That would be a good bit if you happened to own the speakers used as the control. The trouble is that there are a huge range of speakers and that the suitability of a particular amplifier into such unpredictable variables is not something that would make measurements productive.
Providing the measurements in a standard format does at least let those interested in the figures compare them on a level field. I think the subjective reviews of amps usually do include a few sets of speakers and allusions to others based on the reviewers experience with similar designs. In other words, a reviewer can draw a similarity between two amps and then having tested the similar amp with several other speakers, he has effectively doubled the number of speakers that would behave similarly and can with some confidence refer to those as well. That gives a reader a much broader sense of what it may sound like with his speaker if he can then draw a similarity with his own speaker or if he has heard one of the mentioned similar amps with his speaker. So I think the reviewers have been doing this long enough to have sorted out your concerns, although without measurments. In short, it's not practical.
-Bill

I'm not at all cerain what measurements you would prefer correlated with what characteristic sound quality. We know an amplifier with a high output impedance will have frequency response problems when it is driving a speaker with an erratic impedance and/or severe phase angle. The difference in sound quality and stability between minimized local and maximized global feedback networks is generally conceded with the latter coming out on the short end of the comparison. (Though intended audience can affect how vehemently this comparison is made in print.) I doubt we will ever end the squabbles over even vs. odd order harmonic distortion content. Not so long as some people feel any amount of either is to be avoided or another group, with equal conviction, asserts both are inconsequential in light of the many other faults of present day amplifier and speaker design.

So what are we going to measure? And how will we measure it?

Your suggestion goes to the heart of the debate between subjective and objective assessment of an audio amplifier's performance. To that end, your suggestion can also be addressed with the same question I would ask of the objectivist who adamantly believes all amplifiers which measure the same will sound the same. What constitutes an amplifier which measures the same? How do we know when we have enough numbers to ascertain whether these two amplifiers are truly identical or in reality are somewhat dissimilar to one another? Can we assume an amplifier with a different circuit topology will have different measurements? If so, isn't informing us the amplifier uses current dumping circuitry enough to say all that needs to be said?

After we get beyond the basic numbers which we can all agree upon as affecting sound quality in one way or another; where do we go next? How deeply do we need to dive into the measurements to be satisfied we have actually tested the amplifier? And how do we measure in a static environment what may become evident only in a dynamic setting? If you view the amplifier as only one portion of the circuit comprised of amplifier-speaker-back EMF from the speaker, how do we adequately assess what the amplifier is doing with only a "typical" speaker in circuit? If we want nothing more than a "typical" speaker, shouldn't we just settle for a load resistor as we curently utilize? It being benign to the point of adding no value to the discussion.

While grasping the concept each amplifier-speaker-back EMF connection will garner somewhat different results, you seem to miss the aim of the subjective review. That is to make enough combinations of components that the reviewer (working through their own set of biases) can begin identifying the signature of just one of the constituents in that circuit. The amplifier. Knowing the amplifier's "sound" along with its technical limitations, as best as we can define them using current measuring techniques, we as consumers can then make educated choices concerning what speakers might or might not be a worthy companion to the amplifier under review. Beyond that, you would appear to be asking JA to measure the unmeasurable.

While Mr. Atkinson would (possibly) be heartily applauded should he discover the next breakthrough in amplifier technology, you are asking him to define what subjective listeners have claimed to be undefinable with present day measurement techniques and our current understanding of how the dynamics of the circuit actually affect the sound quality of one portion of the circuit.

To what end would these measurements benefit anyone? As soon as there was a discrepancy between numbers and assigned values of sound quality, the objectivists would, being the good political rivals they have proven themselves to be, jump on that divergence "like a goose on a June Bug". Better your suggestion should appear on a forum from the Audio Anarchist than from Stereophile.

While most subjectivists would prefer confirmation for their case, the ability to produce such evidence is a shared responsibility of both camps. Otherwise we, as subjectivists, will only add to the level of dimissive derision we receive when we discuss such things as "skin effect".

What I like about JA's speaker measurements is that he tries to correlate them to what sound distinctions that reviewer found with that model "in their listening room". Cabinet resonances and speaker cross-over abnormalities all play a big part as well as even a potentially marginal design.

Amplifiers can be trickier as we tend to want only a straight wire with gain, but as we all know after Michael Fremer's review of the Wavac, just because that amp may measure more like a "loudspeaker" he still really, really liked it!

