[jcx]

Quote:

Originally Posted by Uncle Erik /img/forum/go_quote.gif SmellyGas, the measurements you mention are usually taken with a dummy load that doesn' change value. Speakers and headphones have impedance curves, as does an amplifier. The power an amp delivers varies as the load changes. I think this is especially noticeable when listening to something as sensitive as headphones. Headphone impedance also varies greatly by brand and model, so it is difficult for one amp to cover, equally, everything that's out there. I'd think you'd find significan variation by ploting the impedance curve of the headphones against the output impedance curve of an amp. Even if two amps had identical distortion, power output, etc., they can still deliver that power very differently based on their output impedance. No, I don't have charts or stats to back that up. However, I do think that's where the difference in amplifier sound mostly comes from. Also, many tubes aren't perfectly linear, so I think that can be an audible difference, too. I've seen sveral listening tests that didn't find an audible difference between solid state amps, but those tests have always excluded tubes.

the "Carver Challenge" does indicate that output impedance tweaking is in general necessary to make 2 vastly different amps sound the same into loudspeakers with very lumpy impedance from the crossover - headphone's impedance curves are generally much smoother

Quote:

Originally Posted by jcx /img/forum/go_quote.gif there is some evidence that the main determinants of an amp's "sound" are the frequency response and output impedance - when not clipping it is usually the case that ss amps are flatter in frequency response and have lower output Z than tube amps so it is relatively easy to modify ss amps in the direction of tube amps: "the Carver Challenge" diyAudio Forums - Blind Listening Tests & Amplifiers which suggests nulling a higher power ss amp against your preferred tube amp at levels where the tube amp isn't clipping - although the ss amp's sound may still diverge from what you liked about the tube amp at higher volume levels simply due to Loudness-Curve changes in perceived frequency balance

no dummy load - Carver nulled the output of both amps with the real load - and Stereophile's Golden Eared reviewers couldn't tell the differnce between tube and ss

 

[SmellyGas]

Quote:

Originally Posted by Uncle Erik /img/forum/go_quote.gif SmellyGas, the measurements you mention are usually taken with a dummy load that doesn' change value. Speakers and headphones have impedance curves, as does an amplifier. The power an amp delivers varies as the load changes.

Many magazines report these measurements into dummy loads. However, The Audio Critic tests loudspeaker amplifiers using a variety of reactive reactive (real-world) loads, including resistive loads (1 ohm to 8 ohm), capacitative (-60, -30 deg), and inductive loads (+30, +60 deg). They call this the PowerCube, and it represents a far more stressful set of loads than you would expect from the vast majority of loudspeakers available. They find that the great majority of loudspeaker amps (typically solid state) they test are capable of driving a very wide variety of loudspeaker loads without difficulty. It is rare that the powercube reveals problems, an they report which unusual loudspeaker characteristics the amp could potentially have problems with.

So, even when resistance (down to 1 ohm), capacitative, and inductive loads are used, the relevant measurements do not change much.

Quote:

I think this is especially noticeable when listening to something as sensitive as headphones. Headphone impedance also varies greatly by brand and model, so it is difficult for one amp to cover, equally, everything that's out there.

Maybe, maybe not. Like I mentioned earlier, there is a preponderance of evidence that LOUDSPEAKER amplifiers probably sound very similar. There is very little useful evidence that HEADPHONE amplifiers do or do not sound different. All you can do is extrapolate the loudspeaker amp data and decide of headphones and headphone amps are sufficiently similar to generalize the conclusions over. One thing is for sure, since the impedance of headphones are so much higher, and the current requirements are so much lower than loudspeakers, some of the challenges with CURRENT delivery present in power amps are not a problem for headphone amps. What else makes headphones different? Well, they operate at much lower voltages and with much higher power sensitivity. This means power supply noise is amplified more...okay, well we can fix that with quieter power supplies. Headphones come in a much wider range of impedance (~64 ohms up to 600 ohms) than loudspeakers (4-8 ohms), with corresponding impedance curves. Should this be a problem? Maybe, maybe not. Again, current delivery problems are less, since the impedance is high. So really, who knows. Just look at Grado's RA-1 amp - it's basically a CMOY in a wooden box that costs 10x as much. Do you really need all the fancy circuitry?

Quote:

I'd think you'd find significan variation by ploting the impedance curve of the headphones against the output impedance curve of an amp. Even if two amps had identical distortion, power output, etc., they can still deliver that power very differently based on their output impedance.

