[proton007]

Hey! Thanks for the notes. 

I didn't put it in initially because I wasn't sure about the technical level, it gets complex pretty fast for beginners, but I guess I'll just put it in.

 

[obobskivich]

Hey! Thanks for the notes. 
I didn't put it in initially because I wasn't sure about what technical level to follow, but I guess I'll just put it in.

I think you did a great job keeping it from being just a huge throw-up of math, and explaining things in a clear manner. Just saw a few "typo-esque" things was all (like Z over R).

If you want some links to include, lmk, I've got a few more.

 

[proton007]

Hey, thanks for the help.

Please send me any links you think will be useful, I'd be glad to add them in. I've added a few of the (simpler) links you referred me to earlier.

 

[doublea71]

I'll keep my eyes peeled for the "Headphone/Amp Impedance Matching For Dummies" on Amazon - I think I need this to be explained to me like a 6 year old....

 

 

[firev1]

Quote:

Great guide, a few notes though:
- Impedance is Z, Resistance is R; they are different things. Source impedance is therefore Zout or Zsource.
- The damping factor section is largely marketing fluffies; for a variety of reasons including cable resistance, and the reactivity of loads (it will inherently vary with frequency). See here for more: http://www.butleraudio.com/damping1.php

I partially disagree with you, that damping factor is marketing fluff, Benchmark has proven that damping factor does make a big difference in audio in terms of distortion test(with results on AP), which has been verified by a non-commercial third party. Anyway's, its notable that your source is yet another commercial tube amp seller, though the math is correct, explanation is missing something .
 
For the cable resistance argument, I say that for speakers and headphones, its a totally different ball game. I refer to "Audio Engineering Explained", on speakers, AC transformers in the loudspeaker wipe out the damping factor's effects on loudspeaker resonance.
 
This is however, notably untrue in the case of headphones as they do not have space to fit in transformers, which brings DF back into the headphone ball game. Damping factor's effects while subtle in some headphones(notably D2000) can be extremely audible in the case of Balanced armatures, for me it was the X10s and Shure 840s driven to the same volume from the ipod as compared to headphone jack out to the O2 before comparisons.
 
 
For anyone saying I'm such a "0 ohm" proponent, I do like to drive my headphones out from my receiver from time to time but I will stick with my O2 for resolution on the bass.
 
References,
http://www.benchmarkmedia.com/discuss/feedback/newsletter/2011/12/2/0-ohm-headphone-amplifier-sonic-advantages-low-impedance-headphone-amp
nwavguy's article on headphone impedance
Audio Engineering Explained for professional recording pg. 199
Sighted and single blind tests by my ears which are highly subjective 

 

[maverickronin]

I think the "marketing fluff" part is how it used to be said they you needed a super low output impedance for a super high damping factor.  There's a point of diminishing returns in there somewhere.

 

[stv014]

Quote:

I think the "marketing fluff" part is how it used to be said they you needed a super low output impedance for a super high damping factor.  There's a point of diminishing returns in there somewhere.

 
Indeed, the Benchmark article claims 100 times or higher increases in distortion compared to a (near) zero output impedance, but it ignores that the distortion of the headphone itself at a given frequency and SPL is even much higher than what is measured at the high impedance output. It shows 0.1% distortion for the HD650 with a damping factor of ~10 at 50 Hz and 106 dB SPL, but InnerFidelity measured more than 1% at the same frequency and only 100 dB SPL. "0.5 dB improvement" in distortion is obviously less impressive

 

[firev1]

Going by the article posted by obobskivich that does not seem the case. I just pointed out the difference in the working of headphones vs speakers. To add on, for the cable argument, damping factor change maybe too low in the case of headphones(impedances are not single digit like speakers)( maybe not so for IEMs eg. XBA-4/3) and maybe accounted for in tuning of the headphones.
 
Recabling it with exotic cables with their greater lengths from multiple twist, will increase resistance and off course, may be the factor accounting for a sonic change in headphones with cables by destroying the damping factor.(my add on, I may be kidding  you).
 
EDIT: Since measurements are taken at the headphone's inputs, acoustic outputs may be dramatically(maybe not night and day but still) affected by the electrical input? 0.5 db is the calculated distortion acoustic output I assume? 
 
@stv014 art lies at the last 0.1% in audio they always say! , especially at >$1000

 

[maverickronin]

The issue with BAs is more their wild impedance swings which can cause pretty big FR changes.

 

[mikeaj]

EDIT: Since measurements are taken at the headphone's inputs, acoustic outputs may be dramatically(maybe not night and day but still) affected by the electrical input? 0.5 db is the calculated distortion acoustic output I assume? 

If I understood it, the idea (very roughly) was to compare 1% distortion of headphones versus 1% + 0.1% distortion of headphones + amp. 10 * log10( 1.1 / 1 ) = 0.41 dB, about 0.5 dB difference. It's not a big deal in the grand scheme of things.

I think you're asking if acoustic outputs of the headphones can be dramatically changed by not-so-dramatic changes in the electrical input? Answer should be "no" probably. Let's break down the input into two components: (1) the original signal and (2) the error signal containing all the distortion products that aren't in the original. Define the input as the sum of those two components. If you've got 0.1% distortion, that means the error signal is a lot smaller than the original signal. If you had perfect headphones in some sense (0% THD, but frequency response need not necessarily be flat, etc.), the response of the headphones to the input would be the sum of the response to the original plus the sum of the response to the error. So the response to the error would be much smaller than the response to the original, since the error signal has much smaller magnitude.

