Hifiman Sundara Review (headphone)

[solderdude]

Only meant to notion that there's more to good sound than FR, group delay, distortion measurements, etc...

But where to draw the line between actual performance and wishful thinking, taste or personal circumstances ?

The headphone cable believers are sure they make 'a world of difference' where this cannot be shown to exist. Mains cables and interlinks and potmeters having an audible sound signature.
The same goes for 'burn-in' vs 'break-in' vs 'brain-in'. The latter cannot be measured but is the most common reason for believing in 'burn-in'.
Sure ... distortion above 500Hz at 114dB SPL is rather pointless but below a few hundred Hz it makes a lot of sense.
FR measurements make sense but seal issues are equally important for most closed headphones.
It can all be measured but takes a lot of measurements and the knowledge to understand what it means.

 

[Ezees]

I've read both of the articles from Rtings (read 'Limitations') and Tyll (inexpensive dynamic HP). My experiences have been with mostly Hifiman planars, from 400i 2020 and Sundara to a lesser degree, Ananda and Arya to greater degrees (reduced treble peaks/sharpness, instrument/vocal layering and wider soundstage, bass texture/definition). Less so with my T50rp and very little with my Monoprice HiFi DJ. may not be to everyone's liking but it was free experimentation. made sure i stayed within the return/warranty window too.

 

[Ezees]

What are those other things beyond what is measurable that make for good sound?

Soundstage width/depth/height perception, perceived detail, bass, mid, treble definition/texture, imaging, perceived instrument separation, real sounding vocals/instruments and size/scale of instruments/vocals... P.S. I have to remember that I'm on the Sundara thread and some of these descriptions may not be applicable with the Sundara as much .

 

[Zensō]

Soundstage width/depth/height perception, perceived detail, bass, mid, treble definition/texture, imaging, perceived instrument separation, real sounding vocals/instruments and size/scale of instruments/vocals... P.S. I have to remember that I'm on the Sundara thread and some of these descriptions may not be applicable with the Sundara as much .

OK, so you’re mostly talking about psychoacoustics. An interesting topic for sure, though not that useful for discussing the reality (as opposed to the personal perception) of audio devices. Unfortunately, when offering subjective impressions, we bring loads of bias to the table, which makes those impressions mostly unreliable and not of much use to others:

 

[Ezees]

OK, so you’re mostly talking about psychoacoustics. An interesting topic for sure, though not that useful for discussing the reality (as opposed to the personal perception) of audio devices. Unfortunately, when offering subjective impressions, we bring loads of bias to the table, which makes those impressions mostly unreliable and not of much use to others:

View attachment 130696

I understand cognitive bias but still try to remain open-minded and searching. Right now I have several sets of cans that I can plug in to "come back to reality" - so to speak. It's good for me to listen to a different setup to reset and re-tune my brain-ear system lest it become a state of "tunnel-hearing". Still keep going back to SMSL su-9, Gustard H20, and Arya as my best HP/setup though, offering the most palpable and richest listening experience I've heard on a HP to date. Apologies for derailing the thread away from Sundara - apologies to all.

 

[Deleted member 26571]

Planars generally need more current vs voltage and some amps lack the ability to deliver enough current (+headroom for musical peaks) under load.

Nonsense. When an amp can provide the required voltage in a load it also can deliver the required current.
Ohm's law also applies to headphones

Not nonsense. Ezees' comment is perfectly correct. He said nothing about "providing the required voltage."

 

[Ezees]

Not nonsense. Ezees' comment is perfectly correct. He said nothing about "providing the required voltage."

Yea, missed Solderdude's post from earlier. Some amps don't have a lot of current capability but have no problem with voltage - example: OTL tube amps and light SS w/ insufficient PS caps... I'm still open to other views though I've been an enthusiast for a while.

 

[aldarrin]

Have you ever played pink noise on a headphone or speaker overnight over the course of a week to note changes in tone, bass capabilities, soundstage, etc?

Yes, it didn't matter. I played this game, and I'm not playing that game again, I spent enough years in a cult (literal). If anything did actually happen, we should be rightly mad at the manufacturer for not doing this for us (if I have do do a time consuming ritual for a product several orders of magnitude over the thing I can buy at a local store and use immediately, that's complete bullshit).

