[moses667]

I frequently read about how different amplifiers change the sound of headphones in different ways... shouldn't a really great amplifier do NOTHING to the signal except increase the amplitude of the signal to make it louder? 
 
If there are changes to the signal, since this isn't a headphone producing noise to our ears, but electrical signal, shouldn't it be fairly easy to measure the input and output signal and see how closely they match? How close the amplifier comes to the ideal (assuming that is, in fact, the ideal)?

 

[mikeaj]

For a number of reasons, these days, that is not the conception a lot of people have for what a "great amplifier" should do.  But certainly if you're talking about high fidelity, that would be the working definition, I would think.
 
It should be noted that the load the amplifier is driving and the output level (well, mostly at the point of clipping and operating out of its capabilities) both influence the performance, i.e. the input/output relationship.  Also, the difference between the input and output is going to change based on what you're inputting.  So you can't measure the amplifier's behavior for every single load with every single possible input signal—even if restricting to just audio-frequency input signals—because you don't have an infinite amount of time to measure an infinite number of possible inputs.  But can you measure a reasonable number of things and end up with a good depiction of the amplifier's behavior for operating conditions of interest?  Yes, but again, some people have a hard time believing that.
 
These tests are done and results published here and there.
 
Also, it's not difficult these days to make amplifiers (particularly headphone amplifiers, given the low power output levels required) at a fairly low cost that perform very well relative to headphones / speakers and also compared to human auditory capability.  So do you call something a great amplifier if it does the job, or if it outperforms other devices in certain ways that don't make a practical difference for music playback purposes?  What about an amplifier that is well suited for some headphones but not for others (e.g. say, too noisy for sensitive IEMs)?  Other considerations such as form factor, features, aesthetics, and cost are also important to people.

 

[bigshot]

With speaker systems, amps are more important. Features like built in EQ and DSP can make a huge difference to sound quality.

 

[Strangelove424]

Quote:

With speaker systems, amps are more important. Features like built in EQ and DSP can make a huge difference to sound quality.

I have a receiver with Audyssey EQ and it's pretty amazing the difference it makes. 
 
In regards to the absolute process of amplifying a signal through an electrical chain un- distorted and as accurately as possible, that's easy to do, but alot of audiophiles (and I include myself in this massive stereotypical lumping) don't really want to hear what's actually there. Distortion can add "fullness" and "richness" etc. but alot of audiophiles won't admit that distortion is causing it, or argue that digital is too crisp and requires distortion to go back to sounding correct. Others understand that they are implementing distortion for the sake of distortion, and embrace it honestly and wholeheartedly, without trying to cover it up with jargon. It's ironic to me that some of the "sweetest" sounding amps have higher THDs and noise floors than those considered "bright" and "analytical". Using a tube amp the majority of time, I'm not one to talk, but atleast I admit I like the sound of distortion and don't cover it up with any fancy talk. But if I didn't start off with a baseline ss amp with measurement-guaranteed neutrality, I'd have no clue where to begin in this hobby, and probably would have gone insane amidst the subjective fog of distortion preferences. This has led me to believe that finding out what the sound really sounds like, objectively verified via measurements, should be priority #1, then tweaking and distorting sound to your liking should come after.  

 

[bigshot]

I prefer a clean amp to start, and then add coloration incrementally to balance it for my ears. One size fits all coloration, like with tube amps don't cut it because I can't adjust them.

 

[Lorspeaker]

the bigger, the heavier the amp the better. its macho.

 

[bigshot]

Don't forget giant glowing analogue VU meters!

 

[moses667]

Quote:

For a number of reasons, these days, that is not the conception a lot of people have for what a "great amplifier" should do.  But certainly if you're talking about high fidelity, that would be the working definition, I would think.
 
It should be noted that the load the amplifier is driving and the output level (well, mostly at the point of clipping and operating out of its capabilities) both influence the performance, i.e. the input/output relationship.  Also, the difference between the input and output is going to change based on what you're inputting.  So you can't measure the amplifier's behavior for every single load with every single possible input signal—even if restricting to just audio-frequency input signals—because you don't have an infinite amount of time to measure an infinite number of possible inputs.  But can you measure a reasonable number of things and end up with a good depiction of the amplifier's behavior for operating conditions of interest?  Yes, but again, some people have a hard time believing that.
 
These tests are done and results published here and there.
 
Also, it's not difficult these days to make amplifiers (particularly headphone amplifiers, given the low power output levels required) at a fairly low cost that perform very well relative to headphones / speakers and also compared to human auditory capability.  So do you call something a great amplifier if it does the job, or if it outperforms other devices in certain ways that don't make a practical difference for music playback purposes?  What about an amplifier that is well suited for some headphones but not for others (e.g. say, too noisy for sensitive IEMs)?  Other considerations such as form factor, features, aesthetics, and cost are also important to people.

Thanks for this answer, you've at least given me some baseline from which to consider much of the opinion going on around here. Do you personally prefer more or less signal "tuning" in an amplifier?

