Hifiman Sundara Open-Back Planar Magnetic Headphones - Official Thread

Ok now I’m really confused, as you’ve given the same thicknesses for the old and new pads - 15mm at the front and 23mm at the back. I zoomed into the first photo I posted so the pad had a diameter of 105mm (as given on the HifiMan store) and measured the size on the screen of the front and back thickness and it does seem like they are 15mm and 23mm too. The store also said they received their pads directly from Hifiman on the 20th of August, so I would have thought they would be the new pads…

I’m afraid I can’t help you there. This is something that in general is hard to tell from pictures.

The confusion continues…I asked Hifiman to send me photos of the Sundara pads they sell separately, which they say are the latest versions, and they sent me this:


Now to me this looks closer to how you described the old pads. And the pads I posted photos of earlier look closer to your ones, with the pleather reaching further over around the edges towards the face of the pad than this one here.


They also sent two other photos showing the front and back sides with a measuring tape that looks like it reads 15mm and 23mm.


The thing is, I’m no longer confident that it’s just the pads that have changed. The official word from HiFiMAN is this:

“There is no official version 2 of the HiFiMAN Sundara. Neither the drivers nor the cups have had any changes or silent revisions to them in the past calendar year. The only changes have been a cosmetic improvement to the dust cover, as well as improvements to the pad structure to ensure reliability and longevity. There have not been any changes to the pad material.”

Here’s my suspicion with all of this - note that this is just a hunch, I don’t have any more inside info, and if anyone has additional information about this by all means post it below and I’ll revise this post:

Several years ago, there was a revision to the Sundara driver - likely to improve reliability. Fang Bian assured me in February that they hadn’t done any changes to the driver in at least one calendar year. He said there may have been some structure changes to the pads - also for the sake of reliability but that’s it.

Then, you have competing reports from reviewers about the sound, and this is where this whole thing gets a bit tricky. So far, every reviewer I’ve spoken to about this who got their review unit from a dealer all got the latest revision - meaning they got the one that you would expect to get if you bought one.

Then there are some reviewers who got their units directly from China. This is what I’m unable to verify but… in some cases (not all), they measured closer to the older version. This also matches my own experience, where two units I evaluated (and measured) that came directly from China were the old version (and they also measured identically to one another so it’s not just unit variation). Contrast this with the unit I got through a dealer for my more recent review and it was clearly the newer revision, as it both sounded and measured differently.

Now I suspect that if you were to buy a Sundara directly from HiFiMAN today (or anywhere else for that matter), it would be the same revision that everyone else is getting through dealers. I can’t confirm any of this, but this all just looks like a situation where some of the review/loaner units that got sent out even in the past year were pulled from old stock. This makes a certain amount of sense, especially if it doesn’t seem like there’s much of a difference between the two revisions on paper - it just looks like the same headphone - but it may have been a bit of an oversight given some of the measurable sonic differences that do exist.

In any case, all signs indicate that they’ve long since moved to the more recent revision for dealers, but there may still have been a bit of a stock delay gap for reviewers specifically, if they’re still sending review units out from old stock. I imagine this isn’t the case anymore, since it’s been so long but I can’t be sure.

What I would advise for anyone looking at buying a Sundara is this - regardless of where you are in the world:

  1. Be careful about buying used at the moment - that’s the most likely way to get the old version
  2. Buy from a reputable dealer
  3. Buy from somewhere with a return policy

Again, I don’t think physical pad measurements are going to be as indicative as it first seemed, because any pad structure changes that have occurred may not have coincided with the older driver revision.


See I have a different hunch…that in fact the Sundara, HE4XX, HE400i and Deva all use the same (or at least very similar) drivers, and it’s the pads that are the main differentiator between them. Just take a look at Oratory1990’s measurements of them all below.

Sundara (likely original pads): https://www.dropbox.com/sh/654wxtmhlq0dhqa/AAD75QfJfKih_8QL-a1O9bESa/Hifiman%20Sundara.pdf

HE4XX: https://www.dropbox.com/sh/654wxtmhlq0dhqa/AAAB2SVYvAYTKwn44x3tcqZBa/Hifiman%20HE4XX.pdf

HE400i: https://www.dropbox.com/sh/654wxtmhlq0dhqa/AAD2c-oE5YmNovJZLa_foGLZa/Hifiman%20HE400i%20(preliminary).pdf

Deva (wired): https://www.dropbox.com/sh/654wxtmhlq0dhqa/AADPEKwIqSbIZcPdKwgHMiV1a/Hifiman%20Deva%20(wired).pdf

They all show dips and peaks in the same or very similar places, just in various amounts, and with different amounts of bass roll-off, both of which could be mostly explained by pad differences (pad seal being a major factor for bass extension of course). As you and Oratory now use the same GRAS 43AG measuring rig, I overlaid your measurement of the Sundara with the new revised pads (thin green line), with his measurement of the Deva (wired, orange line):

These are very similar frequency responses. And I don’t think it’s a coincidence that it looks like the new Sundara pads and the Deva pads are the most similar of all the HifiMan pads, both having the ‘bulge’ you were talking about (the pleather wrapping further around the edge to the ear-facing pad surface), and using the same materials. The only difference being the Deva pad’s slightly oval shaped inner ring that is not perforated like the Sundara’s, which could explain the minimal differences seen in the frequency responses above.

