Originally Posted by DreamWarrior
I love me a good "amplifiers all sound the same" argument....
Though I must admit my bias is towards the scientific, as Bigus said, even as scientifically biased as I may be, we cannot prove a negative. We certainly, as scientists, know amplifiers measure differently. Even right down to the component level; e.g. a resistor has a tolerance to its specs. So, if we measure that exact resistor component across many different instances of the same model amplifier, it will likely measure differently. So, sure, we know there are differences between amplifiers. At a more gross level, we know that, for example, THD figures and slew rates and other measurable things are different between models. We know all sorts of things are different between two "x watts rated" amplifiers -- but we also, to date, know that in controlled
experiments where both amplifiers are operated within their linear operating range
that these differences are inaudible.
Exactly. I've measured dozens of amplifiers. With modern test equipment one can generally find repeatable and statistically significant differences between even the various channels of the same multichannel amplifiers. Two nominally identical monoblocks will measure differently from each other.
It is thus hard to escape the idea that at some point differences become so small that they are insignificant. Yet, how many audiophiles complain about audible differences of this kind? Of course any reasonable listening test will confirm that the audible differences between any two reasonable amplifier channels are moot.
That said -- even removing biases, I wonder how many of these "I hear differences" things are true.
I think that they are true in the sense that they are not intentional falsehoods, and that they are usually also true on the grounds that they are the byproducts of naive listening evaluations. IOW the listener is perceiving a difference and in many cases there really is an audible difference due to the way the evaluation is done.
IOW they are true but irrelevant to the question of inherent audible differences among different pieces of audio gear of a given kind.
Hard to drive loads, loud listening volumes at far away distances, and other factors can push some amplifiers pretty hard into their "non linear operating range."
Perhaps. More likely, inadequate level matchnig and badly done comparisons. Of course there are some ludicrous speaker loads out there - many branded with a M and a L, and sold by people (Company P) who should know better. But, there is no actual technical justification for building a speaker that is too hard for a good AVR to drive.
Of course, this is a place we wouldn't want to stray into because I'd guess that what amplifiers do when they are "on the fringe" can very quite a bit due to circuit design.
You are speculating. I have spent years designing, bench testing and building power amplifiers.
Oscillation would, I believe, increase THD and some designs would suffer more than others, no?
Oscillation often accompanies clipping. Avoid the clipping and you avoid the oscillation. Avoiding the clipping is a simple matter of having a powerful enough amplifier. If you run the numbers you will find that most home audio systems never ever see the sunny side of 50 wpc.
So, how many of these uncontrolled "studies" have truly heard differences due to the fact that two amplifiers of similar wattage spec actually had different current capabilities (which, to the best of my knowledge really is the specification that matters most, no?).
Very few, only because the uncontrolled studies are already invalidated by stronger irrelevant influences, namely level matching and time synchronization. If you really want to do a sensitive listening test you need to be able to listen to the identical same music within a second or two. How many audiophiles have the resources (a switch box) to do this and actually do it? If you want to do a sensitive listening test you need some means for matching levels withn a couple of tenths of a dB. How many audiophiles have the resources to do this and actuallly do it?
I have now invalidated 99%+ of all listening tests that anybody has read of or actually done themselves, over the entire history of audio or just the last month. ;-)
Of course, that difference was audible because one amplifier was running out of steam while the other kept on chugging. Further, its presense is probably no surprise and one which a controlled study would eliminate. But, still, one that could possibly be heard in the lisenting room of an uncontrolled experiment.
You appear to be making this far more complex than it usually is. In reality neither amplifier is usually even close to running out of steam. The whole test has already failed on the grounds of inadequate level matching and failure to actually listen to both amplifiers with the identical same music, and in quick succession.
If you want two amplifiers to sound different, simply play different musical selections or even the identical same musical selection at different times. That is what almost all audiophiles do. Therefore they pretty much universally perceive that all amplifiers sound different, and the way they compare them, they do!