Besides that, the brain is pretty cool and adaptive. I have two monitors stacked vertically that have different color profiles. I can notice it if I look for it, but otherwise my brain knows to compensate for it. I can go from my non-HDR LED TV upstairs to downstairs where the HDR OLED is and tell the difference (the OLED is better by every measurable performance metric, maybe except for longevity), but in the middle of a movie I don't notice it. I'd be beyond shocked if my ears are any different.

Parting shot: I'm not 100% certain, but suspect that the whole burn in idea is a way to buy time for your brain to compensate for a new sound device before the return period ends / gets more expensive.

 

[RHO]

Yea, missed Solderdude's post from earlier. Some amps don't have a lot of current capability but have no problem with voltage - example: OTL tube amps and light SS w/ insufficient PS caps... I'm still open to other views though I've been an enthusiast for a while.

It's a law of nature. If it can deliver the voltage it can deliver the current. There is no way around that in this universe. If it cannot deliver the current the voltage will drop, 100% of the time!

 

[solderdude]

Not nonsense. Ezees' comment is perfectly correct. He said nothing about "providing the required voltage."

Did you see the 'under load' condition ?

Planars generally need more current vs voltage and some amps lack the ability to deliver enough current (+headroom for musical peaks) under load.

I agree with his part of course some amps can deliver a high output voltage in high impedances but lack current capabilities to deliver the same voltage in a low impedance headphone because of current limiting. OTL tube amps and some higher voltage C'Moy amps are nice examples.

RME amp lacks power for sufficient headroom when EQ'ing. May get loud but will have more distortion.

The question is which RME amp we are talking about.
The ADI 2 Pro output stage can provide 1W in 16 ohm and 10V > 100 ohm.
It is anything but lacking.

The baby face pro, however, can not supply much power.

When something get's loud but distorts you are already in clipping territory which may be current or voltage limited.
With the ADI2 you really need some quite inefficient headphones to reach that.

This is the Sundara thread.. 37 ohm, 104dB/V the ADI2 can reach 121dB SPL peak. so even with EQ you can get it very loud without clipping.

 

[2M2B]

All I know is that every set of planar cans that I've owned (even t50rp) sound different (better) now than when they were fresh out of the box. Mental acclimatization taken into account.

I had the same reaction when I went from dynamic based headphones/IEM's to balanced armature one's like the ER4S/SR. After spending 5 months with the ER4SR I noticed BA's had clarity & bass quality I never got from my past dynamic stuff.

 

[Deleted member 26571]

It's a law of nature. If it can deliver the voltage it can deliver the current. There is no way around that in this universe. If it cannot deliver the current the voltage will drop, 100% of the time!

Make a graph with voltage on the horizontal axis and current on the vertical axis. Draw a line from the upper left-hand corner to the lower right-hand corner; that's the boundary of clipping for that amplifier, as a function of voltage and current. It could be a straight line, or a rectangular box, or some arbitrary curve.

Now draw a point on the graph and a line connecting the origin to the point. The slope of the line is the load admittance, and the product of voltage and current at the point is instantaneous power. For any point inside (with respect to the origin) the boundary, the amplifier will deliver the required power without clipping. For a point outside the boundary, the amplifier will deliver neither the demanded current nor voltage, but will clip at the intersection of the boundary and the line.

(Note to self: why am I explaining this?)

Every amplifier will have some absolute maximum voltage it can produce, and some absolute maximum current. They may be relatively independent, in which case your operating area boundary looks like a box. Or they may be interrelated, giving some other curve. The shape of the boundary is determined by the design of the amplifier. As a general rule, there are both current- and voltage-limiting mechanisms in play. The lower the load impedance, the steeper the load line, and the more likely the amplifier will clip due to current limiting rather than voltage limiting. The opposite is true for high-impedance loads.

I assumed solderdude objected to the latter part of Ezees' comment, but I see below that's not the case. I think his complaint is that Ezees said the RME amp lacks power. I took that to mean in the context of driving a low-impedance planar load. But, of course, it's not true in general.

