[MrHaelscheir]

For me, it's the illusion that there could be more out there. It's a trick of the layering, that there are things close, then things further out, and then more things even further out that my mind extrapolate that there could be even more out there. I hear this with most tracks, but electronic tracks are usually best with songs like "Bloom" by Dabin and "With Love Until We Die" by Tristam and "Underneath My Skin" by HALIENE. In those tracks, the vocals are upfront, the main melody is a little further away, and then there are these little synths and effects fluttering in the background that sound far away on the X9K. The Corina and HD800S in comparison put all of the sounds in this even arc in front of me, just with arcs of different sizes.

Thus the X9K manufactures this layering and separation which just sounds delightful to me, like I'm floating in the midst of all of this sound. It does so even though it's not actually as large in soundstage as the HD800S, it just uses that soundstage better.

The best analogy to me is like how a very immersive video game produces the illusion that there's a larger world outside of the player character and their adventures. The play area is inherently bounded and NPCs don't actually have their own lives while the player is away, of course, but the best games make it seem as if there was a living world beyond what the player can see.

For "Bloom", I can concur that this track has a nice spatial background track that comes in and out, rendered reasonably through the HE1000se with Dekoni Elite Hybrid pads which I recently received (I'm finished flooding that thread with measurements for the time being...); I doubt I will ever have another chance to hear the X9000 with that kind of track. My classical orchestral recordings don't often have that presentation without speaker-like crossfeed, its being "wide", but closer up and bounded by the drivers. Some of Susumu Hirasawa's tracks can have some interesting spatial effects mixed in like where I could here a piano-like tune playing vividly in the distance, but also panned upward thanks to some headphone imaging/HRTF distortion compared to speakers.

My Jabra Elite 85h renders the background synthesized sounds more darkly, in this case detracting from their spatial character. My fine-tune-EQed Meze Elite with hybrid pads also intentionally relaxed what I normally find to be too forward 5 kHz to 8 kHz. Hmm. Switching to the unEQed HE1000se with Dekoni Elite Hybrid pads (a bit more forward upper midrange and a bit more relaxed lower treble compared to stock), I am hearing a nice forwardness and clarity to the vocals and presentation; the Arya Stealth with the stock pads is similar, but measurably with a less forward upper midrange and hence vocals; actually, that increase in clarity was an error in my volume-matching.

For "With Love Until We Die", the guitars are imaged from the drivers with a large "soundstage height" which I'd consider as mainly being thanks to the sheer side of the drivers and earpads. I otherwise didn't hear as much layering, at least not as notable as in the previous track. As for my EQed Jabra Elite 85h, it does funny enough and probably measurably so present a warmth per some bass I had still failed to tame. I wouldn't be surprised if the sideward imaging of the guitars and whatnot is technically similar, just subjectively limited in size by the smaller pads.

For "Underneath My Skin", the space of those knocks is rendered reasonably. I'd say I don't notice much differences between my EQed Meze Elite and the HiFiMans other than subtle differences in presentation related to pad size and measurable tonal differences. Anyways, I am certainly not hearing anything comparable to what happens when I turn on my there the sound is actually laid out around a meter before me and I can literally look at the individual sound sources. If you are experiencing convincing near-field-speaker-like holography or "surround sound" without any crossfeed, then I would be impressed.

 

[SolarCetacean]

near-field-speaker-like holography.

That's not what I'm referring to with the X9K's layered soundstaging. It doesn't sound speaker-like, nor is speaker soundstaging what I want from or enjoy about the X9K. I'll explain more later as I'm on my phone right now. But I think the analogy I used about the immersive video game is the clearest expression of what I mean with the illusion of boundless size. Not that anything I hear is actually wider or further away, but that my mind is presented with the illusion that there could be something further. Hence immersive.

 

[SolarCetacean]

Deleted. Double post

Edit: since I have this post here and I mentioned expanding on my previous thoughts (though I had decided to write more about that video game analogy), I'll expand even though I already covered what I had wanted to say.

Other headphones that I've heard like the HD800S and Corina have limited openness in that they present sounds evenly dispersed along an arc. To me, that's like hitting an invisible wall in a video game. You can't go further in a very artificial way which breaks the immersion and signifies that the game map has nothing more to offer. For those other headphones, it's like there's an invisible wall that's very clearly defines how far away a sound can be.

