A DAC listening test challenge. Will it happen?

[dazzerfong]

A low impedance combined with a high sensitivity will absolutely cause hiss, if you connect it to a powerful amp, regardless of the quality of the amp, even on low gain. In fact some amps now have a special IEM match switch to help prevent that from happening.

Everyone knows I’m not a fan of Bigshot, but he’s not wrong here.

I'm clarifying between low impedance vs high sensitivity. It's not low impedance, it's high sensitivity which is prone. If that were the case, the HE6 would hiss - low impedance and low sensitivity. Obviously it doesn't.

As bigshot said, it's 'semantics', but it's just one thing to point out.

Dazzle it’s just semantics. I’m looking at they why and you’re like ok’ing at the what. It’s all good. It isn’t hard to make an audibly transparent player. You have to dig deep to find one that isn’t. I think currently it would be very difficult to find anything like this.

No it's not hard to make an audibly transparent player for some situations, but it's easy to find one that isn't audibly transparent. For one, if it audibly hisses, that's no longer audibly transparent on account of the note that you can hear the hiss!

Whether we like it or not, overly sensitive BA IEMs are here to stay. Yes, there are solutions on the designer's end to address the BA sensitivity, but unfortunately it's present in a lot of them.

And for a long time cheap Android phones were made by manufacturers that didn't care. If it had a 3.5mm output that worked, that was good enough. Good audio is trivial to achieve, but for phone manufacturers it doesn't always matter. Having used tons of Android devices, that was mostly resolved. Of course, now a lot of them don't even have a HO.

Unfortunately not in my experience, and sometimes manufacturers of Android phones come up with some weird ways to overcome limitations. For example, on my old Galaxy S8, if I leave all the DSP off I can hear hiss. However, if I enable UHQ and move the EQ bar ever so slightly, hiss gone. If I don't move the EQ bar, some proprietary processing is applied.

Weird stuff, but -shrug-

 

[Kentajalli]

I'm clarifying For example, on my old Galaxy S8, if I leave all the DSP off I can hear hiss. However, if I enable UHQ and move the EQ bar ever so slightly, hiss gone. If I don't move the EQ bar, some proprietary processing is applied.

Weird stuff, but -shrug-

the reason for that ,although I am not totally familiar with S8, is the fact the a lot of software equalizers on phones, automatically reduce the gain as you push up the equalizer sliders, this is to make sure you don't overload the digital section. so the hiss goes away .
now if that is happens , it indicates the hiss is in digital domain and not the sensitivity of the earphones.
one last thing to consider. on many recordings there is a little tape hiss, you are not referring to that, are you?

 

[bigshot]

No it's not hard to make an audibly transparent player for some situations, but it's easy to find one that isn't audibly transparent. For one, if it audibly hisses, that's no longer audibly transparent on account of the note that you can hear the hiss!

Yes, that is the "what". I'm just saying that the "why" of it isn't the player itself, it's the stuff connected to it. Mismatches with IEMs, interference from telephone circuitry, etc. None of that has anything to do with the player itself. It's shoving two different things together that don't work together. I've explained this in the past. Understanding the context of the "why" of it helps to predict where problems might exist and where they definitely don't.

For instance, I don't see any reason why one would assume that a $40 Walmart DVD player would play a CD any differently than the high end Oppo blu-ray player. But for some reason, people continue to be surprised at that. They both have transparent DAC chips. They both put out line level output through RCA jacks. Neither has problems with interference or impedance/sensitivity issues. The only reason to assume that one sounds better than the other is the price... and I find that to be one of the most untrustworthy ways to judge quality.

Sure, you can make any component function incorrectly. Plug it into 220 instead of 120. Connect it to speakers with blown voice coils. Try to play it at the base of an AM radio tower. Connect massively inefficient headphones without using an amp. Put it in a microwave oven on defrost. That clearly isn't what it's designed for. But if you're talking about using equipment under the conditions it's designed for, it is very hard to find a player that isn't audibly transparent. I've never run across one myself. When was the last time you did?

