Hifiman Ananda Nano Headphone Review

[Ezees]

but everyone on here who hasn't heard the headphone is slamming it over distortion measurements at very high levels that seem inaudible to anyone who has heard them

This is one of the absolute dilemmas of this discussion, IMO: Those who've heard them IRL (ie: HFM's egg-shaped cans) who hear them as superb and class-leaders (again, IRL) - vs those who automatically disparage them over "high distortion" graphs, yet who haven't heard them IRL having no experience with which to judge what counts vs what doesn't count IRL. IOW, the experienced who have been informed by IRL listening - vs the rather ill-informed who have no real-world experience with which to judge them, but yet still judge them anyways without any IRL experience - while steering people away from what could be possible revelations in SQ for the bucks, LOL....

 

[Robbo99999]

That is: At ~94, 104, and 114dB volume levels - almost 100% never a volume that even people who use wanton EQ listen at - and at which most HPs would show distortions. This is the main point of my not at all exhaustive post, LOL. Also, due to HFMs' totally non-damped diaphragms and enclosures - I believe the distortion figures on the graph look way worse than they actually are IRL - partly due to the totally "free" nature of the planar drivers vs virtually all dynamic drivers' inherent damping that wouldn't show as minute of gradations as planars of a distortion graph (the last point is sort of debatable though). Even I tend to listen rather loudly (that is, I think that I do) and even at my loudest: "Effit, I wanna hear how THAT sounds...." volume level, I barely scratch the ~75dB averaged loudness threshold....and 85 to 90dB+ is my 100% cautiously afraid and/or totally unbearable limit - though some music's very short-term dynamic bursts may go beyond these not-set-in-stone values.....

Measured distortion is measured distortion, it doesn't matter what the technology is behind the drivers - it's measured distortion so there's no "I believe the distortion figures on the graph look way worse than they actually are" that you mention - it's measured distortion regardless of the technology behind the drivers, the type of technology plays no part in making it look better or worse beyond what is shown in the measurements, the distortion figures speak for themselves as an absolute.

It's possible you'd be alright if you listen at low volume levels as the lowest level that Amir measures is 94dB (although we'd need measurements below 94dB to be sure), but point is that the headphone has some weaknesses when it comes to distortion that is easily bested by other headphone models & makes.

 

[Robbo99999]

This is one of the absolute dilemmas of this discussion, IMO: Those who've heard them IRL (ie: HFM's egg-shaped cans) who hear them as superb and class-leaders (again, IRL) - vs those who automatically disparage them over "high distortion" graphs, yet who haven't heard them IRL having no experience with which to judge what counts vs what doesn't count IRL. IOW, the experienced who have been informed by IRL listening - vs the rather ill-informed who have no real-world experience with which to judge them, but yet still judge them anyways without any IRL experience - while steering people away from what could be possible revelations in SQ for the bucks, LOL....

I'm pretty sure that the reviewed headphone at $499 is in no way a revelation in Sound Quality (capitalised because you just put SQ) for the bucks that you mention. (It's not magic)

 

[Ezees]

Would you share how you measured the loudness of your listening?

An SPL meter directly in between the Ananda's, the Arya's, and the HEKv2 Stealth's drivers while playing pink noise. That gives me a directly repeatable average SPL that can directly be correlated to any musical track to then equate where on the volume dial is the same SPL. This is as quick and rough estimate that's accurate enough for this purpose, IMO - that is, without breaking out the slide rule (which IDAT is necessary in this case). Of course, I don't really listen to very much classical so that's a(nother?) point where you and I differ......

 

[Ezees]

I'm pretty sure that the reviewed headphone at $499 is in no way a revelation in Sound Quality

On the contrary - whether a $499 HP can be a revelation in SQ depends 100% on the listeners' prior listening experience, IMO. Ex: The HFM Sundara was my first foray into a proper planar HPs (Fostex T50rp Mk3 notwithstanding), and it netted me quite a revelatory listening experience - one that left an indelible reference point that other, later HPs were judged by. The same with the OG Ananda and Arya v2.....

