Hifiman Sundara Review (headphone)

[GaryH]

I am with you on that, however since there are no transparent EQ systems in existence yet, I argue that it can sound better , generally speaking.
I tried and I tried, eventually no EQ was the best sounding to me.
EQ helped one corner but damaged another.

In what specific ways do you think well-implemented EQ is not audibly transparent and can 'damage' the sound?

 

[Ken Tajalli]

In what specific ways do you think well-implemented EQ is not audibly transparent and can 'damage' the sound?

damage is a big word.
I have tried jRiver and Neutron (Android ), both seem to kill something in the sound, in terms of sound staging and small details.
To my ears!
I do use EQ to correct odd recordings .
in case of Sundara , I preferred no EQ most of the time.

 

[McLenin]

I don't get that mentality.
EQ is for fixing the design-shortcuts/mistakes/inadequacies of the headphones.
Yes, it can sound better with EQ, but it can sound perfectly fine without. Not even trying it without is strange to me. why would you do that?
It like buying a new car and getting it towed straight to a tuning shop to make it run better... "because I really don't want to know how my perfectly fine brand new car drives. I know it can drive more comfortable after I get it tuned." If you know you're not going to like how it drives ... buy a different car!
If you know the headphones are not going to sound as they should ... buy different ones.
EQ is for fixing stuff that are clearly wrong. It should not be the default IMHO.
If my headphones don't sound good without EQ, I don't want them. That doesn't mean I don't use EQ. But most of the time I turn it back off after a while. I haven't even set up a EQ profile for my Sundara yet. I don't feel the need. (all other headphones I own do have an EQ profile in EQ APO, which I seldom use lately.)

Fair point and the point is taken. I didn't have enough time to do proper comparison. Will try it more with EQ on and off and see which I like more. You got me on the car analogy, since I love cars, and no, I wouldn't even think about doing that...

 

[GaryH]

To my ears!

Nope. To your ears, eyes and brain. If you want to judge what you actually prefer without the influence of subconscious sighted bias, you need to do a blind comparison.

 

[Ken Tajalli]

Nope. To your ears, eyes and brain. If you want to judge what you actually prefer without the influence of subconscious sighted bias, you need to do a blind comparison.

Yeeh - that blind test again! it has been quoted so many times, and sometimes it is very valid.
Unless I listen to my music blind folded all the time, it won't work for me, perhaps it is just a prejudice, perhaps not.

On a different note:
My Sundara developed a fault and had to send it back!
Meanwhile I picked up a Dan Clark "Aeon Flow X open", got it today.
Going by memory, Sundara was better! eventhough AFXO goes deeper in bass and has a sweeter upper treble, Sundara sounded BIGGER, more grand, AFXO in comparison sounds smaller, sound stage is smaller.
I see if I can pick up a Sundara cheap on eBay.

 

[smallricey]

Yeeh - that blind test again! it has been quoted so many times, and sometimes it is very valid.
Unless I listen to my music blind folded all the time, it won't work for me, perhaps it is just a prejudice, perhaps not.

On a different note:
My Sundara developed a fault and had to send it back!
Meanwhile I picked up a Dan Clark "Aeon Flow X open", got it today.
Going by memory, Sundara was better! eventhough AFXO goes deeper in bass and has a sweeter upper treble, Sundara sounded BIGGER, more grand, AFXO in comparison sounds smaller, sound stage is smaller.
I see if I can pick up a Sundara cheap on eBay.

The point is subliminally your eyes are tricking your brain, so in your mind it does sound better.
Which is good, but in reality it doesn't. So what poor people want is to be good to their wallet too.

So for people who don't care about money and do not listen to music blind folded, you are absolutely correct.
It's pretty damn awesome to improve the sound in everyway you can, even subliminally.

 

[smallricey]

Headphones are like cheese or fine wine... the longer they are lying around the tastier they get.

Or maybe I just got old. Lost some higher freqhearing.

 

[RHO]

The point is subliminally your eyes are tricking your brain, so in your mind it does sound better.
Which is good, but in reality it doesn't.

LOL, so you decide for him which objectively different sounding system he should like more than the other?

 

[smallricey]

LOL, so you decide for him which objectively different sounding system he should like more than the other?

Did I?

 

[RHO]

Did I?

