HIFIMAN Susvara Headphone Review

[CedarX]

I see that Hifiman claims a 6Hz-75kHz FR (!!!) for Susvara. They don’t say under what conditions, or what tolerances.

I wonder if Susvara is the result of a manufacturer chasing some extreme spec. (super-wide bandwidth for example) and losing sight of the basic, “good sound”, characteristics.
This matches some of the comments about the “incredible definition”, “spatial characteristics“, or “technicality” of Susvara: as a listener, you have to forget about the (pedestrian) “good sound” to start appreciating Susvara—for example, what it can reproduce in the 20-75KHz band that no other headphones can…

Sorry Hifiman, not for me! …and I’m the happy owner of a (EQ’d) Sundara… which BTW also has a claimed 6hz-75kHz FR according to Hifiman!

 

[Bernard23]

By 2074, we will likely have full synthesis of any music we like. You type what you want and the system will generate it for you. Then we won't have to worry about how something is "recorded." You would just as well be able to change the tonality just as easily. Using a recording will be just as quaint as black and white film to many people...

How would you capture someone playing an instrument? I doubt that musicians will become extinct in 50 years, we will always want to capture a real individual playing.

 

[oceansize]

How would you capture someone playing an instrument? I doubt that musicians will become extinct in 50 years, we will always want to capture a real individual playing.

Like Elvis? Maybe, in 50 years time, there could be a holographic representation of deceased celebrity singers/musicians (instead of using films) and playing some new AI-generated songs to entertain the masses?

 

[Bernard23]

Like Elvis? Maybe, in 50 years time, there could be a holographic representation of deceased celebrity singers/musicians (instead of using films) and playing some new AI-generated songs to entertain the masses?

That doesn't address the musician of the future present that could well be unique and not a copy of a past. I don't see how any AI can create something that can replace real people playing instruments, and the experience of seeing / hearing that.

 

[oceansize]

That doesn't address the musician of the future present that could well be unique and not a copy of a past. I don't see how any AI can create something that can replace real people playing instruments, and the experience of seeing / hearing that.

It wouldn't interest me, but you could have AI generated musicians playing things that are beyond the technical capabilities of any human - and improvising so that each performance is a unique "live" event.

*edit*

Horrible thought, isn't it?

 

[jhwalker]

That doesn't address the musician of the future present that could well be unique and not a copy of a past. I don't see how any AI can create something that can replace real people playing instruments, and the experience of seeing / hearing that.

You don't have much imagination, then. I can see it quite clearly.

Not that I necessarily LIKE it (I'm a musician myself), but I can certainly see it happening, and a LOT faster than "2074".

 

[Dimitri]

As interesting as peripheraly relevant tangents may be, at best they deserve their own thread.
At best.

 

[Bernard23]

You don't have much imagination, then. I can see it quite clearly.

Not that I necessarily LIKE it (I'm a musician myself), but I can certainly see it happening, and a LOT faster than "2074".

Last comment, but you misundstood my point. AI can create music, but it cannot "create" (replace) a human musician.

 

[amirm]

How would you capture someone playing an instrument? I doubt that musicians will become extinct in 50 years, we will always want to capture a real individual playing.

I created a new thread for this: https://www.audiosciencereview.com/forum/index.php?threads/ai-and-future-of-music-production.50933/

Let's continue there.

 

[srkbear]

Do we know the actual margins on those IEM's? I don't doubt the Hifiman would have larger margins given the price tag, to be clear.

Some people say that GS is still figuring out his ABX box to this very day.

When saying “Hifiman” in this context it’s really the Hifiman of the Susvara period, as their subsequent models have been priced remarkably sensibly. They’ve clearly maintained the outlandish price point of that “flagship” because its apocryphal legend persists (gleefully promoted by dealers posing as review sites), and for unknown reasons people continue to pay $6,000 for the thing (by the way, eight years later they’re well-overdue for a flagship refresh, and I think with recent price drops of the HE1000se, 2024 might be the year we see it).

In recent years, Hifiman has been one of the most fairly-priced players in their segment, and they continue to reduce the prices of their higher-tier models as they maintain sales. I know of no other “audiophile” headphone manufacturer that does this.

Of course it is how it should be in a “Utopian” world (pun definitely intended), because after they recoup their R&D costs and streamline production, their upfront costs should drop accordingly, right? And especially with storefronts such as Amazon, they can move a lot more units via these strategic price drops. No different than the strategies of other Chinese manufacturers such as Topping, SMSL, Loxjie, Sabaj, et al. As I’ve stated previously it’s a rare thing indeed in this hobby to get a deal paying full price, and I think that’s exactly what Hifiman is offering with the HE1000 Stealth, the Arya Organic, the Ananda Nano and the Edition XS.

 

[srkbear]

There is no such thing as "detail retrieval." This is a made up term by audiophiles for headphones which has no grounding in engineering or science of sound reproduction. It classic lay intuition terminology used by people to claim technical knowledge and value when there is none.