 

[Ezees]

Measured distortion is measured distortion, it doesn't matter what the technology is behind the drivers - it's measured distortion so there's no "I believe the distortion figures on the graph look way worse than they actually are" that you mention - it's measured distortion regardless of the technology behind the drivers, the type of technology plays no part in making it look better or worse beyond what is shown in the measurements, the distortion figures speak for themselves as an absolute.

It's possible you'd be alright if you listen at low volume levels as the lowest level that Amir measures is 94dB (although we'd need measurements below 94dB to be sure), but point is that the headphone has some weaknesses when it comes to distortion that is easily bested by other headphone models & makes.

That's IF said distortion is even able to be heard, qualified, and quantified. I'd wager that outside of this specialized test meant to isolate and repeat this particular "distortion" - that for most people it won't even be heard as such. As is the experience of the supermajority of Ananda Nano users (and OG Ananda/Stealth, Arya v2/Stealth, HEKv2/Stealth/SE, and Sus users as well)......

 

[solderdude]

It's possible you'd be alright if you listen at low volume levels as the lowest level that Amir measures is 94dB (although we'd need measurements below 94dB to be sure), but point is that the headphone has some weaknesses when it comes to distortion that is easily bested by other headphone models & makes.

Agreed the Ananda's (and similar models) all have a weakness which is shown to exist at higher SPL.
Part of my 'explanation' might be the distortion measuring method. I have noticed that the distortion increase coincides with massive ringing seen in these drivers.

Futhermore... as we can see Amir also measures speakers (at 0.5m but calculated for 1m) and only shows distortion plots with SPL up to 96dB in which case many percents of distortion is often seen. This is just 2dB above the lowest measurement range (I measure at 90dB) yet the 96dB range for speakers is considered loud.

The point here is that headphones can reproduce subbass better (but need more of it because of lack of tactile feel at higher SPL) and that it is a good thing to measure at higher SPL while still being safe (for the headphone) to measure at high SPL.
With speakers you might blow some tweeters and harm small (mid)woofers as the power levels involved differ immensely at the same SPL.

Chances are most people, just like you, don't ever play loud... maybe occasionally and even then those peaks are short and loud (maybe uncomfortable) while the rest of the time the music stays below heavily distorted levels.

I guess this is what @Ezees means with 'you can't hear it as well as you can see it on plots'. If the measurements were comparable one would need to omit the 104 and 114 from the plots and introduce 84dB plots.
Of course, I am still a proponent of measuring higher up because it shows limits which is a good thing.

Notice that most headphone that aren't Harman compliant in the bass (usually open models) hardly need more than 10dB at the sub-lows in order to comply. Speakers cannot even reach those levels at those frequencies and when they do (with EQ) the distortion is massive as well often even at 86dB already.

So while a proponent of measuring headphones at 114 SPL (up to 200Hz or so) it is pointless to measure beyond 104dB up to a few kHz and above 94dB above 5kHz or so from a 'music program' perspective. The plots of 104 and 114dB up to 20kHz are 'misleading'.

Just like Amir I have listened to sweeps and can tell you with certainty that above 1kHz you really don't want to have that headphone on your head at 104dB let alone 114dB.
It is painful and you immediately throw the headphone of your head or press the stop button.
It is fine up to 150Hz or so at these high SPL levels, with fine I mean it isn't good for the hearing but you aren't feeling it as loud.

 

[Ezees]

Agreed the Ananda's (and similar models) all have a weakness which is shown to exist at higher SPL.
Part of my 'explanation' might be the distortion measuring method. I have noticed that the distortion increase coincides with massive ringing seen in these drivers.

Futhermore... as we can see Amir also measures speakers (at 0.5m but calculated for 1m) and only shows distortion plots with SPL up to 96dB in which case many percents of distortion is often seen. This is just 2dB above the lowest measurement range (I measure at 90dB) yet the 96dB range for speakers is considered loud.

The point here is that headphones can reproduce subbass better (but need more of it because of lack of tactile feel at higher SPL) and that it is a good thing to measure at higher SPL while still being safe (for the headphone) to measure at high SPL.
With speakers you might blow some tweeters and harm small (mid)woofers as the power levels involved differ immensely at the same SPL.