 

[Hifiearspeakers]

I'm clarifying between low impedance vs high sensitivity. It's not low impedance, it's high sensitivity which is prone. If that were the case, the HE6 would hiss - low impedance and low sensitivity. Obviously it doesn't.

As bigshot said, it's 'semantics', but it's just one thing to point out.



No it's not hard to make an audibly transparent player for some situations, but it's easy to find one that isn't audibly transparent. For one, if it audibly hisses, that's no longer audibly transparent on account of the note that you can hear the hiss!

Whether we like it or not, overly sensitive BA IEMs are here to stay. Yes, there are solutions on the designer's end to address the BA sensitivity, but unfortunately it's present in a lot of them.



Unfortunately not in my experience, and sometimes manufacturers of Android phones come up with some weird ways to overcome limitations. For example, on my old Galaxy S8, if I leave all the DSP off I can hear hiss. However, if I enable UHQ and move the EQ bar ever so slightly, hiss gone. If I don't move the EQ bar, some proprietary processing is applied.

Weird stuff, but -shrug-

I disagree. The impedance is just as important as the sensitivity, due to the resistance. Your HE-6 example is actually ridiculous. First of all, you have to take into account the fact that it is a planar. Even planars with a low impedance always require more power than their dynamic counterparts to be driven properly. Yes they have a flat impedance curve, but they also require more juice than the numbers would have you believe. Second of all, the HE-6 has one of the lowest sensitivity ratings of any planar headphone of all time. It requires tons of power to be driven properly. Many have connected it to speaker taps that can put out over 100 watts at 50 ohms.

The Sennheiser 800/800S has a high sensitivity of 102 db’s. But because it also has an impedance of 300 ohms, it can easily be given several watts of power with no amplifier hiss whatsoever.

BOTH metrics matter.

 

[bigshot]

They're related certainly. Not worth arguing over. It's clear that hiss can be created by other things than the amp or DAC, especially with IEMs. That was my original point. Dotting the i and crossing the t has put us off track for the topic... It's not likely that two DACs that are current and designed adequately are going to sound different. Cheap... expensive... they should all sound the same.

 

[Kentajalli]

I'm clarifying between low impedance vs high sensitivity. It's not low impedance, it's high sensitivity which is prone. If that were the case, the HE6 would hiss - low impedance and low sensitivity. Obviously it doesn't.

As bigshot said, it's 'semantics', but it's just one thing to point out.



No it's not hard to make an audibly transparent player for some situations, but it's easy to find one that isn't audibly transparent. For one, if it audibly hisses, that's no longer audibly transparent on account of the note that you can hear the hiss!

Whether we like it or not, overly sensitive BA IEMs are here to stay. Yes, there are solutions on the designer's end to address the BA sensitivity, but unfortunately it's present in a lot of them.



Unfortunately not in my experience, and sometimes manufacturers of Android phones come up with some weird ways to overcome limitations. For example, on my old Galaxy S8, if I leave all the DSP off I can hear hiss. However, if I enable UHQ and move the EQ bar ever so slightly, hiss gone. If I don't move the EQ bar, some proprietary processing is applied.

Weird stuff, but -shrug-

Actually impedance can and does have something to do with it.
You see in a voltage amplifier (they mostly are) - if you double the impedance of the load the power output drops by a factor of two and vice versa.
So the lower the impedance of the headphone, the more power it draws from the amp, and that can reveal the hidden hiss.
On LG phones with quad-dacs , the phone monitors the impedance of the headphone as it gets connected. If it sees a low impedance load, it drops the out gain - if it detects high impedance, it increases it.
Indeed on my LG phone, I used to trick it to go into high output, the trick was easy, simply plug an extension lead into the socket, let the phone detect very high impedance (open circuit actually) and then plug my Ultimate ears into the extension lead.
Presto - I had a lot more volume.