Now all this could easily be confirmed or denied, if either you, Oratory or someone else with a GRAS 43AG has several of these different models, swaps the pads between them and re-measures. Apart from your Sundaras (with revised pads), do you have any other HifiMan headphones on hand?

1 Like

I just reviewed the 400i (2020) and the Deva and they definitely weren’t the same drivers.


Okay this isn’t true. HiFiMAN generally aims for a similar tuning, so if all you’re looking to compare is FR then you’ll see a similarity between their lowest end models and their highest end models. This by no means indicates they’re using the same drivers. You can even tell just by looking at the differences in impedance and sensitivity.


  • Impedance: 37 Ohms
  • Sensitivity: 94 dB


  • Impedance: 18 ohms
  • Sensitivity: 93.5 dB

Moreover, they aren’t all using single-sided magnetic arrays. In fact, the Deva’s driver surprisingly uses a double-sided magnetic array (unlike the Sundara), so these are very different transducers, even if they share similarities in overall tuning. I spoke with Fang Bian about this as well, and he mentioned that while we in the digital media space key in on certain parameters, like whether or not the magnetic array is single or double sided, there are a lot more parameters that go into transducer design - like diaphragm thickness, magnet strength, adhesive material and so on. These other parameters are likely also why the Deva simply doesn’t perform as well as the Sundara, in spite of the fact that it’s using a double-sided array while the Sundara does not.

This means that you can’t expect to throw Sundara pads on a Deva and have it sound like a Sundara, even if the FR is similar. You also can’t expect an HE-400i to sound like a Sundara just by throwing Sundara pads on it.

This also kind of goes into the whole question of what else is contained in FR beyond simply tonal balance, and while I do agree that many of the things we call ‘technicalities’ are captured by FR, this is also likely to be much more fine-grained information than you see in the graph overlay you’ve provided.


Ok read ‘very similar drivers’ as ‘very similarly tuned drivers’ in my previous post. What I was getting at is I suspect the main difference between these headphones is due to the pads, especially considering the significant difference in FR you heard and measured from the Sundara with the original and revised pads. I think it would be interesting and fun to test this theory by swapping pads between them and re-measuring if you have several of the models on hand :slight_smile: In fact, due to the similar design between the Sundara and (old) HE400i, and between the Deva and (new) HE400i, you could even do an informal blind test with them as well.

A quote from Hifiman’s founder - “Everything matters.”.

If you take two identical headphones but use different glue to assemble them they will sound different.

I don’t think there’s too much information you can draw from changing pads around.

It has been a while since I listened to the Sundara and have never heard the original HE400i, so I can’t comment there.

However, between the Deva and the new HE400i, pad changes will not get close to eliminating the differences between them.

You can read my comparison here:

How do you know a pad swap won’t make a big difference unless you try? :slight_smile: @Resolve has shown with his measurements that a pad change on the Sundaras significantly changed their sound, and Oratory1990 has measurements of several headphones showing sometimes drastic changes between pads.

The pads do influence the tuning yes, but I don’t think it’s quite as simple as just swapping pads - especially across models. I have a Deva here so i’ll give that a shot if I can (this could be interesting on its own). But importantly, as it relates to the Sundara, the only way to tell would be to have an old Sundara in and then pad swap to see if it is indeed just the pads. But this I can’t confirm since I don’t have an old revision.


Thanks that would be great, I’d be very interested to see those results. Yes to confirm the effect of the revised Sundara pads a swap with the old ones would be ideal. Maybe Oratory has access to both, seeing as his job is pretty much to measure headphones.

I literally said “everything matters”. Of course swapping pads will change things.

My point was you can’t draw conclusions about internal construction and how that relates to the entire line by doing some pad swaps.

I’m not saying it won’t make a difference to FR, it almost certainly will (whether they will be similar or completely different I do not know) but it will not improve the parts were the driver doesn’t perform as well on the HE400i as on the Deva (speed, detail etc.)

I’m curious. Why do you think it is likely oratory1990 measured the original pads? The measurement is dated Dec 2019.

Because his measurements differ significantly from @Resolve’s below (raw uncompensated, using the same GRAS 43AG rig), which we know were with the new revised pads.

And they differ in the ‘right’ way too. Although not as accurate, Resolve’s original measurements using his old rig (compensated) showed the Sundaras with the original pads having more pronounced dips at 1-3kHz and 6kHz and a peak at 3-5kHz, relative to his measurements of the new unit with the revised pads, see below.



These more pronounced dips and peaks with the original pads are just what you see in Oratory1990’s measurements compared to Resolve’s of the new Sundara with revised pads using the same measuring equipment as Oratory (first graph in this post), strongly suggesting Oratory’s measurements were using the original pads. Oh and the dates on his graphs don’t really tell you anything - that’s just when he measured them, and since he mostly just measures headphones people already own and send in to him, they’re not necessarily recent purchases.

Again, we don’t know when the pad structure changes happened, and we don’t know if that’s solely responsible for the changes to the tuning. If anything, the differences here just support the initial suspicion that some reviewers got old stock units - although I’d have to ask him where he got it from. Regardless, unless one of us has them both side by side, with both pads (and/or potentially the pads that are sold separately from HiFiMAN directly), we won’t know what’s responsible for it. It’s just speculation otherwise. I’ll keep an eye out for an older unit to see if that’s doable because I’d personally really love to do this comparison and get to the bottom of it.

But at the very least, if you follow the purchase advice noted above (i.e., buying from a reputable dealer with a return policy and not through used sources), you should feel confident you’re getting the better measuring headphone.