Chances are most people, just like you, don't ever play loud... maybe occasionally and even then those peaks are short and loud (maybe uncomfortable) while the rest of the time the music stays below heavily distorted levels.

I guess this is what @Ezees means with 'you can't hear it as well as you can see it on plots'. If the measurements were comparable one would need to omit the 104 and 114 from the plots and introduce 84dB plots.
Of course, I am still a proponent of measuring higher up because it shows limits which is a good thing.

Notice that most headphone that aren't Harman compliant in the bass (usually open models) hardly need more than 10dB at the sub-lows in order to comply. Speakers cannot even reach those levels at those frequencies and when they do (with EQ) the distortion is massive as well often even at 86dB already.

So while a proponent of measuring headphones at 114 SPL (up to 200Hz or so) it is pointless to measure beyond 104dB up to a few kHz and above 94dB above 5kHz or so from a 'music program' perspective. The plots of 104 and 114dB up to 20kHz are 'misleading'.

Just like Amir I have listened to sweeps and can tell you with certainty that above 1kHz you really don't want to have that headphone on your head at 104dB let alone 114dB.
It is painful and you immediately throw the headphone of your head or press the stop button.
It is fine up to 150Hz or so at these high SPL levels, with fine I mean it isn't good for the hearing but you aren't feeling it as loud.

Very astute - and very reasonable, IMO....

 

[Robbo99999]

On the contrary - whether a $499 HP can be a revelation in SQ depends 100% on the listeners' prior listening experience, IMO. Ex: The HFM Sundara was my first foray into a proper planar HPs (Fostex T50rp Mk3 notwithstanding), and it netted me quite a revelatory listening experience - one that left an indelible reference point that other, later HPs were judged by. The same with the OG Ananda and Arya v2.....

I see what you're saying, but I'm saying for $499 vs other headphones in this price bracket (& below) that I doubt it would be a revelation in sound quality.....and indeed you were originally talking about it as "revelations in SQ for the bucks".

That's IF said distortion is even able to be heard, qualified, and quantified. I'd wager that outside of this specialized test meant to isolate and repeat this particular "distortion" - that for most people it won't even be heard as such. As is the experience of the supermajority of Ananda Nano users (and OG Ananda/Stealth, Arya v2/Stealth, HEKv2/Stealth/SE, and Sus users as well)......

Perhaps, but we're going around in circles here in the discussion. It doesn't measure well for this variable vs other headphones.

Agreed the Ananda's (and similar models) all have a weakness which is shown to exist at higher SPL.
Part of my 'explanation' might be the distortion measuring method. I have noticed that the distortion increase coincides with massive ringing seen in these drivers.

Futhermore... as we can see Amir also measures speakers (at 0.5m but calculated for 1m) and only shows distortion plots with SPL up to 96dB in which case many percents of distortion is often seen. This is just 2dB above the lowest measurement range (I measure at 90dB) yet the 96dB range for speakers is considered loud.

The point here is that headphones can reproduce subbass better (but need more of it because of lack of tactile feel at higher SPL) and that it is a good thing to measure at higher SPL while still being safe (for the headphone) to measure at high SPL.
With speakers you might blow some tweeters and harm small (mid)woofers as the power levels involved differ immensely at the same SPL.

Chances are most people, just like you, don't ever play loud... maybe occasionally and even then those peaks are short and loud (maybe uncomfortable) while the rest of the time the music stays below heavily distorted levels.

I guess this is what @Ezees means with 'you can't hear it as well as you can see it on plots'. If the measurements were comparable one would need to omit the 104 and 114 from the plots and introduce 84dB plots.
Of course, I am still a proponent of measuring higher up because it shows limits which is a good thing.

Notice that most headphone that aren't Harman compliant in the bass (usually open models) hardly need more than 10dB at the sub-lows in order to comply. Speakers cannot even reach those levels at those frequencies and when they do (with EQ) the distortion is massive as well often even at 86dB already.