My own headphone amplifier will deliver over a Watt into 100 Ohms, but one-tenth that power into 10 Ohms. The reason is that its single-ended Class A output stage is biased by a 150 mA current source. It will swing 30V peak-to-peak into 100 Ohms. As Ezees says, "It doesn't have a lot of current capability, but has no problem with voltage."

Of course, 150 mA is still a good deal of current.

I am drawn to pissing matches like a moth to a flame. I have to get back to work.

 

[Nango]

Make a graph with voltage on the horizontal axis and current on the vertical axis. Draw a line from the upper left-hand corner to the lower right-hand corner; that's the boundary of clipping for that amplifier, as a function of voltage and current. It could be a straight line, or a rectangular box, or some arbitrary curve.

Now draw a point on the graph and a line connecting the origin to the point. The slope of the line is the load impedance, and the product of voltage and current at the point is instantaneous power. For any point inside (with respect to the origin) the boundary, the amplifier will deliver the required power without clipping. For a point outside the boundary, the amplifier will deliver neither the demanded current nor voltage, but will clip at the intersection of the boundary and the line.

(Note to self: why am I explaining this?)

Every amplifier will have some absolute maximum voltage it can produce, and some absolute maximum current. They may be relatively indepenent, in which case your operating area boundary looks like a box. Or they may be interrelated, giving some other curve. The shape of the boundary is determined by the design of the amplifier. As a general rule, there are both current- and voltage-limiting mechanisms in play. The lower the load impedance, the steeper the load line, and the more likely the amplifier will clip due to current limiting rather than voltage limiting. The opposite is true for high-impedance loads.

I assumed solderdude objected to the latter part of Ezees' comment, but I see below that's not the case. I think his complaint is that Ezees said the RME amp lacks power. I took that to mean in the context of driving a low-impedance planar load. But, of course, it's not true in general.

My own headphone amplifier will deliver over a Watt into 100 Ohms, but one-tenth that power into 10 Ohms. The reason is that its single-ended Class A output stage is biased by a 150 mA current source. It will swing 30V peak-to-peak into 100 Ohms. As Ezees says, "It doesn't have a lot of current capability, but has no problem with voltage."

Of course, 150 mA is still a good deal of current.

I am drawn to pissing matches like a moth to a flame. I have to get back to work.

It is the capacity to handle peaks of current, not in absolute terms but in relative terms what makes the music sound dynamic. For that you need at least one oversized toroidal (better two of them, one each channel) able to deliver up to 3x in peaks those 150mA. An amp with no toroidal is not a good amp. The weight of an amp matters!!!!!

 

[RHO]

Make a graph with voltage on the horizontal axis and current on the vertical axis. Draw a line from the upper left-hand corner to the lower right-hand corner; that's the boundary of clipping for that amplifier, as a function of voltage and current. It could be a straight line, or a rectangular box, or some arbitrary curve.

Now draw a point on the graph and a line connecting the origin to the point. The slope of the line is the load impedance, and the product of voltage and current at the point is instantaneous power. For any point inside (with respect to the origin) the boundary, the amplifier will deliver the required power without clipping. For a point outside the boundary, the amplifier will deliver neither the demanded current nor voltage, but will clip at the intersection of the boundary and the line.

(Note to self: why am I explaining this?)

Every amplifier will have some absolute maximum voltage it can produce, and some absolute maximum current. They may be relatively indepenent, in which case your operating area boundary looks like a box. Or they may be interrelated, giving some other curve. The shape of the boundary is determined by the design of the amplifier. As a general rule, there are both current- and voltage-limiting mechanisms in play. The lower the load impedance, the steeper the load line, and the more likely the amplifier will clip due to current limiting rather than voltage limiting. The opposite is true for high-impedance loads.

I assumed solderdude objected to the latter part of Ezees' comment, but I see below that's not the case. I think his complaint is that Ezees said the RME amp lacks power. I took that to mean in the context of driving a low-impedance planar load. But, of course, it's not true in general.

My own headphone amplifier will deliver over a Watt into 100 Ohms, but one-tenth that power into 10 Ohms. The reason is that its single-ended Class A output stage is biased by a 150 mA current source. It will swing 30V peak-to-peak into 100 Ohms. As Ezees says, "It doesn't have a lot of current capability, but has no problem with voltage."