The fact that the measurements are documented after the listening review is a great way to keep bias out of the review. I feel most confident that if any of the "Phile" writers really like the sound of a piece of gear it would be worth owning, regardless of how it measured. Due to the qty of great gear they have had the opportunity to audition and own, they know what great sound is. That is regardless of whether an amp puts out only 500 wetts, vs the 1,000 marketing watts presented as for sale as with the Chord amp. 500 great sounding watts is still well worth owning. Period.

I believe that all amps have a "sonic signature", but you must have speakers with enough resolving power to discern it, and the training to discover it. The front end will also be crucial in extracting all the software has to offer, sending it along to the amp. Cheap out here and all is lost.

Jim, I think most folks who read more than one issue of Stereophile believe amplifiers have a sonic signature. Otherwise they'd move on to Electronic House magazine and try to figure out their remotes. The question d'jour is; what measurements would be needed to describe that sonic signature with any consistency? What measurement would indicate an amplifier would have superior timbral accuracy, enveloping ambience retrieval or knock-your-socks-off soundstaging? Or, on the minus side; what would suggest an amplifier have that dreaded "honkiness" through the upper midrange? Dryness at the top end? Or that the sound simply would be uninteresting as with the AudioValve Baldur 70 in this month's issue. Any thoughts on that matter?

Two ways are the multi freq tests and look at the harmonic content that the amp produces. The other is square wave response. We all have read that tubes are more even harmonic generators and transistors are more odd, but certainly often down in level.

I still think it comes down to a proper, long term audition where you are often not just listening to the music, but listening through it for the "character" of that sound. At this level the fun is out the window as you are trying to find elements that are less pleasing.

JA's comments about using equipment reviews in listening to his own recordings and discerning what differences he is hearing "through the gear" rather than what he remembered on recording day. This is science in its most basic form for our audio hobby. I wish more recording engineers were equipment reviewers.

In the current issue, go to his comments on the Grace 902 VS the Benchmark DAC. Here he clearly states he would have mixed differently by using the Grace or the Benchmark. Here we are only talking about differences in the sonic signature between two high-end products.

Think of how this multiplies at every recording studio who use various monitors, amps, and DAC from who knows who. From that live moment in time it is impossible for you to know you are reporducing what "actually" happened that day.
Yet, these are the very things all to many agonize over when they buy equipment. We do not usually have a Grace or Benchmark for days to mull over what really floats our boat. But, the writers at "Phile" sure provide us with an unmeasureable service by doing it for us.

With all that said there will be some who will prefer the Grace over the Benchmark, and visa versa. As for me with my tinitus getting advanced I am not so sure any more that I can hear all I need to in order to make the "gnat fart at 50 foot" test anymore. I'll have to leave it to the younger folks...who seem to find no problem with MP3s, which should be the direction the hearing impaired, like me, should be going, not those with perfect hearing.

"At this level the fun is out the window as you are trying to find elements that are less pleasing."

Ahhh, that is where you and I differ. I am listening for things which please me and that is most enjoyable.

As to distortion product, I would direct you to Keith Howard's article in the April edition of Stereophile which focuses on just the sort of harmonic content you describe. It would appear harmonic content is a poor predictor of musical enjoyment when it comes to amplification.

Square wave measurements have been a "traditional" way for amplifiers to reveal some of their character. However, sqaure wave performance would tell us what about soundstaging? Ambience retrieval? Harmonic content? What the original post seems to imply is a search for measurements not yet accepted by the mainstream which would tip the scales toward identifying an amplifier that has those indefinable qualities which audiophiles prize.

Finally, while you would hope the recording engineer/producer would have the best sound possible in mind as they settle behind the mixing board, that is seldom the case when commerical recordings are being produced. "Best sound" is even more difficult to pin down in the recording industry than in a group of audiophiles. What most audiophiles desire is not what the market demands and dollars demand the engineer goes where the money flows. If you read one or two interviews in Mix or Soundstage magazines, you will be discouraged to find what the engineers are really trying to achieve and how they employ their technology to reach those ends. This is most particularly true when multi channel recording techniques enter into the mix.

Otherwise, I think we've still not identified any measurements that would predict which amplifier will have Bill Evans or Paul Desmond brought into my room at my disposal.

Please reference Michael Fremer's review of the AudioValve Baldur and the accompanying IM spectrum display. This is why this amp sounded like it did. Figs 8 and 9 do tell some of the story.