 

[dazzerfong]

Actually impedance can and does have something to do with it.
You see in a voltage amplifier (they mostly are) - if you double the impedance of the load the power output drops by a factor of two and vice versa.
So the lower the impedance of the headphone, the more power it draws from the amp, and that can reveal the hidden hiss.
On LG phones with quad-dacs , the phone monitors the impedance of the headphone as it gets connected. If it sees a low impedance load, it drops the out gain - if it detects high impedance, it increases it.
Indeed on my LG phone, I used to trick it to go into high output, the trick was easy, simply plug an extension lead into the socket, let the phone detect very high impedance (open circuit actually) and then plug my Ultimate ears into the extension lead.
Presto - I had a lot more volume.

Somewhat correct, but your explanation is incorrect. Noise floor is something that's measured in voltage, not current. Impedance does not govern power, it governs the balance between current and voltage.

If you double the impedance, yes your theoretical power output drops by a factor of two: however, this has nothing to do with hiss.

For the best way to see if hiss would affect you, you'd use sensitivity, but rather than dB/mw, you'd use dB/V. That would be a much easier way to compare hiss sensitivity of a particular driver.

Hiss is a voltage imperfection, not a current one. Current simply 'follows' the voltage. All of this is moot anyway, because at least for IEMs having high-impedance low-sensitivity almost never is the case.

 

[Kentajalli]

Somewhat correct, but your explanation is incorrect. Noise floor is something that's measured in voltage, not current. Impedance does not govern power, it governs the balance between current and voltage.

If you double the impedance, yes your theoretical power output drops by a factor of two: however, this has nothing to do with hiss.

For the best way to see if hiss would affect you, you'd use sensitivity, but rather than dB/mw, you'd use dB/V. That would be a much easier way to compare hiss sensitivity of a particular driver.

Hiss is a voltage imperfection, not a current one. Current simply 'follows' the voltage. All of this is moot anyway, because at least for IEMs having high-impedance low-sensitivity almost never is the case.

technically it is the efficiency that matters, however impedance can play a part.
take a look at this:

This is according to 1More earphone manufacturer.
Truth is lower impedance IEM's tend to be more efficient than high impedance ones, though the relationship is not always linear.
design plays a good part.
one reason why, is the fact that lower impedance draws more power from the amp.
A low audible hiss of say 1mV into an impedance four times lower, has four times the power - all others being equal, it can be more audible.
This is why I said low impedance CAN have something to do with it! I never said it was the cause - did I?
Impedance is mixed in with an earphones sensitivity to produce audible hiss - frequency response too!

 

[dazzerfong]

technically it is the efficiency that matters, however impedance can play a part.
take a look at this:

This is according to 1More earphone manufacturer.
Truth is lower impedance IEM's tend to be more efficient than high impedance ones, though the relationship is not always linear.
design plays a good part.
one reason why, is the fact that lower impedance draws more power from the amp.
A low audible hiss of say 1mV into an impedance four times lower, has four times the power - all others being equal, it can be more audible.
This is why I said low impedance CAN have something to do with it! I never said it was the cause - did I?
Impedance is mixed in with an earphones sensitivity to produce audible hiss - frequency response too!

Don't know why you're being combative, because I never accused you of saying it was the cause.

Lower impedance does not draw more power from the amp, it draws a bigger proportion of current than voltage. Very important difference there. Audible hiss of 1 mV into an impedance 4 times lower would still be 1 mV, but you'd have 4x the current, correct: that has nothing to do with the audibility as the voltage was all that's needed to determine it's audible or not.

Low impedance cans tend to have high efficiency, thus your attribution. However, it's not the determining factor by and large, it's sensitivity.

 

[Kentajalli]

Don't know why you're being combative, because I never accused you of saying it was the cause.

Lower impedance does not draw more power from the amp, it draws a bigger proportion of current than voltage. Very important difference there. Audible hiss of 1 mV into an impedance 4 times lower would still be 1 mV, but you'd have 4x the current, correct: that has nothing to do with the audibility as the voltage was all that's needed to determine it's audible or not.