I’m guilty of using the term, full disclosure. I know of no other way to describe the discernible reproduction of individual instruments properly and precisely placed in the audible proscenium I can detect with certain headphones, that otherwise just get lost in the mix. It isn’t all about tuning and distortion for me, although I don’t have any intelligent way to describe how.

I’m in no way a technical expert as you know, and you probably also know my journey as an unwitting victim of marketing pseudoscience prior to joining this site. But I am a formally trained musician, and I would like to believe that I can at minimum trust my ears to some extent in terms of the quality of audio reproduction on a macro scale, although I’m far more suspicious in recent years.

Is there another term more appropriate to reflect the accuracy I’m hearing, or do you believe this is cognitive dissonance on my part as well? Sincere, genuinely curious question here please…

 

[majingotan]

For me, detail retrieval is misconstrued as it doesn't exist objectively. A more accurate subjective term would be "presentation". I prefer the default/no DSP/no EQ sonic presentation of Susvara even if its presentation is a lot less preferred by Harman research objectively

 

[plumpudding2]

I’m guilty of using the term, full disclosure. I know of no other way to describe the discernible reproduction of individual instruments properly and precisely placed in the audible proscenium I can detect with certain headphones, that otherwise just get lost in the mix. It isn’t all about tuning and distortion for me, although I don’t have any intelligent way to describe how.

I’m in no way a technical expert as you know, and you probably also know my journey as an unwitting victim of marketing pseudoscience prior to joining this site. But I am a formally trained musician, and I would like to believe that I can at minimum trust my ears to some extent in terms of the quality of audio reproduction on a macro scale, although I’m far more suspicious in recent years.

Is there another term more appropriate to reflect the accuracy I’m hearing, or do you believe this is cognitive dissonance on my part as well? Sincere, genuinely curious question here please…

I think it's a time domain phenomenon, almost all transducers can produce all the necessary frequencies but if they arrive at our ears at the same time the brain positions them together, while if there's a tiny delay the brain places them a small distance apart in the virtual room it creates. This is just my conjecture though don't take this as science.

I do remember the designer of the HD800 (famous for its uncanny layering) saying the angled earcups and ring drivers create a wavefront that closely mimics the way natural sound enters your ear or something like that, so that also hints to a similar mechanism.

I don't think we'll be able to measure any of this until we can determine sound positioning from brain scans, unfortunately.

 

[bidn]

When saying “Hifiman” in this context it’s really the Hifiman of the Susvara period, as their subsequent models have been priced remarkably sensibly. They’ve clearly maintained the outlandish price point of that “flagship” because its apocryphal legend persists (gleefully promoted by dealers posing as review sites), and for unknown reasons people continue to pay $6,000 for the thing (by the way, eight years later they’re well-overdue for a flagship refresh, and I think with recent price drops of the HE1000se, 2024 might be the year we see it).

In recent years, Hifiman has been one of the most fairly-priced players in their segment, and they continue to reduce the prices of their higher-tier models as they maintain sales. I know of no other “audiophile” headphone manufacturer that does this.

Of course it is how it should be in a “Utopian” world (pun definitely intended), because after they recoup their R&D costs and streamline production, their upfront costs should drop accordingly, right? And especially with storefronts such as Amazon, they can move a lot more units via these strategic price drops. No different than the strategies of other Chinese manufacturers such as Topping, SMSL, Loxjie, Sabaj, et al. As I’ve stated previously it’s a rare thing indeed in this hobby to get a deal paying full price, and I think that’s exactly what Hifiman is offering with the HE1000 Stealth, the Arya Organic, the Ananda Nano and the Edition XS.

I agree with you that the price of the Susvara is a rip-off. And the more so when taking into account its very poor manufacturing quality. I remember that, when I was following the Susvara thread on Head-Fi years ago, I read about several persons reporting that the earpads of their Susvara had already fallen apart within 6 months after purchasing them! This combined with the crazy price protected me from wasting my money on these headphones.

I am sorry, but, in view of their very poor material quality, I fear that even the other Hifiman models are also overpriced.
I purchased and own five Hifiman headphones, from the cheap Sundara to the more expensive Shangri La Jr (€ 4500) , and for four out of these five models the earpads fell apart within two years.
I provide more details and a few photographs about this unacceptable Hifiman quality issue in the following post of the Ananda Stealth thread:

Hifiman Ananda Stealth V2 Headphone Review

Guess nobody will confuse these with the Dan Clark Stealth, will they... why would they? They are 10x the price. Silly comparison if they did. Apples and bananas

www.audiosciencereview.com

 

[migo77]

I agree with you that the price of the Susvara is a rip-off. And the more so when taking into account its very poor manufacturing quality. I remember that, when I was following the Susvara thread on Head-Fi years ago, I read about several persons reporting that the earpads of their Susvara had already fallen apart within 6 months after purchasing them! This combined with the crazy price protected me from wasting my money on these headphones.

Yes this is true, but HFM replaced them free of charge when in warranty. Other manufacturers have lot of questions and hesitating to replace something. But yeah you have prepaid that service in initial price...

