. In this respect aeo and auteur are pretty similar. To my ears, aeo is better at tightness while auteur do differentiate bass nuance more. But in each criterion, only by a small margin. HD800 win both in reproducing tighter bass but deep notes are not as clear and clean as aeo/auteur. Empyrean has similarly deep bass but far beyond aeo/auteur in both tightness and differentiation. Ether 2 sounds noticeably cleaner and clearer with a pinch of warmth here but I suspect tons of ownership bias or synergy.

    Like hd600 with dekoni pads or empy, aeo has relatively recessed high mids. I think this part is a key factor of aeo coloration and "fun". Main players of most songs (vocals, guitar, etc) sound a bit distant and less present. Bites of saxophone and cymbal clang are slightly limited, too. But none of them are annoying or bothering. They are controlled very tastefully (Zach is very good at handling such trade-offs). At a cost of above, well-tuned V-shaped tonality introduces lots of fun and involvement. And potentially benefit low-volume listeners, too.

    There are healthy amount of high trebles. Atticus also do this. But aeolus treble is way more refined. Clicking kick drum, shining electronic guitar, crisp snare drum, and sizzling hi hat are very satisfying and pleasing. Needless to say any of these does not sound "too much". Auteur are largely controlled in this region, but do have some sibilance that rarely bothers. With auteur pads, this problem becomes more remarkable (this is why I like eikon pads for auteur). Anyway, aeo still has this issue, to much less extent though. Empy is good at dealing with this region. Probably as skillful as aeo. Ether 2 takes pretty conservative attitude in high trebles. Absence of harshness at a cost of less engagement.

    Technicalities, specifically resolution and dynamics, are good enough to put some multiple-grand headphones in a shame. Senn hd600/800 also do this but they are either limited in low freq (hd600/650) or inherently problematic in tonality without massive mods. Aeo just sound good out of the box, and may scale with better amps and/or sources. Empyrean are a bad (in my book) example of this. Although their tonal balance was really good and entertaining, high quality signal from DMC2 was restricted by empy's limited technicalities (=sins of omissions) and too lazy transience. Aeo have at least no less fun than empy without sacrificing too much. Like I said in another post, I hear more details and more volume gradation with aeo sourced by dangerous cv2 than with auteur sourced by rme adi2.

    Transience is decently fast. Not at the level of focal headphones, but faster than most of others. Auteur have slower and softer attack but I believe auteur's decay feels more natural and smoother than aeo. Driver material difference (tpe vs bio-cellulose) may explain this. I digress though. In terms of cup resonance, neither aeo nor auteur is as evident as atticus. But both have more of it than eikon. To my ears, aeo utilizes more internal resonance than auteur.

    In sum, I think zach made it again. Aeolus is not cannibalizing other products, but holding the own place for a certain group of customers. Indeed aeolus and auteur were in my shortlist last fall when I decided to get my end-game cans. If I knew what I know now, my decision might be harder. While I still prefer auteur to aeolus by meaningful margins, the decision between the two may vary depending on personal tastes, like @famish99 .
     
    Last edited: Feb 2, 2019
  7. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    Here are MiniDSP EARS measurements with SBAF compensation curve. See more in this and related threads: https://www.superbestaudiofriends.o...g-sbaf-compensations-from-minidsp-files.7067/

    upload_2019-2-2_14-28-55.png

    Channel balance is pretty good. Some discrepancy is seen in trebles, but mostly due to positioning. Symmetric positioning between left and right sides is a bit tricky for zmf headphones. Also I am a bit lazy in taking care of positioning these days. Anyway nothing wrong.

    As expected, there is slightly (3-4db) decreased high mids with a noticeable dip at 4k. But I didn't hear that much loss there. Probably mitigated by time-domain behavior around 4k, presented in spectrogram.

    Look at the ruler-flat bass response between 20-1000 hz.. Well done Zach!

    Take a look at how aeo fr differs from those of other headphones. Graphs are normalized at 1k and 1/6 oct smoothed.

    upload_2019-2-2_14-29-37.png

    upload_2019-2-2_14-29-45.png

    upload_2019-2-2_14-29-52.png
     
    Last edited: Feb 2, 2019
  8. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    upload_2019-2-2_14-30-38.png

    Zach's first-gen dynamic headphones had weird peaks in harmonics (they didn't sound bad though). But it seems that he could control them in open-back designs. Ignore a small peak at 4.5k hz, commonly shown in my other EARS measurements. 3rd harmonics exceed D2 for 70-100 hz but my ears couldn't hear remarkable troubles there.
     
  9. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    upload_2019-2-2_14-31-40.png

    upload_2019-2-2_14-31-52.png

    1/12 oct spectrograms look interesting. It shows a hint of cup resonance (400-1k) -- compare this with auteur spectrogram (attached below) -- and slower decay in 4-5 khz region, which may explain why I couldn't perceive 4k loss in fr as severe as the graph suggests. Both aeo and auteur spectrogram shows there are more energy than more physically open designs. I am guessing this is associated with the perception of "somewhat closed" feeling with all zmf headphones I've experienced.

