[pwjazz]

I often read about how headphone X (e.g. HD 600) "scales" with better amplification. I've been trying to objectively establish whether I'm able to discern such differences either through measurement or listening tests. I've now failed to discern such differences often enough that I've concluded that this is bunk. My latest experiment was an ABX listening test which you can read about on Reddit.

 

[colonelkernel8]

Because, generally speaking, it is bunk. If the load impedance is sufficiently higher than the output impedance of the amplifier, and the amplifier can provide enough voltage to reach a target volume and enough current to sustain that voltage, you’re fine. HD600’s are relatively efficient in that they don’t need much power to operate, so current isn’t much of an issue, the question is just if there’s enough voltage to reach the volume you’re happy listening at. Higher impedance headphones don’t necessarily mean they are inefficient or require a ton of power, they might just need more voltage to reach reasonable volume and don’t suffer as much if the output impedance is higher than ideal.

 

[bigshot]

Amplification depends on impedance. If you have a mismatch, you need an amp. If you are in the sweet spot you don't. It has nothing to do with "scaling". It's all about the numbers.

 

[pwjazz]

Higher impedance headphones don’t necessarily mean they are inefficient or require a ton of power, they might just need more voltage to reach reasonable volume and don’t suffer as much if the output impedance is higher than ideal.

Yup. Compared to something like an HE-6, the HD 600 is quite efficient. In my personal stable though, I have the DT 1990, LCD2C and HF5, all of which are more efficient than the HD 600, so for me it's the least efficient headphone with which I can test

 

[tansand]

nvm

 

[ev13wt]

A really bad song, but the bassline is mixed really close up to 0dB...

If the bass hits sound distorted at a reasonably loud volume, even just a tiny bit - an amp can help. At this point you have checked impedance matching. You checked if you are overloading any opamps as well

 

[castleofargh]

I often read about how headphone X (e.g. HD 600) "scales" with better amplification. I've been trying to objectively establish whether I'm able to discern such differences either through measurement or listening tests. I've now failed to discern such differences often enough that I've concluded that this is bunk. My latest experiment was an ABX listening test which you can read about on Reddit.

kudos for you randomized switch. I wish I was informed(or just willing enough) to go and do something like that for myself. do you hear the switch? like can you tell when there was a switch compared to when there isn't one?
I ask someone to help when need to, but everybody knows it's going to be boring and possibly long, so nowadays when I ask for someone to help me with a test, they all remember urgent stuff they need to do ^_^.


about you failing to identify a clear benefit/difference. I present to you this multiple choice:

A/ lol noob, u no gloden irz
B/ your rig isn't good enough, you need great gears to benefit from a better amp
C/ ABX tests almost always result in no difference, just listen! that's when you perceive all the "real" differences.
D/ you lucky dog, you can enjoy your music without having to spill thousands of dollars in gears.
E/ stop lying! you must be one of those people with the "everything sounds the same" agenda, planning to collapse the economy by convincing people that they don't need 400 different models of everything.


/!\ not all options are certified facts

 

[colonelkernel8]

kudos for you randomized switch. I wish I was informed(or just willing enough) to go and do something like that for myself. do you hear the switch? like can you tell when there was a switch compared to when there isn't one?
I ask someone to help when need to, but everybody knows it's going to be boring and possibly long, so nowadays when I ask for someone to help me with a test, they all remember urgent stuff they need to do ^_^.


about you failing to identify a clear benefit/difference. I present to you this multiple choice:

A/ lol noob, u no gloden irz
B/ your rig isn't good enough, you need great gears to benefit from a better amp
C/ ABX tests almost always result in no difference, just listen! that's when you perceive all the "real" differences.
D/ you lucky dog, you can enjoy your music without having to spill thousands of dollars in gears.
E/ stop lying! you must be one of those people with the "everything sounds the same" agenda, planning to collapse the economy by convincing people that they don't need 400 different models of everything.


/!\ not all options are certified facts

Now I don't feel so bad for occasionally getting slightly "trolly".