 

[Buckchester]

Okay, I'm glad you made the goal clear.

So indeed, the measurement is supposed to define the better headphones, and the IEM sounds better than the Susvara.

I appreciate the directness of the reply. I honestly disagree that Susvara sounds worse than Zero 2, and I honestly feel that most people who listen to the two will pick Susvara if they are oblivious to the price.

The reason is that the transient response (at least perceived to me) makes it sound more "real." When you listen to jazz or acoustic tracks, the percussive sounds sound more impactful and realistic.

I studied a bit of motor design in the past, and the speed of motors is real. So although the impulse response should be seen in the FR curve, especially at the treble, it's not true for headphones because the system is more complicated than a simple motor (which in itself can be a bit non-linear as well).

So you still see the treble, but that doesn't mean it has a fast impulse response.

To make it quick, you can either have a light diaphragm, strong magnets, or both. Headphones has the advantage of being able to fit more magnets in a larger space. Planars are basically linear (rather than rotary) motors, and it allows them to do this very quick response with a linear force constant vs. a voicecoil in dynamic drivers that has a non-linear force constant by design.

So these are the things that should give us the "detail" of the recordings, and should be visible in FR plots if the system is simpler and LTI.

But we see many headphones with alot of treble but do not have detail, a give away that the system is not LTI and FR curve is not sufficient.

Anyway I appreciate your response and agree to disagree on this one.

Do you even own a pair of Zero 2s? Are you judging them before listening to them?

 

[John_Siau]

Ignoring minor deviations, we see that the response is identical so there is really no need for the Benchmark AHB2 in this application. The Topping A90 is actually more capable of driving the Susvara. As noted, this is due to AHB2 not having enough voltage drive due to rather "high" impedance so SUSVARA (as compared to a speaker):

This result is entirely incorrect. There is a major error in these measurements!

The AHB2 delivers 80 Vpp on each output in stereo mode (single AHB2) and 160 Vpp in mono mode (using two AHB2 amplifiers). In contrast, the Topping A90 only delivers 49 Vpp. If you do the math, the output voltage of the AHB2 (stereo mode) is 4.26 dB higher than the balanced output of the Topping A90 (not 1 dB lower). In mono mode, the AHB2 would be 10.28 dB higher than the Topping A90.

I suspect the limitation was the input level being fed to the AHB2. Check the input level and the setting of the gain switch on the AHB2.

 

[solderdude]

This result is incorrect.....

I think the SPL measurement is correct as Amir used the Hifiman speaker adapter (HE adapter see below) which has resistors inside.
This lowers the output voltage to the headphone by 2.8dB (from 28Veff to 20Veff) and the AHB2 was not used BTL
The A90 can deliver 17V (balanced) so 1.4dB level difference only.

HE-Adapter

<p><strong style="font-size: medium;">Overview:</strong></p> <p><span style="font-size: medium;">HE-Adapter is a device to protect amplifiers and headphones when using a speaker amplifier to drive HE6se, HE5LE, and HE4 headphones.</span></p> <p><span st

store.hifiman.com

10 ohm as a an amp load and 25ohm series resistor (with the 65ohm headphone).
Those resistors may have had a hard time while testing at 114dB SPL

 

[John_Siau]

I think the SPL measurement is correct as Amir used the Hifiman speaker adapter (HE adapter see below) which has resistors inside.
This lowers the output voltage to the headphone by 2.8dB (from 28Veff to 20Veff) and the AHB2 was not used BTL
The A90 can deliver 17V (balanced) so 1.4dB level difference only.

HE-Adapter

<p><strong style="font-size: medium;">Overview:</strong></p> <p><span style="font-size: medium;">HE-Adapter is a device to protect amplifiers and headphones when using a speaker amplifier to drive HE6se, HE5LE, and HE4 headphones.</span></p> <p><span st

store.hifiman.com

10 ohm as a an amp load and 25ohm series resistor (with the 65ohm headphone).
Those resistors may have had a hard time while testing at 114dB SPL

There is no need to use the HE adapter with the AHB2. If I understand you correctly, the HE adapter was used with the AHB2, but not with the Topping A90. The low output from the AHB2 was due to the adapter and not the amplifier.

We recommend this cable (direct connection with no resistors) when driving headphones from the AHB2:

Benchmark NL4-XLR4 Headphone Adapter Cable for AHB2 Power Amplifier

NL4 to XLR4 Headphone adapter cable for use with power amplifiers. This cable provides a connection from the NL4 SpeakOn outputs on an audio power amplifier to the XLR4 input of a balanced headphone cable. This should only be used with low-sensitivity headphones that require large drive voltages.

benchmarkmedia.com

 

[IXOYE]

Sorry Hifiman, not for me! …and I’m the happy owner of a (EQ’d) Sundara… which BTW also has a claimed 6hz-75kHz FR according to Hifiman!

Sundara is the best sounding headphone from Hifiman imo.