The freq response of the speaker will also play a part as you look at the plot of speaker freq vs impedence and the increases in distortion as the amp output power increases. How an amp interacts with a certain speaker is not measureable since it is an unknown qty when doing just an amp review.

I would agree that we cannot measure it all, but what is measureable tells us much. Just as when JA is able to discern speaker cabinet resonances in measurements and places why certain sonic signatures apply.

It is also why bits are not just bits and why jitter and square wave responses do not tell us all about CD player's sound qualities. Especially when something as simple as a variable output control can lessen the performance of a very good Cd player like the new Quad 99 as compared to its fixed output.

Just the even/odd order harmonic issues is enough to drive many into tube amplifier land and some even further into low power SET land. Better is in the ears of the beholder. He who pays decides.

I would bet that the trained reviewers in "Phile" could look at the typical measurement data JA provides and put many amps in two or three piles based on "suspected" better performance. It may not be totally accurate, but I'd bet on them rather than against them.

Go to www,musicangle.com and check out Michael Fremer's interviews with some recording and mastering engineers and get a sense of their frustration with producers. The engineers generally know, they just don't always win.

So far you have discussed traditional measurements which most everyone agrees will affect the sound quality. The original post would appear to be searching for new techniques and fresh numbers to describe an amplifier's performance. I don't think IM distortion, damping factor or output impedance fall into that category. Everyone is in fairly close agreement regarding their effects. While we weren't discussing CD players, it is not a matter of what jitter and square wave responses do or do not tell us about a player's sound qualities. It would take a truly stuck in the fog Luddite to pretend any of these measurements would not affect the performance of a digital player or an amplifier. So, I'm not in disagreement with you that these numbers are relevant to what we can expect to hear, but these are not new measurements you are suggesting.

You state, "How an amp interacts with a certain speaker is not measureable since it is an unknown qty when doing just an amp review." Is that your final answer? I ask because, as I read it, that is precisely what the original post theorizes should become our new measurement regimen.

Post: I ask because, as I read it, that is precisely what the original post theorizes should become our new measurement regimen.

Now you have my curiosity really high. And just how do you or anyone suppose JA to be able to test an amp with every conceivable "speaker perameter" possiblity in the retail world be achieved? I thought this was the point Mr. Howard did make, although he did say he was hoping to return to it at some point. Some speaker manufacturers would hopefully chime in here as thay have the toughest job on the audio planet.

In spite of the largest set of variables, The "speaker load", and we still expect amplifiers to act like "just a straight wire with gain"? Then we see an amp behave differently when just applying a constant simple, 2,4, or 8 ohm load, not withstanding every phase-angle known to man?

Those of you who are looking for the difinitive, be all-end all, test every set of possiblities a speakers can thrown at it, is not quite real world. I say that when we look at the simple 19 and 20 khz test frequencies, just asking the amp to reproduce those, and what do we see is all kinds of other frequencies showing up that we didn't ask for.

Look at the differences between the Baldur with all the additional freqs at less than -40db from the fundementals. What is even more alarming is the freqs showing up below 6khz. (Mr. Howard's point I believe). It is these large level differences we can hear.

Compare that to the Halcro MC 20 that is over -80db down from 19 & 20 khz with nothing showing up below 6khz. or the Chord, which has become the whipping boy these day at "Phile forums", with both 18 and 21 khz down almost -80db and again nearly nothing below 6khz, as is the Moon at about -70db at 18 and 21 khz and nothing raising its head below 6khz. The Yomomato A-08 at the same 18 and 21 khz harmonics showing up much higher at about -30db from the fundamental.

Then look at the inexpensive Outlaw and its behavior and you can understand why MF might have liked the sound the amp produced and did not run screaming from the room or quickly grow tired of the sound. It does the basic things an amp should do.

Why do we need more testing when, IMOH, we have amps that show these kinds of characteristics and others do act more like a straight wire with gain? I, for one, am satisfied that it would be impossible to test for "all possible perameters" that the real world (speakers) would throw at us.

I suppose JA could give is a broader range of visual square wave tests. Or he give us more adjacent khz test to see if the amps offer more harmonic content in the lower and mid freq ranges.

I only brought up the CD testing because when you agonize over your player purchase that keeps the other harmonics way down from 19 and 20 KHZ and then you run it through the Baldur did you gain or lose? What ever is the weakest link will bring down the whole system, it might even be, and probably is, my speakers. To note that the gear MF assembled with the Baldur in not chopped liver, but highly regarded music delivery products for sure.

