Hi Keith since you are weighing in a lot on the Lyr thread I'm curious since you said the Lyr was "very much too tubey for your tastes- and yet on the Emotiva forum you say the following
" Some vendors also go beyond design results and deliberately exaggerate the "tubey-ness" of their products. The Schiit Lyr, and some of the Yaqin triode amps, come to mind." Can you clarify?
And to avoid confusion about
my point-your k
nowledge of Amplifiers is welcome! And it's a WELCOME break from a certain individual who sings the "ALL AMPS are the SAME SONG" ALL the TIME.
Let's start off with the fact that amplifiers are "simple two-dimensional devices". By that I mean that speakers, for example, have a very complex job. At a live concert you experience a wide variety of sounds, each being generated by a different type of mechanism. Vibrations from a piano string head off at right angles to the string, and bouncing off the cover, while sounds from a horn come out the front in a sort of fan pattern, and sound from a guitar comes partly from strings and partly from a vibrating box with a hole in it. In fact, no speaker has a chance in hell of reproducing all those sounds, each starting at the proper place, and each going in the proper direction. It all ricochets around the room and enters your ears a certain way; which is probably differently than it enters my differently shaped ears. So, in the end, it all becomes a guessing game about which factors there are important, and how we can do the best job of faking them, which may also be different between us, both due to differences in hearing, and due to what we consider important to each of us. In contrast, all an amplifier or recording has to do is to deliver a simple electrical voltage that varies a certain way over time. You really can describe that with a single measurement (and, for stereo, with two measurements). As far as the amplifier is concerned, there is
NO SUCH THING as depth, or sound stage, or what have you. If both channels of electrical signal are exactly correct, then the result will be correct (as correct as the original recording is anyway). If two amplifiers have different sound stages, then this
MUST be traceable directly back to the fact that the electrical signal they put out is different.
This has interesting implications, the biggest of which is that there is such a thing (at least theoretically) as a "perfect amplifier". And, following this to its logical conclusion, if you had more than one perfect amplifier they would be (and sound)
EXACTLY THE SAME. Excluding differences in recordings and speakers, if both amplifier deliver the same identical perfect signal, then they will and must sound identical. This means that a perfect solid state amplifier, a perfect tube amplifier, and a perfect magic amplifier that runs on unicorn pee will all sound the same. Period. End. This is pure logic.
Therefore, logic tells us that, when certain amplifiers sound "very different", they must be producing a different electrical output. Now, in the early days of tubes, the goal of high fidelity was to eliminate all that variation and produce that one perfect output. And, if you had asked an early designer of either tube or solid state equipment their goal, they would have told you that both should sound totally clean and uncolored, and so
BOTH SHOULD SOUND THE SAME. Let me rephrase that: If a tube amp and a solid state amp don't sound identical, then one or the other (or both) is flawed. When tube designs were current technology, the fact that minute differences between different brands, and even different batches, of tubes was quite well known. However, back then, there was no such idea as "tube rolling". In fact, it was the job of the tube designer to design circuits that were "good enough" that
THEY SOUNDED EXACTLY THE SAME REGARDLESS OF WHICH BRAND OF TUBE YOU USED. The whole point of a good design was that it was immune to those minor and unavoidable variations - which were seen as flaws. (A vintner may see the tiny amounts of chemicals in different wines as significant; to someone who produces distilled water for lab applications they're all just flaws to be eliminated; hi-fi was "distilled water".)
Literally, if I had an amplifier fifty years ago that sounded noticeably different with different brands of tubes, the only question would have been whether the design of the amplifier was so bad that it was unable to compensate for those differences, or if we'd picked a tube type that was so inconsistent and unreliable that a poor tube choice was our mistake. The "obvious" goal was that a good design, with any equivalent "decent quality tube" in it, should sound exactly the same.
The idea that some people seem to have, for example, that "single ended triodes are magically accurate" is just plain silly. I can't rule out the possibility that there is some electrical characteristic of triodes that is magically wonderful, but I can easily prove that the measurable amounts of known audible distortions (like second harmonic) are so high in a single ended triode that, if there was some actual good thing about them, it
WOULD be a miracle if it could be heard above all the extra (and easily measurable) coloration and distortion. (While I suppose it's possible that some people somehow fail to be annoyed by 5% THD, but are sensitive to minute amounts of some so-far-unmeasured other type of distortion, I tend to suspect that they simply like the way 5% THD sounds...)
I've heard a few on those Yagin triode amps and, without getting into the endless debate about "whether they sound better", it's pretty obvious that they sound
VERY different from a relatively uncolored solid state design (or even a relatively uncolored tube design). You can claim that the engineers at Yaqin, rather than choose parts and circuit designs to minimize the coloration, have instead chosen them to maximize it, or you can just assume they're incompetent, but the fact remains that their products have a
LOT of coloration. (It is actually somewhat complicated to make the correct design choices to minimize the colorations involved, so it could simply be lack of technical ability, or unwillingness to buy better parts - but I personally suspect it's deliberate.)
Now, in the case of Schiit audio, and the Lyr, the guys (and gals) over at Schiit Audio are quite competent, so I don't for a minute think anything they've done was accidental. They designed the Lyr to sound different from solid state units like the Asgard and the Mjolnir. Furthermore, they haven't done their best to design the Lyr to be immune to the slight variations between tubes. They know that people enjoy tube rolling, so it would be foolish for them to try and make a design that it didn't work with - right? Rather, they've made sure to use a design that
WILL sound different when you use it with different tubes, by
NOT incorporating circuit and design elements that would tend to render those differences inaudible. (This includes using certain types of circuits, avoiding others, and using certain types of parts in certain locations. Nobody at Schiit Audio would ever say: "We did our best to make the Lyr sound just like the Asgard, and this is the best we could do.") The Lyr uses a MOSFET output stage (solid state); it would be trivial to use one more solid state stage to deliver all the gain they need; instead they've gone to the extra effort to use a tube in that role, and the only legitimate reason for doing so is "because it sounds different" (since the output stage is solid state, they haven't "avoided solid state because there's something bad about it".)
With the current availability of parts, tubes are (at least slightly) more difficult to use than transistors so, to put it bluntly, if a tube amplifier sounded exactly the same as a solid-state equivalent, and nobody could tell them apart, then you'd have to be an idiot to use tubes, since the tube design would cost more to build, cost more to purchase, and be less reliable in the end as well. The
ONLY reason people buy tubes is because they
LIKE the differences in how they sound.
To my ears, the Lyr sounds
VERY different from solid state units like their Asgard (which, to me, sounds quite similar to other uncolored solid-state equivalent units).
(I've never measured one, and probably never will; since I'm not trying to replicate the differences I have little motivation to specifically understand what they are.)