Low impedance cans tend to have high efficiency, thus your attribution. However, it's not the determining factor by and large, it's sensitivity.

I am sorry, didn't mean to sound combative.
To answer your querry,
In a voltage amplifier of reasonably low output impedance, which almost all amplifiers are, the output voltage remains the same for differing load impedance.
A low impedance load draws more current (you are correct ) and since the voltage remains the same, the power output goes up - simple ohm's law physics.
And I am sorry, you are wrong, it is not mere voltage that translates into audio energy, but power.
hence the efficiency is given as a function of power, not voltage.
But it is not as simple as that, there are many variables at play, and impedance is but one of them.
I am not a combative guy at all, argumentative sure!

 

[castleofargh]

If we did leave the same output voltage and reduced only the impedance of the load, then the increase in power would generate more noise in the circuit(heat). I think that's pretty much what @Kentajalli is trying to say and he's correct AFAIK.

I also think that if we wished to replace a headphone or IEM by another one for the sake of lowering background noise, we'd go for what Dazz has been saying. Because sensi is the most obvious way to estimate the final dB output for a noise of X volt. That would typically have way more impact than a small increase in some noises strictly caused by increased current while disregarding everything else.

Did I reconcile both points of views or did I manage to create a 3 way argument? ^_^



ps from hell: If the noise increases because of current flowing in the circuit and that signal is music, is it still noise? Did you discuss correlated noise, distortions, or just plain good ol' noise? If some noise increases in a forest of bigger noises, is there still someone to measure or hear it? What is love? What does the fox say? Can thread stay on topic for more than 1 page? So many mysteries...

 

[Kentajalli]

If we did leave the same output voltage and reduced only the impedance of the load, then the increase in power would generate more noise in the circuit(heat). I think that's pretty much what @Kentajalli is trying to say and he's correct AFAIK.

I also think that if we wished to replace a headphone or IEM by another one for the sake of lowering background noise, we'd go for what Dazz has been saying. Because sensi is the most obvious way to estimate the final dB output for a noise of X volt. That would typically have way more impact than a small increase in some noises strictly caused by increased current while disregarding everything else.

Did I reconcile both points of views or did I manage to create a 3 way argument? ^_^

No, you didn't - you just told me that my powers-of-explanation are lacking.
Let us take this back to its origin.
@bigshot said that hiss was a symptom of low impedance, to which @dazzerfong replied and said impedance has nothing to do with it, it is the sensitivity of the earphones that is the cause (if memory serves). I stated Impedance does have something to do with it.
Now the word sensitive has been used a lo, and Efficiency has been confused with it.
Technically efficiency is how loud a sound transducer, be it speakers, headphones can go for a given amount of power at a frequency of 1kHz.
For speakers it is in dB for one watt (often mistaken for per-watt), in headphones it could be in dB for 10 or 100 milliwatt.
- Earphones are reffered to as Sensitive (to my knowledge) because they can go loud for a given volume setting compared to others (may or may not be efficient).
- Hiss is not just any noise, it is the Audible part of background noise, when there is no music! so music does not come into it.
The amount of music related unwanted crap is Distortion, that is not noise nor hiss.
- So where does this hiss come from?, and why some earphones produce it profoundly when others dont ?
Hiss or upper midrange (where our ears are very sensitive) portion of background noise - is usually a hash of noise that gets generated at the output of the DAC section and/or in the preamp section. Traditionally poweramp sections are pretty much silent (unless badly designed and/or noisy powersupply).
The poweramp section amplifies this noise and outputs it into the earphones.
Noise in lab environments is measured in mV or a minus db number ( db level compared to max output of the unit).
- My assertion is that most Sensitive earphones are low impedance efficient ones - why? because the low impedance by its nature draws more current (power) from the poweramp section at a lower volume setting , so they go louder!
To the average listener that is all that matters! - stick a 300 ohm pair, and you need to pump up that volume to get the same loudness (if you are lucky).
- Now in almost all DAC/amps , the volume control is digital, this means the amount of background noise is not affected by volume control and is constant at a certain mV all the time.
When music is not playing or at very low volumes, this noise becomes significant and audible - specially if you have low impedance earphones, since the same mV of noise, causes more power output. Again stick a 300 ohm headphones (even an efficient one) and you hear nothing!
Consider the following situation:
- Headphone A has impedance of 50 ohms and a sensitivity (efficiency) of 100dB for100 milliwatt.
- Headphone B has impedance of 100 ohms and a sensitivity of 100dB for 100 milliwatt.
The amp needs to have a voltage output of 3.16V to get headphone B to produce 100 dB as opposed to just 2.23V to get headphone A to produce the same100 dB.
- At either voltage levels, the power ouput is 100 milliwatts (do the math).
- However!, the volume control on the device only affects the Voltage output. meaning for headphone A it would be at a lower volume setting (Sensitive headphone?).
- at no music, or zero volume setting, a 5mV hiss level sounds louder on headphone A compared to B.
Did I get my point across?? I wonder . . .