The fact that JA could audibly distinguish between the much smaller differences in the Grace and BenchMark DACs is also very telling. These differences are much, much smaller than from the Baldur to the Moon, Chord, or Halcro amps. Could it be that JA would have heard these minute differences auditioning on the Baldur? Maybe not.

It may be that for the likes of Moon, Halcro, Chord, Levinson, Boulder, Pass, McIntosh and others we DO need some new testing perameters (to seperate each other better), but for all the rest what we have does tell us much. Maybe just not all as others would like.

Jim, I think if you'll look back at my first post, you'll see that I never have suggested there is a way to measure the dynamics of the amplifier/speaker/back EMF circuit. You seem to be missing the point that I'm in agreement with you on this subject.

Possibly this isn't worth discussing since there are only two respondents going around the Maypole. But the original post, as I read it, asked for additional measurements which would predict how an amplifier would "sound". You and I have agreed that traditional measurements tell us "this" amount of information. I suggested "that" amount of information will not tell us how an amplifier recreates a soundstage, captures the recording venue ambience nor will it tell me whether the amplifier will touch me with emotion. Those are qualities for which we have no measurement techniques at the present time.

So, the question is, beyond the "traditional" measurements, what should we be looking for? Or, am I reading you incorrectly and you feel IM distortion, et al, will tell you all you need to know about an amplifier as it cohabitates with various speakers?

"The fact that JA could audibly distinguish between the much smaller differences in the Grace and BenchMark DACs is also very telling."

I hesitate to suggest this lest I be once again misunderstood, but isn't that why we read Stereophile? Isn't that exactly how we should all listen? If we are not learning how to listen by reading the experience of the reviewers in Stereophile, why are we reading the magazine? I would ask how JA learned to listen in this manner. And, I suspect, the answer would be by becoming intimately familiar with live music and the emotions and reactions it stimulates in his memory. So, Jim, how do we measure that quality? That is far beyond what IM numbers would indicate. Or, do you think not?

I admit to finding measurements fascinating.

Not a be all, end all, but a gloriously mysterious realm with much left to uncover.

I enjoyed reading about the sonic impact of the different connection schemes for the Grace headphone system and how well the sound correlated with the jitter levels of the connection.

Much still left to learn, but I think there would be really cool things to try and see now:

1) You know those square wave plots for amps? I'd like to see the same waves plotted as output from several different speakers or speaker types when they measure that. I'd like to see if the differences on the scope translate to different speaker outputs.

2) You know those truncated sine waves they post for digital playback? I'd like to see what they look like coming out of a speaker, as well.

3) How about speaker reviews with different waterfall plots and such with the speaker being driven by different amps?

3) How about measuring speaker frequency response results with different cables? I'd even like to see some of the classic things like impedance, inductance, or capacitance measured for cables. (Can they do that?)

4) I'd really like to see plots of different phono cartridge performance on test discs to see if we can correlate tracking ability or frequency response or any of that other stuff with what we hear subjectively.

__________________
__________________

The bedrock of hi-fi is tinkering and curiosity about how things can be made better.

You would think, by our very nature, that we would all be intruiged and curious as to what makes this great stuff tick and how we can learn more about what it's doing.

Now, with the web, we could have ten pages of cartridge measurements to look at as we read one of Mr. Fremer's reviews or try and match what Mr. Phillips is describing about an amp and how that is played out via measurements of what it makes speakers do.

We have these great measurement toys, I say play with them more!

(A side kudo to Mr. Fremer - I like how he acknowledges JA's existence in reviews and ponders the connection between what he hears and what JA will come up with. It shows a good audio spirit.)

I have a bunch of the old Shure test discs you are welcome to borrow if the mood hit you to become Dr. Frankenturntable. "Doctor, its alive!!!!!!!"

When I got started into this hobby I remember Julian Hirsch of Stereo Review showing cart produced square-wave photos that did tell part of the story. That was when tts were the norm.

The speaker measurements might be more problematic in that JA would have to discern what the microphone and mic pre were adding to the measurements, but is would still be fun for us and a whole lot more work for him.

I do believe we would be disappointed in what we would see. It should sober us up and wonder why we would complain about amps at all.

