Hifiman Sundara Review (headphone)

[Robbo99999]

serial number. Just PM hifiman. They replied to me the next day.
Serial numbers are under the pads and on the warranty card.

From Reddit:
End of 2017 - Sundara released at $500
Beginning of 2019 - Sundara dropped to $350
End of 2019 - stealth revision changing the pads and dust filter to increase durability
September 2021 - first reports of the softer 4XX cable, but many people still got them with the old cable even in 2022, not clear if it was regional or maybe based on stock
June 2022 - adjusted process to drop the impedance to 32Ω - nobody measured them yet to confirm any audible changes but Hifiman assures it sounds the same. Looks like everyone is finally getting them with the new cable
July 2022 - new packaging with included table stand
The only confirmed change in the sound happened at the end of 2019 when they changed the pads (or should I say the glue in the pads that slightly changed their shape). Not clear if there are any audible changes since the impedance drop. Probably not.

The differences between old and 2021 Sundara were not big on my test fixture:

though quite significant on KEMAR:
View attachment 261601
Much smoother ear gain on the 2020

Maybe there's some unit to unit variation that is obscuring the comparison between flat plate & GRAS. As an aside, the 2020 Sundara does look pretty good re frequency response.

 

[JanesJr1]

Maybe there's some unit to unit variation that is obscuring the comparison between flat plate & GRAS. As an aside, the 2020 Sundara does look pretty good re frequency response.

Agreed; I just set up the 2020 EQ and polished off 10 reference tracks on my Sundara without feeling any need to tweak, at least not yet. Rare!

(PS But did you mean the out-of-the-box FR for the 2020 Sundara looked good or the 2020 EQ? I meant I like the 2020 Oratory EQ.)

 

[oratory1990]

Maybe there's some unit to unit variation that is obscuring the comparison between flat plate & GRAS. As an aside, the 2020 Sundara does look pretty good re frequency response.

Above 1 kHz both the pre-2020 an the post-2020 Sundaras have about ±2 dB variation (2x standard deviation, equal to 95 % confidence intervals).

Nonetheless, I would expect the difference between the earpad revisions to also show up on a flat-plate measurement rig..

 

[solderdude]

Hmmmm. seems I will have to update the plot in my Sundara article.

 

[Robbo99999]

Agreed; I just set up the 2020 EQ and polished off 10 reference tracks on my Sundara without feeling any need to tweak, at least not yet. Rare!

(PS But did you mean the out-of-the-box FR for the 2020 Sundara looked good or the 2020 EQ? I meant I like the 2020 Oratory EQ.)

Nice, how does your EQ'd 2020 Sundara compare to your DCA Closed X EQ'd? Regarding my comment on the 2020 frequency response, I think it looks like a pretty nice stock frequency response there for the 2020 Sundara, and it also looks very EQ'able (easy to EQ accurately to the target), which is proven by how Oratory was able to EQ accurately to the target without any deviations - it looks like a real nice platform on which to EQ. (He didn't need to use any sharp filters on it).

Above 1 kHz both the pre-2020 an the post-2020 Sundaras have about ±2 dB variation (2x standard deviation, equal to 95 % confidence intervals).

Nonetheless, I would expect the difference between the earpad revisions to also show up on a flat-plate measurement rig..

That's good to know that you'd expect the differences to show up on a flat plate, I was expecting the difference to be seen there too, and it means there is still value in doing measurements on non-standard rigs as long as you are using it to show comparative differences, eg pad changes and unit to unit variation for example.

View attachment 261706

Hmmmm. seems I will have to update the plot in my Sundara article.

Thing is Solderdude, you did those measurements and that's what you got, so wouldn't the only way you could revisit it would be to measure some other units of the 2018 & 2020 Sundara - to make sure unit to unit variation (or at least "rogue outlier" units) are being removed - that's assuming you're confident that your measurement methodology was solid to arrive at the frequency response graphs that you published......because if you just measure the same units again then you "should" just get the same result again. Have you decided yet how you're going to revisit it?

 

[solderdude]

wouldn't the only way you could revisit it would be to measure some other units of the 2018 & 2020 Sundara - to make sure unit to unit variation (or at least "rogue outlier" units) are being removed

Well... if 10 people would send various Sundara in I would be game.
I do not have any Sundara at the moment nor am I going to purchase one again.

Have you decided yet how you're going to revisit it?

Why would I do that?
My measurement nicely fits with the one from Oratory.
I don't know of many others that have measured old and new versions either.
Both show about 4dB more energy between 3k and 5k and la few dB less between 1k and 2k so they are not different enough to make me revisit the Sundara's
Of course above 6kHz there will be differences.

 

[Robbo99999]

Well... if 10 people would send various Sundara in I would be game.
I do not have any Sundara at the moment nor am I going to purchase one again.



Why would I do that?
My measurement nicely fits with the one from Oratory.
I don't know of many others that have measured old and new versions either.
Both show about 4dB more energy between 3k and 5k and la few dB less between 1k and 2k so they are not different enough to make me revisit the Sundara's
Of course above 6kHz there will be differences.

You just said yourself in the post I quoted, that you need to update your plot - so what did you mean by that? I thought most people in this thread that reacted to your plot commented/insinuated that they would expect to see more difference between the two versions of the headphone in your graph - StaticV3 / Oratory. I'm not against your measurements on a flat plate, so don't get me wrong & get offended, I think your measurements can be useful, especially in this scenario to compare the differences between the two different versions of the same headphone model. Don't forget, I have the miniDSP EARS rig which is also a non-standard rig, so I know the value that these non-standard rigs like flat plates, miniDSP EARS etc can bring.......I'd rather have a GRAS (lol), but I can't afford (& wouldn't) buy one! (There is not much difference in your graph between 1-2kHz, for example, but is on the GRAS, and seem to be somewhat smaller differences elsewhere).