Same headphones with and without EQ. He like without EQ better. You told him he was wrong about that.
Research that lead to the Harman target used headphones (always the same one) and applied EQ (or not, or more, or less) and asked if listeners liked it more or not. They did accept the response from the participants.
If you accept that, why do you think @KenTajalli can not like his headphones more without EQ?

 

[Ken Tajalli]

At any rate, after a couple of days with a "Dan Clark Aeon Open X", I ordered another Sundara!
It arrived this morning, Sundara is much better sounding (to me).
OK the Aeon's had more bass kick and treble sparkle, but Sundara is natural sounding. AOX is plasticky!
No it is not the upper bass bloom they are famous for, the timbre is all wrong and when going gets tough (heavy orchestral) they loose the plot.
One thing for AOX driver, it can go loud, loud and not distort! and frequency extreme extensions are very good, but .....
I shall put it up on eBay soon.
If you got a Sundara, value it, and don't believe hype about "Other" expensive alternatives.
Edualized or not.

 

[GaryH]

Same headphones with and without EQ. He like without EQ better. You told him he was wrong about that.
Research that lead to the Harman target used headphones (always the same one) and applied EQ (or not, or more, or less) and asked if listeners liked it more or not. They did accept the response from the participants.
If you accept that, why do you think @KenTajalli can not like his headphones more without EQ?

The huge difference you're ignoring is the listening tests in Harman's research were controlled, level-matched, and crucially, blinded.

 

[RHO]

The huge difference you're ignoring is the listening tests in Harman's research were controlled, level-matched, and crucially, blinded.

They were blinded in a way that they always had the same headphones on (EQ'd to different levels). As does @KenTajalli (his personal Sundara). Level matched ... we have to ask him if he applied pre-gain in his EQ settings and/or changed volume levels accordingly.
So in essence they were doing the same thing. Listening to headphones with various amounts of EQ en giving subjective feedback about what they heard. They did ask for subjective discriptions, along with preferences to try to match "audiophile terms" with certain FR characteristics. (like treble peeks => more [micro]detail etc.)

 

[GaryH]

They were blinded in a way that they always had the same headphones on (EQ'd to different levels). As does @KenTajalli (his personal Sundara).

Err no...you need to be blind to the variable you're actually testing, which in this case is the EQed/unEQed frequency response. If you know which you are listening to, it is in no way a blind test of your preference of with/without EQ, and so can be influenced by innumerable subconscious cognitive biases.

 

[Ken Tajalli]

They were blinded in a way that they always had the same headphones on (EQ'd to different levels). As does @KenTajalli (his personal Sundara). Level matched ... we have to ask him if he applied pre-gain in his EQ settings and/or changed volume levels accordingly.
So in essence they were doing the same thing. Listening to headphones with various amounts of EQ en giving subjective feedback about what they heard. They did ask for subjective discriptions, along with preferences to try to match "audiophile terms" with certain FR characteristics. (like treble peeks => more [micro]detail etc.)

Here is the problem with level matching:
Say I boost midrange (1 kHz) in one instance and next time instead, I boost bass by same amount (100Hz)
If I apply a preamp of -2dB in both cases, the first setting would sound louder ! when doing an A/B comparison as our ears are more sensitive to 1kHz compared to 100Hz.
When it comes to equalization, simple preamp level matching won't work, and you always know which is which - so blind tests are out too.
I can tell when a voice has lost something after EQ, a violin looses a bit of texture - these are very minute effects, but they bother me sometimes.
Ofcourse when it comes to a recording that obviously is bass light, I will reach out for that slider, but try to keep it to absolute necessity.

 

[RHO]

Err no...you need to be blind to the variable you're actually testing, which in this case is the EQed/unEQed frequency response. If you know which you are listening to, it is in no way a blind test of your preference of with/without EQ, and so can be influenced by innumerable subconscious cognitive biases.

I see your point now.
He did indeed know when there was EQ applied or not. (or he should have followed Amir's procedure for checking EQ preferences while reviewing speakers/headphones)

Still, when I decide to use EQ or not I also do it sighted. I just switch it on and a small amount of time later I switch it off. And then I decide if I like EQ better today or not. Some days I do, some days I don't. Does it matter if it's sighted in these cases. I don't think so. It's not that I make a permanent decision. And I'm not making a conclusion that one is definitely better than the other. Just in that moment I like EQ more/less. If I would turf my decisions over time I could come to the conclusion I like to EQ more/less often than not. => As if I'm in the process of determining if I like EQ more or not. Only I never come to the conclusion, because I don't feel the need to. I enjoy both. (in essence I have 2x the amount of headphones I physically own )
That is different from what @KenTajalli did. That was more a EQ: YES/NO decision.

