This has been kind of bugging me for a while (not just to pick on you, but in general):
How does one define what transparent "sounds like"? The O2 was designed to "sound transparent", so how can one justify something "sounds more transparent" and how can one actually actually confirm this? The designer of the O2 says it's transparent, but how does he know? I have heard multiple people say a Leckerton amp sounds more detailed than the O2; does this mean it's more coloured than the O2?
The last part means nothing, unless it can be demonstrated in double blind tests. Even then, it has yet to be ascertained whether it's actually an increase of detail you'd be talking about, or just more treble (or whatever).
The reason amps can be 'transparent' (which is not the same as 'transparent sounding', which is total nonsense because nobody knows what that even means) is because amping a signal is a science. If an amp measures well, it will perform accordingly. If it measures worse, it will sound different. The problem is that some people may feel that it sounds 'better' when it technically performs worse. At that point all bets are off and even the most badly designed amp can be conceivably sold. Another problem with audio is that impressions can hardly be verified, so even the most objective listener will be exposed to his own subjective preferences when evaluating gear.
All of that does not take anything away from the fact that amping is a science and that scientifically and technically, you can build an amp that objectively performs as well as possible when it comes to sound quality. Contrary to popular belief on headfi, an amp is not able to do anything magical that goes beyond measurements. It can not magically alter the amount of detail that is heard beyond what measurements account for. It can not be made to sound warmer or colder beyond making adjustments to frequency response (which is undesirable and I don't personally know any amps that shoot for strange deviations but I'm sure they're out there) or accounting for impedance mismatches. Likewise for soundstage, etc. The mistake that is often made is that people seem to think an amp is a way to beautify the sound, while in fact all it's supposed to do is to carry out the technical task of amping the signal. It's not even able to 'improve' the sound; all it can do is deviate from good to great measurements negatively and have people experience that as an improvement. It's comparable to brushing your teeth: You can do it with a super expensive brush or with a wooden one, but at the end of the day the ultimate performance would be to get rid of all the plaque. If the wooden one can do it as well as the super expensive one, they're both equal when it comes to the end result. Such is the case with amping as well, except that it just so happens to be that amping has been figured out and can be had for cheap. That is to say, the super expensive high tech brush can be bought for about a hundred bucks in the shape of a fiio e9k.
What remains are bad amps, exoticly designed amps and super expensive amps, that have you believe some magic is going inside that somehow manages to re-invent the wheel of amping. Newsflash: if there really was some kind of added functionality that improves sound quality that you can find in a kilobuck tube audio product, it would be the number one thing to sell the product with. Just the fact that that doesn't happen, should tell you plenty.
The belief is also that anything that's proven to measure well "only cares about measurements but not about sounding good" and "is analytical and cold sounding". Just in the last month I have seen amps/DACs such as the Benchmark and O2 been accused of such multiple times, ofcourse without so much as a word on double blind testing.
TLDR amping can be transparent because there are scientific goals set for it. When the amp achieves those goals - which is a very, very realistic possiblity - the amp can be considered transparent because it does nothing besides carrying out the scientific task of amplification, with errors only beyond human hearing (which is part of its technical purpose).
Edited by SunshineReggae - 3/23/13 at 7:39pm