 

[pwjazz]

kudos for you randomized switch. I wish I was informed(or just willing enough) to go and do something like that for myself. do you hear the switch? like can you tell when there was a switch compared to when there isn't one?
I ask someone to help when need to, but everybody knows it's going to be boring and possibly long, so nowadays when I ask for someone to help me with a test, they all remember urgent stuff they need to do ^_^.


about you failing to identify a clear benefit/difference. I present to you this multiple choice:

A/ lol noob, u no gloden irz
B/ your rig isn't good enough, you need great gears to benefit from a better amp
C/ ABX tests almost always result in no difference, just listen! that's when you perceive all the "real" differences.
D/ you lucky dog, you can enjoy your music without having to spill thousands of dollars in gears.
E/ stop lying! you must be one of those people with the "everything sounds the same" agenda, planning to collapse the economy by convincing people that they don't need 400 different models of everything.


/!\ not all options are certified facts

I'll pick D!

Regarding audibility of the switch ... The relays are externally loud. Worse, they sound different depending on the direction they're switching. There's also audible switching noise on the line due to swayback.

So, I cheat. On each change, I add an extra 50 flips of the switch within about 50ms. It's loud and annoying but it keeps me from cheating by listening to the switch.

 

[amirm]

I often read about how headphone X (e.g. HD 600) "scales" with better amplification. I've been trying to objectively establish whether I'm able to discern such differences either through measurement or listening tests. I've now failed to discern such differences often enough that I've concluded that this is bunk. My latest experiment was an ABX listening test which you can read about on Reddit.

Thanks for the effort and finding such an easy platform to build this on. I purchased the parts and will measure/give it a college try to see if I can replicate your results.

 

[71 dB]

Amplification depends on impedance. If you have a mismatch, you need an amp. If you are in the sweet spot you don't. It has nothing to do with "scaling". It's all about the numbers.

Scaling is not all about the numbers?

In my opinion the output impedance of headphone amps is a rare area of sound reproduction that hasn't been taken care of properly. We have cans of low and hing impedance. Flat impedance. Curvy impedance. High and low damping ratios. It's anarchy and there is no consensus how to do it right. I was pretty shocked myself when I realized the existence of this mess back in 2011. Before that I had been an ignoramus thinking these things have been taken care of before I was born. Not the case.

 

[amirm]

Well said. Check out this graph of testing the headphone amplifier of the RME ADI-2 DAC (one of the best DAC/headphone amps I have tested) with real headphones versus dummy (resistive) load:

There is as much as 40 dB of degradation from headphone load versus dummy load! And of course nobody specs amps with real headphones (you would be lucky if the spec had both channels driven versus one, or was at reasonable distortion).

 

[bigshot]

Could they be designing headphones to make them deliberately persnickety? I don't know myself. I wouldn't buy portable headphones that require amping. Just one more thing to lug around. I'd rather just use cans that you can plug into everything.

 

[colonelkernel8]

Well said. Check out this graph of testing the headphone amplifier of the RME ADI-2 DAC (one of the best DAC/headphone amps I have tested) with real headphones versus dummy (resistive) load:

There is as much as 40 dB of degradation from headphone load versus dummy load! And of course nobody specs amps with real headphones (you would be lucky if the spec had both channels driven versus one, or was at reasonable distortion).

Hmm. This is a bit of a head-scratcher. Are you sure what you're measuring is what you *think* you're measuring in this case? Of *course* we'd expect to see much less impedance at lower frequencies because it's effectively an inductive load (of course with some differences, the voice coil moving inside a magnet and having some mass is going to make it at system at least a few orders more complex) so there's going to be a a pretty big immediate current requirement (especially since the output impedance is so low) and the voltage drop across the load is going to shrink drastically. Of course we aren't seeing +40 dB of THD+N in terms of the actual sound output at frequencies from 10 to 500 Hz. So these plots may not be telling the whole story.

 

[amirm]

Hmm. This is a bit of a head-scratcher. Are you sure what you're measuring is what you *think* you're measuring in this case? Of *course* we'd expect to see much less impedance at lower frequencies because it's effectively an inductive load (of course with some differences, the voice coil moving inside a magnet and having some mass is going to make it at system at least a few orders more complex) so there's going to be a a pretty big immediate current requirement (especially since the output impedance is so low) and the voltage drop across the load is going to shrink drastically. Of course we aren't seeing +40 dB of THD+N in terms of the actual sound output at frequencies from 10 to 500 Hz. So these plots may not be telling the whole story.

Where did you get the "of course" part? The measurements are tapped at the input to the headphone. Where do you think that signal is going if not into the headphone?

There are two white papers on this, one from Benchmark and another from Texas Instruments. You can start reading from here: https://www.audiosciencereview.com/...ments-of-rme-adi-2-dac.2582/page-5#post-74064