 

[RHO]

Here is the problem with level matching:
Say I boost midrange (1 kHz) in one instance and next time instead, I boost bass by same amount (100Hz)
If I apply a preamp of -2dB in both cases, the first setting would sound louder ! when doing an A/B comparison as our ears are more sensitive to 1kHz compared to 100Hz.
When it comes to equalization, simple preamp level matching won't work, and you always know which is which - so blind tests are out too.
I can tell when a voice has lost something after EQ, a violin looses a bit of texture - these are very minute effects, but they bother me sometimes.
Ofcourse when it comes to a recording that obviously is bass light, I will reach out for that slider, but try to keep it to absolute necessity.

You have a baseline here. the un-EQ-ed headphones with 0dB pre-gain.
I you would boost 1kHz and find it too loud, would that mean there was no deficiency in 1kHz output? If you found that it was not too loud it would mean there was a deficiency in 1kHz output in relation to the rest of the frequencies.
Now that I see how you tested it was not blind, I can see why someone could conclude the addition of 1kHz would always sound worse, even if there was a deficiency at 1kHz, in the case that that person dislikes EQ in general. If that person would not know if the EQ was on or not... the conclusion could be different.

 

[Ken Tajalli]

You have a baseline here. the un-EQ-ed headphones with 0dB pre-gain.
I you would boost 1kHz and find it too loud, would that mean there was no deficiency in 1kHz output? If you found that it was not too loud it would mean there was a deficiency in 1kHz output in relation to the rest of the frequencies.
Now that I see how you tested it was not blind, I can see why someone could conclude the addition of 1kHz would always sound worse, even if there was a deficiency at 1kHz, in the case that that person dislikes EQ in general. If that person would not know if the EQ was on or not... the conclusion could be different.

That is not what I am saying at all.
It is not a question of 1kHz being too loud.
I am trying to highlight an issue with preamp level matching for the purpose of A/B testing for EQ on/off.
If I boost midrange by 2db and then use preamp and apply -2dB in order to level match, I don't think it works, as my ears would subjectively pick up that midrange as louder (compared to no eq), even though the level at peak frequency may be the same.
Blind level-matched A/B comparison on equalized sound is nigh impossible.
I have very little against eq'ing, I do use it - but try to avoid it, not because of my prejudices, because sometimes it is not worth it.

 

[RHO]

That is not what I am saying at all.
It is not a question of 1kHz being too loud.
I am trying to highlight an issue with preamp level matching for the purpose of A/B testing for EQ on/off.
If I boost midrange by 2db and then use preamp and apply -2dB in order to level match, I don't think it works, as my ears would subjectively pick up that midrange as louder (compared to no eq), even though the level at peak frequency may be the same.
Blind level-matched A/B comparison on equalized sound is nigh impossible.
I have very little against eq'ing, I do use it - but try to avoid it, not because of my prejudices, because sometimes it is not worth it.

No, you should not apply pre-gain if you want to compare the 2.
A/B comparison between 2 different headphones is an other game. There you should choose your frequency you want to level match very carefully.
But in this case you are comparing Sundara v.s. the exact same Sundara + EQ. They are level matched.

 

[GaryH]

Here is the problem with level matching:
Say I boost midrange (1 kHz) in one instance and next time instead, I boost bass by same amount (100Hz)
If I apply a preamp of -2dB in both cases, the first setting would sound louder ! when doing an A/B comparison as our ears are more sensitive to 1kHz compared to 100Hz.
When it comes to equalization, simple preamp level matching won't work, and you always know which is which - so blind tests are out too.

That's why Harman matched overall loudnes (not simply preamp level) using the ITU-R BS 1770 model. But the more important issue is sighted listening. To make sure it's blind, you would have to create the EQ profile on your source (better get a friend to do this so you don't even know what will change), randomize whether it's on or off (e.g. mashing the on/off button with your eyes closed, or again better get a friend to help), all before knowingly listening to the headphone stock sighted, if you really want to know what your true preference is.