On the X9000, even though there's no actual soundstage beyond a certain point, the positioning and layering of sounds within its soundstage tricks the brain into thinking that there could be something else beyond. Since you have some elements close, then more sounds a bit further away, and then the backing harmonies beyond those, and the little background sounds even further beyond that, the mind assumes that there could well be yet more beyond, even if there isn't. Just like a clever boundary in a game map effectively restricts the player yet allows the player to imagine a world that's active and alive beyond the slice that's simulated in the map. Intellectually, every player will understand that the map is limited, and the wider game world is all an illusion of clever map design and skyboxes, but in the moment, it should feel like there's more to the world. Same deal with the X9000. I can reason about how its soundstage is limited and that it's layering is just a trick of the brain, but when I listened to it, even the second time around, I was sucked in and immersed. Hence I cannot relate to comments that the X9000 is sterile or unmusical, though I recognize that others in turn may not relate to my impressions and experiences, because I enjoyed every moment I listened to it.

 

[gammi]

Regarding the "X9000 is thin sounding" discussion recently. I went back to watch GoldenSound's Aperio review today for uhhh reasons and he actually mentions in that he's only ever heard three estats that had bass slam. CRBN, Bravura, and X9000. (and then the Aperio)

I really think anyone who felt the X9000 was thin was listening on a chain that isn't good for the X9000 or has a head/ear shape that somehow is a bad match.

Go to 15min:

 

[jamesjames]

@SolarCetacean Count me jealous of the comfort you are experiencing with these headphones. Is it simply a perfect fit for your head, as in proper seal at the front and back of the pads with even pressure throughout?

As for what you described with transient decay, unless it rather has to do with treble quantity, it could be correlated with the cumulative spectral decay (CSD; how each frequency decays for an impulse input, hence which resonances are excited) measurements. E.g. With you and others noticing that extra sizzle or whatever to transients on the Susvara, that might be the "hideous" CSD managing to be audible in actual music rather than just when listening to isolated impulses like in http://pcfarina.eng.unipr.it/Acustica-samples/Dirac.wav.


Figure 1: Susvara CSD from https://www.soundstagenetwork.com/i...fiman-susvara-headphones&catid=263&Itemid=203.

I would suspect the DCA Corina to manage to have a much cleaner CSD at the expense of dulled transients from the damping required to achieve such. I did measure a higher than average but controlled upper midrange CSD on the X9000 and Sennheiser HE-1, so estats with less damping probably do inherently trade off transient sharpness for a less clean transient decay which in this case you seem to prefer the sound of. My holy grail is to be able to couple both incisive transients with a clean decay so I can enjoy both that sharpness, impact, and tactility, as well as a "black background" not polluted by any decay products, my only hearing the actual decay within the recording. The Final Audio D8000 pro was probably the closest to that, but it encountered sub-bass clipping or rattle more quickly than the other headphones I tried, and its presentation is too small for my liking.

I did recently figure out how to display the step response of my headphones as calculated from the logarithmic sine sweep and hence impulse response; these make objective considerations of headphone "speed" much more intuitive to spot out and correlate with experience:

The step response in the below is always in teal.


Figure 2: HiFiMan Arya Stealth impulse and step response. Here, you can see that the powerful incisiveness I head out of it can be correlated to a sharp step response. On the other hand, that edge is dirtier, and the decay is dominated by a 4.3 kHz resonance, though one can EQ that peak down.


Figure 3: Meze Elite stock hybrid pads impulse and step response. Here, the attack of the step response is clearly rounder and hence duller.


Figure 4: Impulse and step response for the Meze Elite hybrid pads with my flat and Harman-like "V3 PEQ". Here, the magnitude and phase response corrections achieved with my minimum-phase EQ have incurred a speed and sharpness comparable to the Arya Stealth's stock performance, its having a cleaner step response decay and thanks to my EQ, greater sub-bass extension (if it extended all the way down to 0 Hz, then the step response would not decay like thus). The Stax SR-X9000 had a yet faster initial rise time, but a subsequent initially more rapid drop-off before rebounding and returning to normal decay, whether or not this adds to a sense of increased incisiveness; its impulse response also had a less prominent secondary spike which I guess contributes to its initially faster decay.