I hope Stephen is not becoming too stressed about this forum. I know this is falling under part of his job description, but it is supposed to be fun. I am glad the casino dudes are gone. What a mess,

Quote:
I hope Stephen is not becoming too stressed about this forum. I know this is falling under part of his job description, but it is supposed to be fun. I am glad the casino dudes are gone. What a mess,

Oh, I just need to lighten up. I really hate those casino dudes, though.

Jim, I doubt that I could segregate amplifiers based on measurements. Overall I don't find them very useful at all. And so many people have so many opinions regarding how a circuit should perform that I don't feel like getting involved in the argument any longer. Whether an amplifier, speaker, etc. has this or that number for this or that measurement is of little value to me unless the product is grossly distorting the signal. I'm guessing this is the reason for Sam Tellig's popularity and longevity. From his early days as the Audio Cheapskate, he has never, to my knowledge, taken a measurement of any piece of audio other than how hard is it to lift. His ears are his validation and that is what matters to him, and I suspect, to his readers.

I used to glance at square wave performance but have decided there's not much point any longer. The rise time of an output device will not tell me whether I will like an amplifier or not. I've seen and sold amplifiers with "terrific" numbers that did nothing to make me want to take it home. And, by most conventional numbers at least, my McIntosh tube amplifiers don't score that well on a test bench when compared to a $500 receiver. That doesn't mean I see no place for measurements, but I have little use for them.

I'm in favor of merely listening and not judging anything with how it performs in a static situation when viewed in isolation from the other components of the system. While I view specifications as a necessity for a designer, I can't imagine a competent high end designer who wouldn't trust their ears to tell them what they've designed. To that end, I wonder how designers get over ruled, or the issue is ignored, when a poorly chosen volume control brings down the performance of a piece of equipment as it did with the Quad digital player. Of course, I do remember when models were upgraded to Mk.II status after a bad review. So possibly not all designers hear as well as I would like to imagine. I know quite well many do not share my values for what is important in the final sound. I am therefore expecting a follow up review of the Halcro amplifier when the problem has been resolved. Simple poor numbers are too easily avoided to think this was a properly functioning unit.

Mr. Vigne, I think I'm catching your drift.

Would it be fair to say you are mostly about the "musicality" of a sytem?

If so, I agree. There is something to music that is ineffable and regardless of spec, good gear captures that essence.

Cheers.

QUOTE: To that end, I wonder how designers get over ruled, or the issue is ignored, when a poorly chosen volume control brings down the performance of a piece of equipment as it did with the Quad digital player. end quote

I am not interested in beating a dead horse either, but this is exactly what Michael Fremer found with the Baldur amp. All he did was listen. So if the Baldur engineering team decided "that" is the sound they wanted that is fine. Like you, people do not buy Mac gear based on measurements. Neither did AD when he bought the Lamm's.

Measurements can tell us some of why an amp sounds like it does if we are curious to know. Most manufacturers "voice" their amps to what they like and what they think will sell. Mac can choose to update the 275 while Quad can just reissue their mono blocks as they were originally designed using the latest components that could match that design. It still sounded good to AD. Great? A conscious choice.

The Quad CD 99 volume control is a separate issue. The fixed outputs still make this a near class A player. The variable outputs are there as an extra as are the digital inputs which is a very nice touch by Quad.

How could this be used:

I often record our church choir with trax. I have used the direct outs to feed my recording rig and control it with a channel fader into my laptop. The variable out I use to feed a separate power amp that feeds the trax playback speakers the choir hears when performing. This allows me to change that level without messing with the actual recorded trax sound level or quality. This works very well. The actual quality of the sound level the choir hears is not that big of an issue in this instance.

Just another tool. A welcome one that works well.

Quote:
Mr. Vigne, I think I'm catching your drift.

Would it be fair to say you are mostly about the "musicality" of a sytem?

If so, I agree. There is something to music that is ineffable and regardless of spec, good gear captures that essence.

Cheers.

Years ago, after reading Robert J. Reina's review of the Creek 4240SE, I became intrigued by his findings regarding its relative neutrality, especially given the cost. It didn't measure as perfectly as I would've hoped, but after auditioning it, I ended up buying it. Over the years, I suped it up by replacing the stock power cord with an aftermarket power cord, a few key resistors, new opamps throughout (including phono stage), and by mounting carbon fibre cones to the chassis - Shakti stones figured into the equation as well, but those aren't really mods. After speaking to Mike Creek on the phone and purchasing The Art of Electronics by Horowitz and Hill, I was going to take out the zener diodes used to regulate the voltage to two opamps and replace them with LM 317 and 337 regulator based circuits to firm up the bass. I was going to gang some caps to lower inductance and replace the two large stock ones and replace the ALPS pot with a stepped attentuator. I had already squeezed quite a bit more performance out of this unit than it had any right to offer, and I'm certain I could've taken it to another level. But, I reached a point of diminishing returns and felt if I wanted to get the performance I really craved, I'd have to lay down some more serious green to dance with the bigger boys.