Of course, 150 mA is still a good deal of current.

I am drawn to pissing matches like a moth to a flame. I have to get back to work.

I see now that I reversed current and voltage in my first sentence.
What I ment to say is that for a certain load, when a certain voltage is applied the amount of current that will flow is a known value. If the amp cannot deliver that current, the voltage will drop. It's Ohms law. You cannot apply that same voltage over that same load and not make the current flow. If the amplifier cannot deliver the current the voltage has to change. (Drop in this case)
I see now that this was not what Ezees was trying to refute.

 

[aldarrin]

It is the capacity to handle peaks of current, not in absolute terms but in relative terms what makes the music sound dynamic. For that you need at least one oversized toroidal (better two of them, one each channel) able to deliver up to 3x in peaks those 150mA. An amp with no toroidal is not a good amp. The weight of an amp matters!!!!!

(emphasis mine)

I may be going on a limb here, but this is almost certainly not true.

 

[solderdude]

It is the capacity to handle peaks of current, not in absolute terms but in relative terms what makes the music sound dynamic. For that you need at least one oversized toroidal (better two of them, one each channel) able to deliver up to 3x in peaks those 150mA. An amp with no toroidal is not a good amp. The weight of an amp matters!!!!!

All headphone amps have regulated power supplies. 150mA is VERY little current. When a headphone amp can deliver 150mA peak it can also do this continuously. It makes no difference if a toroid, SMPS, R-core or 'normal' (E core) is used that have similar power ratings.

For speaker amps with non regulated power supplies then yes.

 

[Nango]

"instantaneous current capability" at +/- 40 Amps" ..... Folks, that's what counts!!!

https://www.tnt-audio.com/ampli/hk630_e.html

 

[solderdude]

The HK630 used via the HP out can only provide 75mA max with it's 330 ohm output R.

Via it's speaker terminals 12W (600mA) in Sundara = 130dB SPL (for a short while before it starts smoking)

 

[JoséEFS]

Hello, a question about the equalization of the sundara.
I asked in a group on another page if the equalization of these headphones was correct (using AMIR or ORATORY), but the answers were a little strange to what I have read in this forum. For example:

First answer:
Hello, I do not understand well, you bought some Sundara and you use equalizer ????
If the headphones are good, it would not be necessary, unless you do not like their profile, in that case, it would be normal to buy headphones that have a profile to your liking.
On the other hand, you cannot improve the quality with the equalizer, you only change the profile
Also by equalizing upwards, you can damage the hearing aids.

Second answer:
First you see that you are going to listen to the headphones and do you want to equalize them?
You have not even done the Burning ????
Don't you see anything weird about that?
First listen to them ... get tired of listening to them, then see if you find that there is "something" that you really do not like and then and only there you put eq where you believe or feel that something is missing.

My sundara are new, but I had the idea that a good EQ "improved" the listening experience. So am I not correct?
I'm sorry for my English.
Greetings

 

[GDK]

Hello, a question about the equalization of the sundara.
I asked in a group on another page if the equalization of these headphones was correct (using AMIR or ORATORY), but the answers were a little strange to what I have read in this forum. For example:

First answer:
Hello, I do not understand well, you bought some Sundara and you use equalizer ????
If the headphones are good, it would not be necessary, unless you do not like their profile, in that case, it would be normal to buy headphones that have a profile to your liking.
On the other hand, you cannot improve the quality with the equalizer, you only change the profile
Also by equalizing upwards, you can damage the hearing aids.

Second answer:
First you see that you are going to listen to the headphones and do you want to equalize them?
You have not even done the Burning ????
Don't you see anything weird about that?
First listen to them ... get tired of listening to them, then see if you find that there is "something" that you really do not like and then and only there you put eq where you believe or feel that something is missing.

My sundara are new, but I had the idea that a good EQ "improved" the listening experience. So am I not correct?
I'm sorry for my English.
Greetings

I use Oratory’s EQ with my Sundaras and I think it makes a huge (positive) difference. I would never listen to them without it.