In practice, for real-world headphones, the response of the sum (of original + error) isn't quite the sum of the responses (to original and to error, separately), like it was in the theoretical example above. It should be mostly close though, close enough to say that small differences in electrical inputs shouldn't create very disproportionately large changes in the acoustic output.

The other question to address is whether or not the distortion of the amplifier is of a significantly different character than the distortion of the headphones. If they were of very different types, then maybe you could pick out the added distortion of the amp, even if its magnitude were smaller than the distortion of the headphones. However, the kinds of distortion added by the amplifier are mostly the same kinds of second, third, etc. harmonic and intermodulation distortion products that the headphones themselves already are producing. If the amp is producing a 2nd harmonic distortion product and the headphone producing a 2nd harmonic distortion product of the amp's (small) 2nd harmonic distortion product, this will be really really small in magnitude, below the noise floor, etc. Also, the 2nd harmonic of a 2nd harmonic is just the 4th harmonic of the original: 2 octaves above, so not a huge deal anyway. I'm not sure if this is always true, but it doesn't seem like the distortion of the amp for a good amp and the distortion of the headphones are so significantly different in character that they aren't effectively "adding" to each other. So really, it should be the larger effect of the headphones that usually dominates.

 

[obobskivich]

I partially disagree with you, that damping factor is marketing fluff, Benchmark has proven that damping factor does make a big difference in audio in terms of distortion test(with results on AP), which has been verified by a non-commercial third party. Anyway's, its notable that your source is yet another commercial tube amp seller, though the math is correct, explanation is missing something .

For the cable resistance argument, I say that for speakers and headphones, its a totally different ball game. I refer to "Audio Engineering Explained", on speakers, AC transformers in the loudspeaker wipe out the damping factor's effects on loudspeaker resonance.

This is however, notably untrue in the case of headphones as they do not have space to fit in transformers, which brings DF back into the headphone ball game. Damping factor's effects while subtle in some headphones(notably D2000) can be extremely audible in the case of Balanced armatures, for me it was the X10s and Shure 840s driven to the same volume from the ipod as compared to headphone jack out to the O2 before comparisons.


For anyone saying I'm such a "0 ohm" proponent, I do like to drive my headphones out from my receiver from time to time but I will stick with my O2 for resolution on the bass.

References,
http://www.benchmarkmedia.com/discuss/feedback/newsletter/2011/12/2/0-ohm-headphone-amplifier-sonic-advantages-low-impedance-headphone-amp
nwavguy's article on headphone impedance
Audio Engineering Explained for professional recording pg. 199
Sighted and single blind tests by my ears which are highly subjective 

My source isn't a "tube amp seller" - Butler has just re-produced some old papers in HTML (I figured it'd be easier for mobile users to view); you can get them as PDFs from Roger Rusell's website (and see the original JBL Research and Harman Research tags on them as they appeared in print in the 1960s and 1970s). They're citing Toole and Augsperger; who are more or less "authorities" in this field. There's no transformers in speakers confused; impedance is just as reactive with speakers as it is with headphones (unless you have a Zobel; MDR-F1 and MDR-MA900 are the only headphones I'm aware of with one). Again, DF is just an imagination of the marketing world. It is not the same as arguing for low Zout; Zout will influence attenuation across FR based on Z.

 

[firev1]

I was looking at the 70V speaker systems as while as ESL speakers, I apologise for any mistakes I made. I always thought that DF was the one that had to do with FR changes more so than just low Zout(since line outs ain't that low impedance either) but its just about impedance swings relative to the Zout is that right?

 

[RexAeterna]

high damping factor is from low output impedance and vice versa. usually anything above 70@8ohms is cause of too much/abuse of negative feedback. also you forget not all amps mention where that damping factor was measured. most amps only measure it at 1khz. some amps do however list damping factor of multiple frequencies like 20/50hz,1khz, and 10khz or don't even list the frequency measured at all. i don't even think headamps measure like speaker amps cause they never list the damping factor at any particular load at all and just say it's at that specific output impedance and that's it.

 

[Chris J]

Quote:

Great guide, a few notes though:
- Impedance is Z, Resistance is R; they are different things. Source impedance is therefore Zout or Zsource.
 

 
 
Actually Impedance is the algebraic sum of Resistance and Inductive & Capacitive Reactance,  so they are two inter-related things.

 

[Chris J]

Quote:

 
Indeed, the Benchmark article claims 100 times or higher increases in distortion compared to a (near) zero output impedance, but it ignores that the distortion of the headphone itself at a given frequency and SPL is even much higher than what is measured at the high impedance output. It shows 0.1% distortion for the HD650 with a damping factor of ~10 at 50 Hz and 106 dB SPL, but InnerFidelity measured more than 1% at the same frequency and only 100 dB SPL. "0.5 dB improvement" in distortion is obviously less impressive

 
The point is that the distortion from the amp drops when the output impedance drops (assuming you have a fairly reactive headphone), you want the amp to be as clean as possible, before you feed that signal into a headphone.    The headphone will just take that distortion and distort it even more.
 
Unless you are a distortion loving fan of SETs........................