So while a proponent of measuring headphones at 114 SPL (up to 200Hz or so) it is pointless to measure beyond 104dB up to a few kHz and above 94dB above 5kHz or so from a 'music program' perspective. The plots of 104 and 114dB up to 20kHz are 'misleading'.

Just like Amir I have listened to sweeps and can tell you with certainty that above 1kHz you really don't want to have that headphone on your head at 104dB let alone 114dB.
It is painful and you immediately throw the headphone of your head or press the stop button.
It is fine up to 150Hz or so at these high SPL levels, with fine I mean it isn't good for the hearing but you aren't feeling it as loud.

I agree with a lot of that. When you compare it against Amir's speaker measurements.....yes it does make you think that it would be useful to have headphone distortion measurements also at those lower levels at which we measure speaker distortion. Yeah, I'd like to see an 86 or 84dB measurement for distortion in headphones here on ASR.

EDIT: I started a poll to see if there was even any appetite for people to see distortion measurements at lower SPL figures, eg 86dB:

Would people like to see 86dB distortion measurements in Headphone Reviews here on ASR, in order to bring it more inline with speaker reviews?

My thinking is that it'd be interesting & useful to see headphone distortion measurements at a lower SPL level in addition to what is currently done here (94dB/104dB/114dB), so I'm thinking 86dB as a proposed extra SPL for testing - and this would directly mirror the 86dB SPL that speakers are...

www.audiosciencereview.com

 

[solderdude]

Amir would need to build a soundproof chamber to test headphones in. Especially for low frequencies this is hard (and expensive) to do. When this is not done background noises will become dominant over distortion making the plots harder to explain as distortion products would be below the noise level or very close to it.

When he has one he could also make isolation measurements which is interesting.

see what you're saying, but I'm saying for $499 vs other headphones in this price bracket (& below) that I doubt it would be a revelation in sound quality.....and indeed you were originally talking about it as "revelations in SQ for the bucks".

A revelation it will not be unless coming from not so great sounding headphones. The $ 500.- price bracket is filled with competitive models. The Ananda's sure don't sound poor though.

 

[staticV3]

Amir would need to build a soundproof chamber to test headphones in. Especially for low frequencies this is hard (and expensive) to do.

One could use coherent averaging to measure THD in poor SNR environments, though I'm not sure whether the APx suite supports this.

 

[solderdude]

yep but that would likely slow down the measurement considerably.

 

[Robbo99999]

Amir would need to build a soundproof chamber to test headphones in. Especially for low frequencies this is hard (and expensive) to do. When this is not done background noises will become dominant over distortion making the plots harder to explain as distortion products would be below the noise level or very close to it.

When he has one he could also make isolation measurements which is interesting.

A revelation it will not be unless coming from not so great sounding headphones. The $ 500.- price bracket is filled with competitive models. The Ananda's sure don't sound poor though.

I've got some historical measurements of one of my K702 at 85dB, measured on miniDSP EARS, and that's in a rural type home environment so it's not loud around here, but no special isolation chambers and normal computer fan noise being the most major constant (but these are Noctua case fans spinning at their lowest rpm most of the time, you can hear it though). Yes, so following is one example, I suppose this is showing the result is below the noise floor, but at the same time it's good enough to show that there are no distortion problems at 85dB:

And in the other channel of same headphone shows slight distortion issue, so just showing that it's "easy" to measure distortion at home at 85dB:

So I'm thinking this should be doable without isolation chamber, do you agree?

 

[solderdude]

Any idea how much noise the fans of Amir's AP makes ?
It's no fun sitting next to one.
I have no idea how close that is to fixture though.

As you can see the distortion in your plots is at the noise floor level so that determines the plot.
Make a sweep with the headphones on the fixture but unplugged. Then you can see the noise floor (be it room, mic or pre-amp)
I measure at 90dB and 85dB is just 5dB below that and am already bothered by the noise floor. I have measured at 80dB as some headphones already start to show dynamic compression already. I have also measured FR at 70dB and 60dB SPL but can forget about distortion plots.