Since headphones have such high impedance relative to loudspeakers, and designing a low-output impedance amp isn't terribly difficult (I'm talking solid-state), I don't see output impedance being a major problem relative to the headphone load. Whether or not headphones represent a particulary difficult inductive or capactitative load, i do not know - I am not an audio engineer, nor do i claim to be. However, interestingly, none of the headphone amplifier design articles I have read really talk about this, which leads me to believe that it's not a major problem. The most common problems mentioned are current availability and power supply noise, which I think are easily addressed.

Also, having listened to a variety of headphone amps with completely different topologies (from CMOY to tube) with a variety of differened headphones, all in A/B comparison, I did not appreciate major differences. Perhaps small differences would have been revealed had I listened for hours...perhaps not. None of

Quote:

No, I don't have charts or stats to back that up. However, I do think that's where the difference in amplifier sound mostly comes from. Also, many tubes aren't perfectly linear, so I think that can be an audible difference, too. I've seen sveral listening tests that didn't find an audible difference between solid state amps, but those tests have always excluded tubes.

If a tube introduces frequency response nonlinearity, then I see this as a flaw, not a benefit. Amplifiers are supposed to reproduce the low-level signal perfectly without adding anything. The less that is added (or subtracted), the better. People here seem to think that amplifiers "add" to the pleasantness of the signal. I think that is a gross misconception. A perfect amplifier does nothing but amplify, that is, add AMPLITUDE to the low-level signal, and introduce nothing else, like noise and distortion.

 

[JaZZ]

Quote:

Originally Posted by SmellyGas /img/forum/go_quote.gif A perfect amplifier does nothing but amplify, that is, add AMPLITUDE to the low-level signal, and introduce nothing else, like noise and distortion.

That's the ideal we dream of, but reality is quite different – at least for audiophiles with trained/sensitive ears. I've tried many headphone amps, and none sounds exactly like the input signal.
.

 

[SmellyGas]

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif That's the ideal we dream of, but reality is quite different – at least for audiophiles with trained/sensitive ears. I've tried many headphone amps, and none sounds exactly like the input signal. .

Most people on this forum report that they have heard differences between headphone amplifiers. However, since differences "heard" can also be attributed to differences in listening volume, imperfect "audio" memory, influences of positive expectations ("placebo effect"), and peer pressure (it is very UNcool if you cannot hear differences), it is impossible for a scientifically-oriented person to accept non-controlled, non-blinded reports from individuals posting on an anonymous internet forum as anything more than "anecdotal" evidence.

With so many published reports of people adamant that they can hear audible differences in equipment (cables, amps, tweaks, etc.), only to discover that the differences "disappear" when they are blinded, I don't think anyone can blame me if I'm a little skeptical, especially with the good data available on loudspeaker amplifiers to extrapolate from.

 

[royalcrown]

Quote:

Originally Posted by grawk /img/forum/go_quote.gif It may go without saying that we should, but that's the trick, isn't it? Instead, it should go without saying that we should be comparing real world amplifiers.

What solid-state amplifiers do not meet these criteria? Every solid-state headphone amp I've had my hands on, and every amp that I've seen measurements for (measurements abound on google and such) measure extremely well, and extremely similarly. You have a valid point when it comes to tube amplifiers, but those audible differences reveal themselves in measurements (extremely high THD, nonlinear FR, etc) and thus do not undermine the original claim.

I think the only case where this may differ is with output impedance, but a) it's easy enough to find one with a low output impedance, b) it's even easier to add output impedance if so desired, and c) these differences are measurable anyway.

 

[royalcrown]

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif That's the ideal we dream of, but reality is quite different – at least for audiophiles with trained/sensitive ears. I've tried many headphone amps, and none sounds exactly like the input signal. .

Almost every headphone amp that I've seen (including some really cheap ones) measure better than the source (redbook CD). Even if you could hear amplifier differences, the limiting factor would be the dynamic range/noise floor/etc. of the CD before anything else.

 

[SmellyGas]

Quote:

Originally Posted by royalcrown /img/forum/go_quote.gif Almost every headphone amp that I've seen (including some really cheap ones) measure better than the source (redbook CD). Even if you could hear amplifier differences, the limiting factor would be the dynamic range/noise floor/etc. of the CD before anything else.

Exactly. I mean, the gold standard test, would just be to test the headphone amplifier under load and see if the frequency response or distortion is altered by the load conditions.