 

[moses667]

Quote:

I prefer a clean amp to start, and then add coloration incrementally to balance it for my ears. One size fits all coloration, like with tube amps don't cut it because I can't adjust them.

 
Quote:

But if I didn't start off with a baseline ss amp with measurement-guaranteed neutrally, I'd have no clue where to begin in this hobby, and probably would have gone insane amidst the subjective fog of distortion preferences. This has led me to believe that finding out what the sound really sounds like, objectively verified via measurements, should be priority #1, then tweaking and distorting sound to your liking should come after.  

 
I'm very interested in learning to hear something as close as possible to a neutral baseline to help me judge what I'm hearing, both with amplifiers and headphones. How would I best go about judging a clean/neutral amplifier from one less so?

 

[bigshot]

Most modern solid state amps are clean and balanced. The same with CD/DVD/Bluray players. Colored is the exception.

 

[Strangelove424]

Measurements such as noise, total harmonic distortion + noise (THD+N), and to a lesser extent crosstalk, should be close to nothing... either hundreds or thousandths of a percent or, in the case of dB ratings, below the noise floor and out of audible range. Clean/neutral SS amps are easy to come by, but specs can often be incomplete, even on expensive audiophile gear. Professional equipment is less stricken by this affliction, but that's a different discussion for a different thread. If you want the most neutral amp you can get, try to find as many hard numbers as humanly possible, and balance that out with subjective impressions of the amp. 

 

[proton007]

Quote:

I'm very interested in learning to hear something as close as possible to a neutral baseline to help me judge what I'm hearing, both with amplifiers and headphones. How would I best go about judging a clean/neutral amplifier from one less so?

 
I think you got the basics bang on. An amplifier is primarily an electrical device more than the artsy sound enhancer. And any measure of performance that applies to an electrical device applies here.
This is common knowledge, there are 4 aspects of an amplifier: frequency response, gain, noise, and distortion.
 
So an ideal amp would be: the same gain for all frequencies, with minimal noise and distortion.
 
These features can help find out the more neutral and cleaner amps out there. Generally, the better the specs, the better the amp is, but whether these differences are audible is another matter.
 
Now, the audible properties of these specs:
Frequency response : same as headphones (bass, treble, mids etc).
Gain: Amplification (volume increase).
Noise : Most commonly heard as hissing sound.
Distortion : Can be Harmonic Distortion, or Clipping.

 

[PleasantSounds]

Quote:

 
Now, the audible properties of these specs:
Frequency response : same as headphones (bass, treble, mids etc).
Gain: Amplification (volume increase).
Noise : Most commonly heard as hissing sound.
Distortion : Can be Harmonic Distortion, or Clipping.

 
I'd also add channel separation to this list, as it may impact the soundstage and detail. There are noticeable differences between amps in that respect.
 
And let's not forget about the output impedance, which is THE critical parameter for pairing amps with headphones. Low impedance headphones will sound better with a very low output impedance amplifier, otherwise their reproduction of low frequencies (where most of the power goes) is crippled. High impedance headphones may be better off with a higher output impedance amp - it will achieve similar power output through higher voltage swing.
 
Combination of both high current ability and high voltage means increasing power output, and that in effect increases harmonic distortions. To reduce harmonic distortion while keeping the power output you have to move to more expensive components and technologies (if they exist).

 

[bigshot]

What amp has a problem with channel separation? Every one I've seen has bleedthough well below the basement. And as for impedence and power vs THD, I've never seen any of that to be a problem either. As long as an amp is powerful enough to do the job, it's fine.

 

[jaddie]

Quote:

 
I'd also add channel separation to this list, as it may impact the soundstage and detail. There are noticeable differences between amps in that respect.
 
And let's not forget about the output impedance, which is THE critical parameter for pairing amps with headphones. Low impedance headphones will sound better with a very low output impedance amplifier, otherwise their reproduction of low frequencies (where most of the power goes) is crippled. High impedance headphones may be better off with a higher output impedance amp - it will achieve similar power output through higher voltage swing.
 
Combination of both high current ability and high voltage means increasing power output, and that in effect increases harmonic distortions. To reduce harmonic distortion while keeping the power output you have to move to more expensive components and technologies (if they exist).

Channel sep shouldn't be an issue at all in a well designed amp.  Of course there are a lot of not-so-well designed headphone amps out there right now, but achieving poor channel separation to the point where it influences sound stage noticeably and negatively would pretty much take a deliberate act.  Funny thing about that, there are some pretty well respected headphone amps that provide a bit of cross-feed out-of-head processing, which if on when channel sep is measured, would make it look terrible...but would sound great. 
 
Still, mostly the issue is being able to drive the load, having the right gain structure, etc.  That bit is one of the things that makes me a little crazy, some of these tin-can designs out there have no business trying to drive a 30 ohm load. 
 
Increasing power output doesn't in and of itself increase harmonic distortion, in fact more probably the reverse if you consider load drive.  Again, it's design, and actually parts cost isn't that big a factor.  The cost is in good astute engineering.