From the chart you can pickup four headphones of 107dB sensitivity with impedances from 24 ohm to 64 ohm.
The Sony is more hissy compared to the Stanton 2000. yet they both have the same sensitivity.
The Ortofon is going to be more Hissy than Stanton, because of its higher sensitivity eventhough they both have same impedance.
BUT it is about as hissy as the Sony! Why?
because the extra sensivity but higher impedance sort of cancel each other out compared to Sony.
gets weired ha?

 

[castleofargh]

- My assertion is that most Sensitive earphones are low impedance efficient ones - why? because the low impedance by its nature draws more current (power) from the poweramp section at a lower volume setting , so they go louder!

lol, I forgot the computer all night with a half done reply. Greta Thunberg will come murder me with an eco-friendly crowbar.

It's correct electrically to expect higher power at 1V into a lower impedance load(P=U/R), and it's somehow correct that lower impedance drivers tend to have higher sensi. But I see no reason to try and link those 2 ideas together. The actual sensi of the transducer is impacted by the diaphragm, size, number or turns of the coil, actual mass of that coil, etc. Focusing only on impedance to find a relation between current and loudness is a mistake IMO.
Sensi can very well dominate over impedance specs in determining how much current will flow into the load. it's completely possible for a very sensitive IEM, the very one that most easily picks up on background noise, to require less current than some low sensi headphone. If the IEM needs ludicrously small voltage to reach the same SPL output as the headphone, then the current flow will also be pretty small.

Also be careful when checking lists of specs for headphones/IEMs because:
-The specs aren't always very accurate(picked from websites that mistook per volt and per mW values, or just a lot of variations in the way the manufacturer measured it, or plain variability between the pairs manufactured). I've had all sorts of results when measuring stuff at home.
-The specs even when correct are at 1kHz. they say nothing about the impedance or SPL at other frequencies.

With vastly different FR in headphones and IEMs, it's easy to find hiss more or less noticeable even with closely matched sensi specs.

 

[bigshot]

It might just be me, but I don't know what the point is here. If IEMs hiss, you look at the sensitivity and impedance first to see if there is a mismatch.

 

[magicscreen]

A little help.
Buy a Sony MDR-1AM2 headphone for this.
"Great headphones have quality levels which allow you to hear the characteristics of the amplifier and digital-to-analog converter (DAC) that process your sound signals. In other words, after you’ve got a strong pair of headphones, the quality if the sound you get will begin to depend on the rest of your audio setup.
And that’s what makes Sony’s latest pair of circumaural headphones, the MDR-1AM2, stand out so well in an otherwise crowded market."
review

I forget to say it has 3Hz-100kHz frequency response. So you can hear if the DACs have different low-pass filter.