Figure 5: Impulse and step response for the Meze Elite hybrid pads with my personal HRTF applied by SPART AmbiBIN and EQed with Equalizer APO to my in-ear measurement for a free-field speaker panned 30 degrees left relative to my left ear. This is assessing my current binaural head-tracking setup where I simulate a perfectly neutral stereo triangle within an anechoic chamber. Transients are still incisive, though having a more rounded initiation, and the decay was initially faster and dirtier; note that this was without my preferred 5 dB 100 Hz low shelf.


Figure 6: Same as above, but with both the left and right channels playing in phase. Here, the decay envelope is more bolstered, possibly contributing more weight while the sound from the left channel alone might have been sharp but thin.


Figure 5: Audio-Technica ATH-M50xBT impulse and step response. My first audiophile headphone from 2019. Other than that initial spike, the step response is quite rounded or dulled.


Figure 6: Jabra Elite 85h impulse and step response. My work-at-home headset from 2021. Here, without EQ, the step response is quite duller yet than the stock Meze Elite. Otherwise, both this and the ATH-M50xBT had exceptionally clean CSDs falling below my measurements' noise floor.

In short, the perceived transient sharpness I heard when listening to http://pcfarina.eng.unipr.it/Acustica-samples/Dirac.wav correlated well with the measured step responses.

This is very interesting. I've long speculated that the time-domain issue with planars and e-stats - as shown by CSDs - is more audible than commonly suggested.

 

[MrHaelscheir]

This is very interesting. I've long speculated that the time-domain issue with planars and e-stats - as shown by CSDs - is more audible than commonly suggested.

One thing that I found was that the upper midrange and treble group delay and CSD could be cleaned up a bit (group delay made to look less noisy, and decay levels getting a bit lower) by holding up some acoustic foam next to one's head while taking the measurement, so like I had read suggested somewhere on ASR, that group delay noisiness may indeed be related to radiated open-back sound reflecting off of surfaces in the room and making their way back to the headphone. After that, there is a good chance that these artifacts come from internal reflections rather than the driver, hence why an otherwise very smooth frequency response seemingly lacking in resonance issues may have a dirty group delay or CSD; the Susvara may be a quite different story considering how jagged its FR is. E.g. IIRC, the X9000's pads have leather inner pad materials, which might contribute reflections, and likewise the angled protective grille. Closed-backs or those open-backs with clean CSDs (e.g. AKG) thus typically damp those internal reflections.

 

[jamesjames]

That's a good point, I think, re room reflections affecting the picture - I've read some material by Keith Howard (ex HiFi News reviewer) who makes a similar point. The 'resonance' issue has always seemed to me tremendously complex - meaning I could never feel confident about identifying particular causes for particular measurements. But my own experience is that the characteristic messy CSD correlates with a certain planar/e-stat 'sound'.

 

[Audio Addict]

I was wondering if anyone else in this thread lives in middle of nowhere Central Illinois and would like to stop by and listen to the RSA B-21 Raider prototype I have for a few weeks. I don't have much experience with amps outside the RSA world. To me this amp seems to be very special in its presentation with the X9000 and the 1266.

I wish I had more experience with the other amplifiers but this amplifier and the X9000 were meant for each. Then add its performance with the Abyss 1266 and you have to ask how could it get better? This is just using the Exogal Comet Plus with Roon running on a Mac Mini M2Pro.

 

[jamesjames]

That's a good point, I think, re room reflections affecting the picture - I've read some material by Keith Howard (ex HiFi News reviewer) who makes a similar point - and that this is often simply ignored. The 'resonance' issue has always seemed to me tremendously complex - meaning I could never feel confident about identifying particular causes for particular measurements. But my own experience is that the characteristic messy CSD correlates with a certain planar/e-stat 'sound'.

And to finish that idea, one reason I've suspected the Susvara (and other HifiMan phones) might be affected by an inherent time-domain behaviour issue is that I've encountered planars - the Meze Empyrean and Elite - which seem to me (for what it's worth) to have avoided what I hear as the characteristic planar sound. And they appear to be cleaner at higher frequencies (I've not been able to track down the measurements - but I believe I have seen CSDs for Meze phones which looked relatively clean). I gather also that HifiMan doesn't use diaphragm plastics having high internal losses which would dampen breakup resonances (as some others do) - and I wonder whether this might also be a factor.

 

[jamesjames]

I should add the Meze phones don't really appeal to me - but I do find them more engaging than other planars.