EDIT: you sneaky bugger (lol), you edited your post to include a different plot vs the one I replied to earlier, lol, - that's what you meant by "update"! No wonder everyone said there were no differences earlier, lol! What was your earlier posted plot from then?

 

[solderdude]

You just said yourself in the post I quoted, that you need to update your plot - so what did you mean by that?

I needed to update the plot on my website and is where I posted it from.
Had a closer look and it appeared I used a comparison plot in my 2020 article based on the wrong traces.
I noticed this when I was scrutinizing my plots after StaticV3's and Oratory's posts.
As Oratory mentioned there should be a similar difference visible on a FP and his plot.
The one shown now IS the updated one.
I just still need to update the article on my website... too busy with the new OLLO.

 

[Robbo99999]

I needed to update the plot on my website. The one shown now IS the updated one.

So how come you posted your original graph, this one?

Where's that one from, why did you change it?

P.S. I don't really appreciate your winks, it insinuates that I'm asking a stupid question.

 

[solderdude]

see post #749,

P.S. I don't really appreciate your winks, it insinuates that I'm asking a stupid question.

Misinterpretation, it was not meant as such, just that I already updated the plot.

 

[Robbo99999]

see post #749,

Ah, you used the wrong traces. I'm curious though, that seems a bit bizarre, how does that happen, which ones did you mistakenly use then in your original article & original post? (What were they representing?)

 

[solderdude]

It seems to have been 2 plots with different seatings of the same headphone (2018) were used instead of the 2020 and 2018 version and probably thought... ah well the differences do not seem that big.. and used it on the website.
Will update it one of these days with the correct one including the comparisons with other headphones and links between Sundara and Sundara 2020 articles.
Mistakes are easily made and am glad I was pointed to it otherwise I would never have known.

 

[Robbo99999]

It seems to have been 2 plots with different seatings of the same headphone (2018) were used instead of the 2020 and 2018 version and probably thought... ah well the differences do not seem that big.. and used it on the website.
Will update it one of these days with the correct one including the comparisons with other headphones and links between Sundara and Sundara 2020 articles.
Mistakes are easily made and am glad I was pointed to it otherwise I would never have known.

Fair enough then!

 

[mechcentric]

May be it's my head's shape or some other effect, changing to dekoni pads changed how sundara sound with noticeable increase in bass and reduction in brightness.
Given how popular this headphone is, If there was a dekoni pad around suitable for he series, it would be interesting to see comparable data with hifiman pali pad vs dekoni on sundara
Many thanks

dekoni sheepskin pads greatly improved bass response to my ears

 

[Nikolaos]

The recommended EQ setting for the Hifiman Sundara is the following:

The original version (before 2020)
The version with the revised earpads (after 2020)

The settings on GitHub are from AutoEQ, not from me. They just use my measurements as a source.

Hi I own Sundara after 2020. On the last right column on the EQ set what BW stands for ? Bass ???
Thanks

 

[staticV3]

Hi I own Sundara after 2020. On the last right column on the EQ set what BW stands for ? Bass ???
Thanks

BW stands for Bandwidth. A Parametric filter can be defined either with a Q-value, or with Bandwidth. Different PEQ programs witll use either one or the other.
oratory's PDFs have both, so that you're covered either way.

 

[eric tee]

After gave my last set of headphones (hd58x) to my son. I was in kind of a need for new ones. Mostly have been using speakers (Genelec G2 or 8340+7350 sub) and use of headphones have been very random for several years.

Today I got Sundara`s in mail, and I have to say that after 4 hours of listening. I am very pleasently surpriced. Bought these, after reading Amir`s review (thank You very much!). Just when I`m writing this, listening one of my "reference" tracks for low freqs`(I am not basshead), Flight of the Cosmic Hippo. Satisfying, even without EQ, which I`m going to try later.

Althought, there in some lack of "body" in sub-bass department. What I get out of these, is punchy and clean enough. For my preferences.
Resolution is very good and balance of overall tonality is on the neutral side which I prefer. Comfortwise, no issues. After few hours of listening, well spent 350 euros. Reference sound for me, after loads of different setups in several decades is pretty much "Genelec sound", if that helps to estimate my writings.

Only thing that somewhat bothers me, is the not so good QC reputation of Hifiman. In fact my first choise was dt 900 pro x, but that was out the stock, so ended up to these.

Finally, I want to sincerely thank all partitipans of this site. I have been on hifi since 80`s, as customer and music lover. Never seen so much crappy "reviews" as nowdays. Loads of sites and YouTube "reviews" with straight connection to industry, one way or an other, or other way biased.

 

[_Piotr]

The leather strap broke in my unit after using it for almost 3 years. I like them a lot and thought it would be wasteful just to buy a new pair.
Bought a cheap headband from aliexpress and somehow didn't destroy the thing in the process.

 

[funkymoneyboy]

I was looking forward to my Sundaras. They came today. Left driver didn’t work. They looked poor and cheaply made and the cable was also bad quality. I won’t be buying this brand again. Shocking quality control.

 

[Doodski]

I was looking forward to my Sundaras. They came today. Left driver didn’t work. They looked poor and cheaply made and the cable was also bad quality. I won’t be buying this brand again. Shocking quality control.

That does not sound good! Very poor quality control.