Oh wait, this has been done! (I just found it) This has been done with headphone amps and 30 ohm and 330 ohm resistive loads using RMAA measurements. On a wide variety of headphone amplifiers I saw tested, they ALL had flat frequency responses and inaudible levels of distortion. This means that UNLESS the reactive load of headphones can somehow alter frequency response and distortion more than the resistive load (maybe, maybe not), I sincerely doubt there will be large audible differences among dedicated headphone amps.

NOW, someone on head-fi has shown that with non-dedicated headhpone amps (built in to a stereo receiver) paired with only certain headphones, there ARE alterations in RMAA-measured frequency response - and in this situation, YES, I bet you could hear a difference...however, we're talking not talking about the headphone jack on a Pioneer receiver, we're talking about dedicated headphone amplifiers. Food for thought.

 

[JaZZ]

Quote:

Originally Posted by Oedipus Rex /img/forum/go_quote.gif I would like to hear a skeptic point of view on electrostatic amps.

Quote:

Originally Posted by SmellyGas /img/forum/go_quote.gif ...I sincerely doubt there will be large audible differences among dedicated headphone amps.

I guess that's the response the thread opener wants to get.


Quote:

Originally Posted by royalcrown /img/forum/go_quote.gif Almost every headphone amp that I've seen (including some really cheap ones) measure better than the source (redbook CD). Even if you could hear amplifier differences, the limiting factor would be the dynamic range/noise floor/etc. of the CD before anything else.

I don't really understand your argument. An audible noise floor doesn't prevent amplifiers from sounding different, just as little as headphones. I know that most contemporary amplifiers measure virtuelly perfect, but (luckily!) that doesn't influence my perception.
.

 

[SmellyGas]

Quote:

Quote: Originally Posted by Oedipus Rex I would like to hear a skeptic point of view on electrostatic amps. Quote: Originally Posted by SmellyGas ...I sincerely doubt there will be large audible differences among dedicated headphone amps. I guess that's the response the thread opener wants to get.

That was my opinion on DYNAMIC headphone amplifiers. The OP wanted a point of view on ELECTROSTATIC amplifiers, and as I wrote previously, I felt that there was insufficient evidence available to confidently conclude that they do or do not have significant differences.


Quote:

Quote: Originally Posted by royalcrown Almost every headphone amp that I've seen (including some really cheap ones) measure better than the source (redbook CD). Even if you could hear amplifier differences, the limiting factor would be the dynamic range/noise floor/etc. of the CD before anything else. I don't really understand your argument. An audible noise floor doesn't prevent amplifiers from sounding different, just as little as headphones. I know that most contemporary amplifiers measure virtuelly perfect, but (luckily!) that doesn't influence my perception.

Actually, to be technical, the RMAA measurements I've seen on dedicated headphone amplifiers show slightly worse SNR's, in the ~90-93dB range, whereas a good quality DAC at 16-bits should do be able to do same or better. However, I don't think anyone is making an argument that noise floor or dynamic range determines sound quality, as long as it is sufficient (90dB is just fine). The argument is that if headphone amplifiers have pretty much flat frequency response output from 20-20khz and low distortion when driving headphones, how could they possibly sound different?

 

[royalcrown]

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif I don't really understand your argument. An audible noise floor doesn't prevent amplifiers from sounding different, just as little as headphones. I know that most contemporary amplifiers measure virtuelly perfect, but (luckily!) that doesn't influence my perception.

An audible noise floor does not prevent amplifiers from sounding different, but the point I'm trying to make is that the noise floor is inaudible. Even if you had someone with literally perfect hearing - that is, that the person could hear any sound in any frequency range, the person would hear the noise floor of the CD before the noise floor of the amplifier. The same goes for every aspect of headphone amplification - let's go down the chart:

Frequency response: The maximum bandwidth possible must be less than 1/2 the sampling rate - in practice, redbook brickwalls past 20hz-20khz, and so if the FR is flat from 20hz-20KHz, then it is audibly perfect in terms of frequency response. I can point out several such headphone amplifiers that do this wonderfully.

SNR/dynamic range: limited to 6.02 db per bit. In the case of redbook audio, 16 bits x 6.02 = 96.32db. The presonus HP4 achieves 95db of dynamic range, and I doubt that any music that won't render you deaf can even reach such a high dynamic range.

THD: .002% from an affordable professional headphone amp. Far, far below the audible threshold, and I'd like to see anyone distinguish between such small amounts of harmonic distortion.