    See other spectrograms for comparison.


    upload_2019-2-2_14-32-15.png
    (HD600)

    upload_2019-2-2_14-32-28.png
    (Ether 2)

    upload_2019-2-2_14-32-53.png
    (HD800)

    upload_2019-2-2_14-33-4.png
    (SR-207)
     
  10. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    Forgot attaching auteur spectrogram.

    upload_2019-2-2_14-40-52.png
    (Auteur teak)
     
  11. zach915m

    zach915m MOT: ZMF Headphones

    Pyrate
    Joined:
    Oct 5, 2015
    Likes Received:
    2,517
    Trophy Points:
    93
    Just wanted to make one note about the Ziricote Aeolus tour unit - I am pretty sure it has Unverse perf pads on it, as I sent it to @Elnrik before we got the final final version of the universe pads to be used on the Aeolus. I will send @Elnrik all the possible pads that can go on the Aeolus (that sound good IMHO) so that these can be compared subjectively as well. From warmest to most "neutral" it goes Aeolus Pads>Universe Pads>Verite Pads.

    So to clarify, what I did was change the density of the foam slightly and the outer covering material to match the Universe pads best for the Verite and Aeolus. About a year and a half ago, the universe pad was a pad I developed for the auteur, but ended up going by the wayside. I did like the pad so much to put it in general production - but I didn't have a headphone to pair it with yet.

    So when I was finalizing the Verite and Aeolus I realized the pad needed to be amended to work best with each respective headphone so minor changes were made. The Universe pad is what's used on the Verite, and soon, I will list the "Aeolus" pad on our site as well, which is what is used on the Aeolus. In short the Aeolus has thinner Lambskin with denser foam, and the universe pad for the Verite has thicker Sheepskin and less dense foam, it's also ever so slightly taller/deeper than the Aeolus pad.

    Here's some measurements showing the Aeolus pads, Universe pads and the Verite pads on our test rig (on the Aeolus). I use compensation that's very close to DF, but I made some adjustments that skew it a touch towards the InnerFidelity curves so that the curve matches as closely with what I'm hearing as possible. This is all done on an AP 515, 7101 and an AECM206 with an anechoic chamber box we built.

    Aeolus_Pads_1.jpg

    Here's a quick take on the THD for the Aeolus:

    Aeolus_THD.jpg


    In the coming months I'll be dedicating a page on our site to measurements of our headphones and my tuning process and all that so information on my process is more readily available. It's something I get asked about a lot and I'm happy to share!
     
    Last edited: Feb 3, 2019
  12. Taverius

    Taverius Smells like sausages

    Pyrate
    Joined:
    Dec 27, 2017
    Likes Received:
    3,026
    Trophy Points:
    113
    Location:
    Rapallo, Italy
    @zach915m do I need to have another pad added to my order on top of the perf suede?

    I'll admit I'm having a hard time keeping track :oops:
     
  13. zach915m

    zach915m MOT: ZMF Headphones

    Pyrate
    Joined:
    Oct 5, 2015
    Likes Received:
    2,517
    Trophy Points:
    93
    Send me an email I'll make sure you have everything you need!

    IMHO between the stock aeolus pads, Universe Suede perf and Verite pads you'll have all the needed sounds of the Aeolus. Most users find one pad they stick with though. If I had to choose two pads I'd go with Universe Suede perf and Aeolus pads for my preferences.
     
  14. Elnrik

    Elnrik Super Friendly

    Pyrate
    Joined:
    Jan 10, 2017
    Likes Received:
    8,973
    Trophy Points:
    113
    Location:
    Denver CO
    Home Page:
    I had no idea about the pad change. Sorry if this caused any problems and confusion.
     
  15. zach915m

    zach915m MOT: ZMF Headphones

    Pyrate
    Joined:
    Oct 5, 2015
    Likes Received:
    2,517
    Trophy Points:
    93
    It's totally my fault for not remembering that I sent your set before the aeolus pads were in.

    None of the production sets were sent out with Universe pads - but they can ofcourse be used without issue to match the measurements above.

    EDIT: It should be noted that all the other measurements in this thread are accurate to the production model as they were matched or used the production pads and that the production pads are the only difference between these sets.
     
    Last edited: Feb 3, 2019
  16. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Aeolus Frequency Response and HD on EARS using SBAF compensation on another set. Looks consistent with @Vtory's pair - considering different operators - which will result in some variance. Keep in mind that I run headphones near 100db for the distortion tests.
    ZMF Aeolus Frequency Response.png
    Aeolus Dist.png

    One thing to be aware for spectrograms is to normalize the T=0 values at their respective frequencies. The Aeolus may only appear to have "slower decay" in the 400Hz region compared to the HD600 because of its FR tilt. In other words, match volumes in the areas which are you looking at, otherwise one of them might be cheating.