An interesting thing happened in those 9 or 10 years of tweaking. I still wanted to know things measured well and that the amp was reasonably accurate (Halcro specs would be comforting), but I found I mostly cared about the sound. I never saw any measurements of the GamuT dual mono amps, but, the sound and performance were ultimately the only things that mattered. I bought one without ever seeing any charts or measurements... something I never would've done in the preceding years. I still have no idea how they measure... I suspect the square waves are a bit rounded and the amp is a 2nd harmonic distortion monster, but none of it matters when I'm hearing the timbre and soundstaging this Danish delight dishes out. I feel the goosebumps on my skin and the chills down my spine when listening to music and know I made the right decision.

Just in passing, remember that should the Gamut T's be a push-pull amplifier the nature of the circuit will null most second order harmonics. What is left would probably be interesting for someone to know, but as you suggest, Jeff, the music is and should be what makes us smile at the thought that we chose wisely.

Diminishing returns are really what separate those of us with no part of Bill Gates or Mark Cuban's portfolios from those gentlemen's shopping habits. After a while you are often begin chasing something merely different rather than truly "better". How many of us have seen the audiophile who changes components more often than the rest of us change breakfast cereal? Having sold to too many people with too much money trying to divest themself of their curse, I occasionally left the shop feeling I was better off having to settle for where my returns got me.

Jim, I think we are once again in agreeement that designers will "voice" their products by tweaking parts. But, in keeping with the spirit of the original post, do you assume they do their tweaking while also considering measurements? The original question seemed to go to wishing for numbers which would correlate to perceived sound quality. Would replacing, let's say, a capacitor (metal film for paper in oil) or yanking a few zeneer diodes change the measurements in any appreciable way? Would they change the measurements only within a circuit or possibly the entire component? If they made for a change reflected in poor measurements, what would the designer have to choose between? We have all probably read of the decisions Conrad Johnson made while creating their flagship pre amplifiers. Not taking anything away from their talents as designers, how does their method of replacing one component part with another relate, if at all, to the question of the thread? (As a side note, where does this stop? Are we, the consumer, really well served by a designer who opts for an obsolete NOS tube when a readily available part would have served "nearly" as well? What happens when that tube needs to be replaced?)

Regarding my preference for a "musical" system, yes, that is my leaning when I judge a system or component. The problem being that "musical" can be merely an excuse for poor judgement. I decided long ago that my system would never be 100% "accurate" based on the diminishing returns (monetary and auditory) which I faced. When I return from a live performance, I am not disappointed with my system's capabilites and I am often impressed by what I hear from a fairly modest, and by many folk's recollection ancient, collection of components. However, live music reminds me where my dimishing returns are and how far from the reality of live music most systems deviate. I established my set of criteria for what my system must do based upon my preferences/dislikes and budget. That the system serve the music well is primary. How it achieves that goal is personal. I have no way to measure what the system is doing to get these results. And that appears to be the main thrust of the original post. What will tell us whether we will like an amplifier?

Lest I leave the wrong impression, I was obviously aware that McIntosh has a reputation for accuracy. Just ask the many audiology labs which still form a large segment of Mac's sales about accuracy, repeatability and dependability. These are, for me, at least a portion of what I consider when I think about a new purchase and the issue relates to my question regarding obsolete NOS tubes. Though probably not in the perview of the original post, how do you measure these things? I have found tubes which I consider more "musical" but should the original design have included tubes which satisfy my soul or satisfy the listener who wishes the same sound after replacing the tubes? This is off the topic slightly, but is also, I feel, a bit of whether we will "like" an amplifier. Would we possibly like it a bit more if we heard it with different tubes? This was a controversy several years ago when reviewers regularly replaced the stock tubes to create an amplifier which met their tastes. Therefore, they were not reviewing the amplifier as supplied by the manufacturer. As I said, off topic. But the question remains, how do we measure "musicallity" and "likeability"?