Stereo crosstalk: A complete non-issue when it comes to amplifiers.

I guess the question really is: what parameter do amplifiers actually differ in?

 

[jcx]

Quote:

Originally Posted by royalcrown /img/forum/go_quote.gif ... I guess the question really is: what parameter do amplifiers actually differ in?

As I keep mentioning headphone amps in general, and particularly electrostatic amps differ in peak V,I capabilities: electrostatic headphones have such low sensitivities that published max V of many amps could lead to clipping at real world dynamic peak SPL

Once you start clipping you have violated the minimum condition that makes the "all amps sound alike" proposition remotely plausible

peak Vswing has been postulated as a primary distinguishing parameter of electrostatic amps - which still leaves the question of the lack of respect for the +/-600 V Koss E/90 amplifier - has no one really ever tried it with Stax Omega?

 

[moonboy403]

Quote:

Originally Posted by royalcrown /img/forum/go_quote.gif THD: .002% from an affordable professional headphone amp. Far, far below the audible threshold, and I'd like to see anyone distinguish between such small amounts of harmonic distortion.

A THD spec of .002% measured over what? 1khz as is the usual norm published by most manufacturer? Then what's to say that nothing drastic is happening over the audible range?

 

[JaZZ]

Quote:

Originally Posted by royalcrown /img/forum/go_quote.gif I guess the question really is: what parameter do amplifiers actually differ in?

A question I have asked myself and others repeatedly. Because they sound significantly different (to my ears), be they dynamic or electrostatic, tubes or solid-state. And (logically) they alter the input signal, not just in terms of amplitude.

If you don't believe it, switch two headphone amps in series! If you're concerned about the high load impedance, use voltage dividers.
.

 

[JaZZ]

Quote:

Originally Posted by SmellyGas /img/forum/go_quote.gif That was my opinion on DYNAMIC headphone amplifiers. The OP wanted a point of view on ELECTROSTATIC amplifiers, and as I wrote previously, I felt that there was insufficient evidence available to confidently conclude that they do or do not have significant differences. Actually, to be technical, the RMAA measurements I've seen on dedicated headphone amplifiers show slightly worse SNR's, in the ~90-93dB range, whereas a good quality DAC at 16-bits should do be able to do same or better. However, I don't think anyone is making an argument that noise floor or dynamic range determines sound quality, as long as it is sufficient (90dB is just fine). The argument is that if headphone amplifiers have pretty much flat frequency response output from 20-20khz and low distortion when driving headphones, how could they possibly sound different?

I'm not sure if what the OP needs – in the form of «skeptical» feedback – is technical specifications instead of actual experience with the corresponding amps, but it's absolutely possible.
.

 

[SmellyGas]

Quote:

Originally Posted by jcx /img/forum/go_quote.gif As I keep mentioning headphone amps in general, and particularly electrostatic amps differ in peak V,I capabilities: electrostatic headphones have such low sensitivities that published max V of many amps could lead to clipping at real world dynamic peak SPL

Can you clarify that further? According to RMAA data, the wide variety of headphone amplifiers tested did not exhibit any such clipping when driving a load, based on the low levels of distortion measured. Can you provide examples, calculations, or links to what you're referring to?

Quote:

peak Vswing has been postulated as a primary distinguishing parameter of electrostatic amps - which still leaves the question of the lack of respect for the +/-600 V Koss E/90 amplifier - has no one really ever tried it with Stax Omega?

postulated by WHOM and based on what?

Quote:

Originally Posted by moonboy403 /img/forum/go_quote.gif A THD spec of .002% measured over what? 1khz as is the usual norm published by most manufacturer? Then what's to say that nothing drastic is happening over the audible range?

RMAA data tests THD over the entire frequency range, and for headphone amps driving a load, THD stays sufficiently low throughout.

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif I'm not sure if what the OP needs – in the form of «skeptical» feedback – is technical specifications instead of actual experience with the corresponding amps, but it's absolutely possible. .

The best form of evidence would be a blinded, controlled, level-matched listening test comparing electrostats. The next best form of evidence might be many people who have done A/B unblinded level-matched comparisons. The next best form would be A/B unblinded non-levell matched comparisons. The next best form would be measurements of frequency response and distortion on two amps in order to show that they were equivalent. The absolute lowest/weakest form of evidence would consist of listeners who are unblinded, spent lots of money, compared two amps perhaps on the same sitting (perhaps not), are influenced by peer pressure, and post on an anonymous internet forum.