Thank you for the vote of confidence....
Your thought for #1 makes sense, but it simply wasn't the case. Back when tube equipment was all there was, what audiophile culture there was concentrated on different types of speakers, and different circuit topology for other equipment (push pull vs single ended; Williamson ultralinear vs pentode wiring; cathode coupled vs plate coupled). The reality was that tubes were treated about like light bulbs - if you needed a 12AX7, you bought a 12AX7 - and the guy behind the counter gave you whatever brand he happened to carry. He'd probably ask if a "used tested good" one was OK, for $1 less, or if you really wanted a new one. It really never occurred to anyone that different tubes might sound significantly different or, if they did, that one would be better than the other. (Bear in mind that an RCA 12AX7 was two or three dollars, and a "premium" Telefunken one was $1 more.) And you couldn't go to the "audiophile store" for more discerning service - because it didn't exist.
There were slight variations, just like there are with light bulbs, and a few "premium" vendors tried to justify higher prices by claiming lower noise and other improvements (cathode coating formulation can actually affect the amount and character of noise you get; and better internal construction can make a tube less sensitive to microphonics due to vibration). By and large, though, a few people had favorite brands, but most people considered them interchangeable (just like some people have a favorite brand of gas, but most do not). Equipment manufacturers virtually
NEVER recommended a specific brand of tube - they called them out by number, although most equipment came with certain brands of tubes (presumably because they bought them in bulk).
Furthermore, most engineers considered that a properly designed circuit
SHOULD equally accept any brand of the correct number tube - and a circuit that performed differently with different ones was overly sensitive. Likewise, it was rare to match power tubes, and almost unheard of to match any other kind (back then "matching preamp tubes" meant finding a set with the same picture on the side of them in the same color - really). While there were some very sensitive circuits that were designed around the specific parameters of a particular brand of tube, this was
NEVER the case with audio equipment - where "having trouble finding a tube that worked in your stereo" would have been viewed as a serious problem.
At this point I should clarify something.... the differences between tubes in terms of how they sound are almost all
RANDOM. Back then tubes simply weren't "rated" in terms of sound quality; they were all ostensibly intended to sound the same. A designer might use a higher voltage tube because it produced lower distortion, and might pick a tube with lower grid-plate capacitance for a radio transmitter, but that was about it.... and those were different types (numbers) of tubes...not different brands of the same tube. You might even find a few situations where one brand would accept a higher voltage, or have a slightly higher plate dissipation, but sound quality per-se wasn't usually considered. (I don't ever recall seeing an ad claiming that a certain brand of tube sounded better - or that anybody should "upgrade" the tubes that came with their equipment unless they required replacement.)
Now, just as different "120 VAC" light bulbs from different manufacturers may vary somewhat in terms of color and brightness, so different "equivalent" tubes can vary. But, when you hear that a Telefunken 12AX7 "sounds clearer and brighter in device X" than an RCA 12AX7, you aren't hearing some magical difference between a premium tube and a garden variety one; what you're hearing is the slight difference in plate-to-grid capacitance between the two manufacturers' tubes interacting with the circuit design of device X to produce a pleasant variation in frequency response.
This is why I always advise people who want to tube roll to try a variety of tubes - and not specifically expensive ones. Since there is no practical difference in the "quality" of the various tubes, the "huge" differences that audiophiles hear are really the more or less random interactions between design variations in different brands of tubes and specific equipment. (So the $200 Telefunken long-grey-plate 12AX7 tube that sounds great in your friend's preamp may sound awful in yours; and the $5 RCA black-plate one that sounds like mud in his might sound better in yours. What you're really hearing is that one has an extra 0.5 pF of grid capacitance.) Unless you have the exact same equipment as the reviewer, you're more likely to find that perfect tube by buying a mixed pack of ten you've never heard of (at $5 apiece) than you are by buying that "$100 premium audiophile connoisseur tube" - which may or may not sound good at all in
YOUR equipment.
I suspect that this reality will make tube rolling a lot more fun for some people - and a lot less for others.
{A lot of what you hear today is what some would call "revisionist history" - like "SET amps"being "magical". Historically, the first amplifiers were single ended triodes. They were then replaced by push-pull models, which everybody agreed was a step up because they had more power and lower distortion. Pentodes then replaced triodes because they again had better performance, lower distortion, and more power - for a given situation. Ditto for the classic beam tetrode (a pentode clone for patent reasons). In 1965, single ended triode amplifiers were reserved for cheap phonographs where there wasn't enough budget to put in a "good modern push pull amp". A typical engineer from 1965 would have found the very idea that someone would deliberately choose a SET over a good push pull pentode to be humor on the level of surrealism.}
I'm not sure whether I agree with your logic on the second point at all...
I do agree that often commercial recordings can "use a little help" - but I'm not at all sure that I agree with your logic there vis-a-vis tubes.