 

[jamesjames]

The Empyrean CSD can be found at Keith Howard's Headphone Test Lab page: https://headphonetestlab.co.uk/home-welcome.

 

[MrHaelscheir]

And to finish that idea, one reason I've suspected the Susvara (and other HifiMan phones) might be affected by an inherent time-domain behaviour issue is that I've encountered planars - the Meze Empyrean and Elite - which seem to me (for what it's worth) to have avoided what I hear as the characteristic planar sound. And they appear to be cleaner at higher frequencies (I've not been able to track down the measurements - but I believe I have seen CSDs for Meze phones which looked relatively clean). I gather also that HifiMan doesn't use diaphragm plastics having high internal losses which would dampen breakup resonances (as some others do) - and I wonder whether this might also be a factor.

I had first seen the Meze Elite's CSD in https://www.soundstagenetwork.com/i...e-audio-elite-headphones&catid=263&Itemid=203 where things looked decent in the midrange and treble compared to HiFiMans. I did a detailed measurement comparison between the Meze Elite, HE1000se, and Arya Stealth within https://www.head-fi.org/threads/hifiman-he1000-se.886228/post-17865003 (post #4,788 and on), its maintaining a pretty decent CSD even after EQing the upper midrange up. Its sine sweep and multi-tone distortion performance is exquisite for a planar. I also found that absorbing the sound radiated from the grilles does improve the CSD measurement as likely influenced by reflections from external surfaces.

 

[jamesjames]

That's very interesting indeed, thanks. As mentioned above, to my ear the Elite is the best planar out there by some margin. And I would prefer it to the Stax phones I've owned too. The more I see of your results the more convinced I am that the planar resonance issue isn't just theoretical (ie, it's actually audible).

 

[gammi]

I had first seen the Meze Elite's CSD in https://www.soundstagenetwork.com/i...e-audio-elite-headphones&catid=263&Itemid=203 where things looked decent in the midrange and treble compared to HiFiMans. I did a detailed measurement comparison between the Meze Elite, HE1000se, and Arya Stealth within https://www.head-fi.org/threads/hifiman-he1000-se.886228/post-17865003 (post #4,788 and on), its maintaining a pretty decent CSD even after EQing the upper midrange up. Its sine sweep and multi-tone distortion performance is exquisite for a planar. I also found that absorbing the sound radiated from the grilles does improve the CSD measurement as likely influenced by reflections from external surfaces.

Mildly off-topic here but I would LOVE to see a CSD measurement of Expanse since it has that AMTS on the driver that absorbs backpropagated waves. I bet it has a unique result. Maybe I'll figure out how to make the measurements on my own pair.

 

[MrHaelscheir]

Mildly off-topic here but I would LOVE to see a CSD measurement of Expanse since it has that AMTS on the driver that absorbs backpropagated waves. I bet it has a unique result. Maybe I'll figure out how to make the measurements on my own pair.

I can say that Earfish did recently respond to an inquiry about an issue with the head-tracking app, though they did say they had been quite busy with university matters. Receiving the measurement kit from them would be the most convenient way to get measurements of any headphone and see results specific to your own ears, plus the indispensable experience of having your personal HRTF measurement for use in binaural head-tracking or EQing to a speaker response. The default Axagon ADA-17 audio interface suffices for basic frequency response and distortion measurements, though you won't get accurate phase response and hence impulse, step, and probably CSD measurements without a proper audio interface for synchronizing samples. As you may have seen in some of my posts, connecting Earfish's unbalanced plug-in power in-ear microphones to an audio interface can be a bit of a hassle. I would warn that Earfish's in-ear microphones can get pretty uncomfortable in longer measurement sessions especially after pressing them hard to help ensure a more consistent seating between measurement sessions; even with a 192 kHz sample rate through an audio interface, the lowest-noise averaged measurements would require sitting still and quiet for 3 minutes straight, but "good enough" measurements can be taken in 21 seconds or less. miniDSP EARS may have distortion performance limited by its electronics, and can't be reliably used to EQ to your own ears' physiology. Once everything is connected properly and you have calibrated the SPL levels, it is as simple as taking a sine sweep of the respective driver which will capture/calculate all the measurements you need (except for multitone which needs use of REW's signal generator and spectrum analyzer). Otherwise, it would be cool to get more comparative data on harmonic and multitone distortion for the X9000 and DCA Expanse in a lower-noise setting.