    As far as 4kHz, dips are always harder to hear than peaks, especially dips less than 1/3 octave.
     
  17. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    Definitely agree. So both normalized and unnormalized results are meaningful. Or at least we need to take care of sound pressure at t=0 instead of just comparing colors. In general, unnormalized ones are better to present absolute level of sound energy remaining at certain time point, while normalized ones indicate relative degree of decay more intuitively.

    The tricky thing is that "90 to 70 db spl" and "70 to 50 db spl" are not always the same. I think if every condition is the same, the latter takes more time due to the nature of time-decay process. In normalized scale, both must be the same (-20db).

    Anyway very good point. Worth noting.
     
  18. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    I'd update measurements for Aeolus with perforated universal pads. Measured at 95dbSPL @ 300hz except distortions -- measured at 100db spl -- with newly-measured step response (regularly added to my measurement protocol).

    upload_2019-2-22_13-27-25.png

    upload_2019-2-22_13-27-32.png
    (noise:brown, D2:red, D3: darker yellow, D4: light green, D5: lighter yellow)

    upload_2019-2-22_13-27-46.png

    upload_2019-2-22_13-27-52.png

    (It's my typo to write down "universe" instead of "universal".. but kinda think universe sounds cooler lol)

    Now the results are even more consistent with @purr1n except slightly lower levels of harmonics.
    It's also interesting to see aeolus have reversed polarity.

    PS. Auteur also had the same polarity. Maybe applies to all zmfs? (verification needed)
     
    Last edited: Feb 22, 2019
  19. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    I am enjoying the loaner period part 2. Different pads clearly have widened capability and competency.

    Having four pads (one shipped with aeolus) in the house: perforated universal lambskin, perforated aeolus lambskin, perforated verite lambskin, perforated universal suede. For convenience. I'll call them universal, aeo, verite, and suede pads as short names.

    One-sentence summary of subjective impression for each set of pads:
    • Universal pads: the least preferred by comparison.
    • Aeolus pads: Warmer, more zmf-like, and much more entertaining (the best fun among theses four). Similar tuning direction with universal one, but with much better treble balance and coherence.
    • Verite pads:I totally understand why @cskippy preferred this way more to universal pads. Transform aeolus to truly reference-grade neutrality.
    • Suede pads: the most preferred. Overall sound becomes even tighter and better controlled, with a bonus of top-level neutrality. Imho aeolus with suede pads OUTPERFORM auteur (with either auteur or eikon pads), which have been my de-facto neutrality king until recently.
    So interestingly there are two groups: fun (universal and aeo pads) and reference (verite and suede pads)
    And yes, suede pads raised my evaluation toward aeolus. If I go for aeolus, suede pads would be the top pick for myself.

    Measurements and more detailed discussions to come. Hopefully by Sunday.

    PS. Aeolus pads are noticeably smaller than the other three pads, which gave me very hard time to attach them to aeolus. Aeolus users who want to change to these pads MUST WATCH Zach's youtube video MANY TIMES (particularly how Zach attached ori pads to auteur).



    PS2. Roughly speculating verite headphones with suede universal pads would be the real killer.

    (Quotes from verite discussion)

    I believe suede pads can be a good candidate for the mentioned needs and wants. Don't have verite to test this though.
     
    Last edited: Feb 25, 2019
  20. Vtory

    Vtory Audiophile™

    Pyrate MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    10,831
    Trophy Points:
    113
    Location:
    East Coast
    One thing to note is that for time-domain analyses, I prefer to use spectrogram based on WAVELET.
    The default setting of rew spectrogram is fast-Fourier transform (FFT), not wavelet.

    Remember that wavelet-based spectrogram looks different from fft-based one. Here is the example of fft-based one (aeolus with univ pads).

    upload_2019-2-22_14-45-55.png

    My choice and preference are based on the following beliefs:
    • FFT results vary extremely depending on signal windowing (window type and window time -- the latter is associated with time-frequency resolution tradeoffs). No right or standardized choice for acoustic measurements afaik. Each windowing parameter has its own merits and demerits and should be chosen based on specific purposes.
    • ^ also largely (slightly different problems but still confusing parameter choices) applies to waterfall. I don't like 3d-presentation of waterfall plots either, because of visual challenges (e.g., harder to read levels and frequency).
    • By contrast, constant-Q wavelet is easier to control. Invariant frequency resolution is more intuitive to choose (1/12 oct is my preference). I also believe wavelet-spectrogram can reflect how we actually hear each frequency slice in a more straightforward way.

    EDIT: The screenshot above is to illustrate confusing and complicated options of fft. But someone asked cleaner pic. Attached below:
    Aeolus spectro univ - fft 2.jpg
     
    Last edited: Feb 23, 2019

Share This Page