For example, if I hear something that sounds "harsh" it could either mean that the high end is boosted, or that there is distortion in the high end (like with MP3 files). If the high end is boosted, then a little bit of EQ would be the "proper fix"; if the high end is distorted, then a high-cut filter would be more in order. I can see how, for example, a bright component could "cancel out" the effects of a dull component - but, to me, it seems more straightforward (and more flexible) to correct each component rather than endlessly search for faults that cancel out perfectly.
However, to me, what tubes add is a slight "blurring" of the high end (sometimes but not necessarily associated with a drooping HF EQ) and extra harmonic distortion. If the distortion is predominantly second harmonic, it will add a bit of "sparkle" to the upper mids, which can sound very nice with vocals, and can even improve intelligibility. Now, I can see how you might believe that "sounds nice" and want to add it to everything you play, but I
DON'T see it as specifically correcting either of those other problems. In fact, in general, I don't see the colorations introduced by tubes as specifically
FIXING any other common problems. (At most, the high-frequency droop exhibited by some tube equipment might compensate for an overly bright high end, but a high end that is so deficient that it works well as a high-cut filter for MP3 files would eliminate most of the desirable high end on recordings without that problem.)
Personally (and I'm dead serious here), there seem to be a whole "class" of audiophiles who generalize the idea that there is one specific type of distortion caused by anything digital, and that they can "eradicate it" by replacing it with analog coloration (sort of like how a dog tries to eradicate the scent of other dogs from his favorite tree). They seem to think of it like "digital paint and analog paint" - and that you can "fix" anything that's been "digitalized" by "painting analog" over it in sufficient quantities. The reality is that a bad MP3 file played through a SET amplifier isn't going to sound analog - it's going to sound like a bad MP3 file with extra second harmonic distortion; the latter won't cancel out the former (at best it may "bury" it in so much distortion that you don't notice it any more.)
As for the
AMOUNT of difference produced by tube rolling - that will depend on the individual equipment you're using, but I wouldn't be expecting anything what I would call dramatic. (Let me phrase it differently; I'm not aware of any maker of tube equipment who qualifies their published frequency response curve as being valid only with certain tubes.)
I would expect differences in the noise floor - both the amount and character (or grain) of the noise (actually the spectrum).
(But we're talking a few dB out of a noise floor that should darned near inaudible anyway with good equipment.)
I would expect differences in the amount of harmonic distortion (both quantity and spectrum) - possibly significant.
(The tube internal topology, circuit topology, gain, and operating point will all affect this - and some of those will vary between tubes.)
I would expect slight differences in frequency response; but, with good equipment, these should be slight as well.
They will, however, quite possibly vary with different tubes - since their inter-electrode capacitances will be different.
Remember, though, that tubes are intended by design to be interchangeable.... so don't expect
HUGE differences.
If you asked a "hi-fi expert" from 1965 how to reduce the excess brightness of your system, he would tell you to turn down the treble ..... or he might even suggest stuffing some fiberglas into your tweeter.
He would
NOT suggest changing your tubes, or your interconnect cables, or your speaker wires.
These are all things imagined by the "nouveau audiophiles" of today.
Since you mention a background in audio engineering, then I put this to you......
If you have a bunch of bad recordings you want to fix, isn't the best solution to digitize them all, then apply the appropriate correction to each one individually?
That way you can correct the EQ on the one with bad EQ, and use a high-cut filter on the one whose high end is simply beyond hope of repair.
In fact, with the variety of editing options available today, you have about a thousand more options - many of them quite useful....
And a copy of Adobe Audition, plus a copy of Izotope Ozone, together cost less than all but the cheapest tube equipment.
Buy the tube equipment because you
LIKE the way it sounds; not because you hope it will do a great job of fixing something....
Not surprised at your indepth and insightiful reply. Thank you for that. I'd like to bring up a couple of point if I may...
1. While all tube were interchangeable at the time, could that be due to the quality or composition of the components not being "accurate" or sensitive enough so that resulted in a lack of perceived difference between the tubes in different gear? I'm probably not articulating my question properly but I hope that you undestand what I'm trying to convey.
2. I'm almost dreading opening this can of worms but I am familiar enough with Audio Recording and Engineering to know that it is rare for any recording to come out as the artist actually desired. Obviously this is due to a plethora of different factors (cash, equipment, time, engineer, mastering, producer vs band's wants, studio sound vs live etc etc). It's even more true for certain genres of music as well. Not to mention quality of speakers and system that the cd/vinyl listener can afford. The amount of factors are staggering. So a case could be made that the addition of the right tubes with the right gear can overcome some audio issues (for lack of a better word) and have the reproduced sound come closer to what was intended than otherwise
I won't even get into the amount of remasters that are out there even from bands like Pink Floyd (which seem to remaster every few months or so haha).
That arguement aside, adding the tubes can tame harshness and help with other unwanted qualities of a bad recording as well. Obviously I'm not discounting your position at all as everyone should have the ability to listen to the music how they see fit. That goes without saying of course. I'm just hoping that tube rolling won't be too subtle as to make it pointless.
All that being said I've never questioned your abilities (and I don't think you've taken it that way), but was concerned about any potential bias. I'm glad it was as unfounded as I assumed and I'm all the more excited to hear what you come up with.