 

[Mr. Haelscheir]

Yeah. In https://www.audiosciencereview.com/...out-headphone-measurements.18451/post-1956895 (post #1,210), sitting still for three minutes straight to capture and average eight 4M length 192 kHz sample rate sweeps and hoping the measurement didn't get messed up wasn't fun, though my new in-ear mic mounts were much more "comfortable" than the previous ear-tips. Though my PC fan mind my own heartbeat for some in-ear mic seatings limits the noise floor averaging, it still seems like the coupling of my unbalanced VXLR+ adapters with my MOTU M2 incurs the main noise limitation with a hump in the midrange compared to the VXLR Pro that incurs its own third-order harmonic distortion (unless that is a property of the MOTU M2's balanced operation rather than the adapter's transformer, but the balanced loopback FFT for the same levels didn't have that noise hump). But I could still get impressive results out of 4M-length FFTs (if I am interpreting things correctly). All 84 dBA distortion sweeps I've done (except for the HiFiMans in the upper midrange) were largely noise floor limited (i.e. all should have inaudible distortion at normal to "loud" listening levels), even with the max averaging, so I acknowledge that I only get visible differences at listening extremes purely to demonstrate exceptionality. I have yet to try 84 dBA pink spectrum multi-tone which would probably push the Meze Elite's distortion products to the 4M FFT 10 dB noise floor (I think a non-distorting balanced plug-in-power adapter would have brought 4M FFT noise floor down to 0 dB above the bass, much overkill for headphone evaluation).

 

[solderdude]

Yes, with multitone and FFT you can look a little deeper in the noise floor but some of the external present things (hum, fans etc) will still be visible but can be identified as not being distortion and one could write remarks on the plot where these 'poles' come from. Of course the FR is also visible when one has enough frequencies.

 

[Ken Tajalli]

Any idea how much noise the fans of Amir's AP makes ?
It's no fun sitting next to one.
I have no idea how close that is to fixture though.

I use a laptop on battery power, set to passive cooling. with SSD, the laptop is dead silent.

 

[solderdude]

I don't think this can replace the AP Amir uses for headphone measurements though.

 

[Robbo99999]

Any idea how much noise the fans of Amir's AP makes ?
It's no fun sitting next to one.
I have no idea how close that is to fixture though.

As you can see the distortion in your plots is at the noise floor level so that determines the plot.
Make a sweep with the headphones on the fixture but unplugged. Then you can see the noise floor (be it room, mic or pre-amp)
I measure at 90dB and 85dB is just 5dB below that and am already bothered by the noise floor. I have measured at 80dB as some headphones already start to show dynamic compression already. I have also measured FR at 70dB and 60dB SPL but can forget about distortion plots.

In my screenshot the around 0.75% bass distortion is above noise floor, because it's a solid dark line. If it's faint line in REW then that means it's below noise floor, but point is that was below 0.2% in the mids (& 0.75% in the bass) so it seems that it's still possible to measure headphones to check that the distortion is still low at 85dB in the case of my graphs without any fancy noise isolation, unless I misinterpreted them. (I take your point re doing sweep with headphones unplugged to detect noise floor, but I can't bring myself to get the miniDSP EARS out of storage just now, but I understand how that test would make sense).

Praps Amir can just put the AP in styrofoam box or something just for the duration of the distortion tests, which means it wouldn't be covered up long enough to overheat the AP. Maybe somekind of stryofoam or otherwise sound insulated "igloo" (fairly large to prevent overheating) could just be put/dropped over the AP during the distortion tests. That's gotta be pretty simple to make right, doesn't even have to be pretty - could be gaffer tape & cushions or something! Sound insulated cardboard box, etc.... Or maybe even a manual override on the AP fans, just a manually inputted Kill Switch into the circuit that you flip for the short duration of the test to completely kill the fans.....just a conveniently located manual switch that could be added into the circuit. Surely the AP would not overheat with zero fan RPM for the short duration of the distortion test (might be the easiest solution actually).

 

[Ken Tajalli]

I don't think this can replace the AP Amir uses for headphone measurements though.

Perhaps not, but don't you think the AP is an overkill for headphone measurements?
DACs are one thing . . .
A good DAC/ADC and a laptop is all the electronics you need.