Differences in Valves from Different Manufacturers

There is always so much controversy about this issue that I thought I would post the info I often quote in its own place.  We can then simply refer to it in other threads when it is necessary.  I'm convinced that very few people bother to follow links to see proof that they already don't want to agree with so here it is on our own forum for all to look at, think about and criticise.  Firstly here is the link:

Valve Make - Frequency Response Comparison

EDIT: It is worth noting that Amplified Parts has now pulled this graph of the curves! They continue to display the totally misleading and meaningless "block graphs" of how great the differences between valves supposedly are still not giving you any idea of how small the range of each of those valve plots is.

This is the data lifted from that external site, (to which I have no connection).

Valve Comparison Frequency Responses

The site sells valves from different manufacturers so has an agenda.  It is therefore stuffed with descriptions of "...lots of top end harmonic texture" and "a pleasant top-end chime and a leaner tone overall" and...  Well you know the drill by now!  (Where is a vomiting emoji when you really need one?   )  With that fact in mind you will see that the picture attached here is intended to prove that these valves do, in fact, all sound different.  That's the first main point here, this info is the best that could be produced to prove that valve makes DO sound different!

Their setup is this.  They have built a standard Fender preamp circuit exactly to Fender spec.  They populate it with each valve in turn.  They then accurately measure the frequency response of that very same stage in each case.  This is exemplary.  That is just the setup we have needed to show the differences between valves once and for all.  It means that each valve is being used with exactly the same supporting circuitry and thus rules out, firstly, any differences being caused by two independent sets of the other circuitry and, secondly, our ears and brain fooling us into believing that there are differences we can hear when there aren't as you can't fool a spectrum analyser.  These results can't be fudged - - - - at least in themselves, (more of that later).

The original page starts with a number of bargraphs of their differences, one for each model of valve, intended to show the balance of that particular valves bass, mid and treble response and its overall gain.  These go on interminably valve after valve but wait.  If you look back to their opening section "Graph Interpretation" and think about what they say there you will realise they are already opening the door to the truth.  They firstly point out that there are differences IN GAIN and say nothing about frequency response.  They are then careful to add that every manufacturer's valve stock has a range of gain within it into which their valves fit and even from the same manufacturer these will not all be the same!  At the bottom is the only relevant piece of data on the page.  This is the graph of the frequency response of each valve as measured in exactly the same circuitry and plotted together on the same axes.  That is what my second link above shows.

The resulting graph is a joke in terms of showing that valves have differences!  What it actually shows is that each valve has EXACTLY the same frequency response.  They do show a range of gain but look at the vertical axis.  The difference between all of them is only 2.8dB.  If you remove the lowest of all you can even see they are pretty much within 1.5dB of each other.  This is only in overall gain.  IN TERMS OF FREQUENCY RESPONSE THEY ARE ACTUALLY EXACTLY THE SAME TO WITHIN A MINUTE FRACTION OF A DECIBEL ACROSS THE ENTIRE AUDIO SPECTRUM.  Just imagine each of those graphs slid up or down to sit over all of the others.  How much difference would you see in their shape?  It is small fractions of a dB.  And it is the differences in shape not overall level which give rise to your tonal differences.  The gain differences are easily balanced out with your Gain or Volume controls which will slide the signal levels up to match.

So how do they use this data to demonstrate the differences they still claim exist?  This is the "fudging" to which I referred earlier.  They invent their own display method!  The bargraphs for each valve are a meaningless con.  The scale of 1-10 which represents the overall differences in the range is only 2.8dB wide.  Each notch on it therefore represents around 0.3dB difference.  What is the maximum difference that any valve shows between its Bass, Mid and Treble values?  It is 2 notches on their scale.  Out of 28 valve types tested and displayed, 3 show completely flat responses with no difference at all, 17 show a maximum of 1 notch for 0.3dB, and 8 of them show a maximum difference of 2 notches or 0.6dB.  Even in hifi terms, much more demanding in tonal balance than guitar work, that is considered inaudible.  And that is all very rough interpretation of those curves.  If you move just a few cycles left or right of the Bass/Mid/Treb measurement points you will get different results.

I submit this for you to look at and criticise.  If I am wrong in this then please explain where I have made my mistake.  A site which sets out to show that there are differences between valves performs a good well set out test, gets its data in accurately, then fluffs up the results to show differences to be significant when they plain aren't.  I URGE YOU, DON'T BE FOOLED!  TRUST YOUR OWN INTELLECT AND NOT THOSE OF SOME SELF STYLED GUITAR GURU AND CHECK THIS OUT, THEN DON'T BE AFRAID TO SPREAD THE MESSAGE.  YOU HAVE THE PROOF.  (And no, I'm not a guitar guru, I'm just a good engineer who trusts only claims backed by evidence and not common knowledge.)

You can't fool someone who doesn't want to be fooled.  You make it possible to fool them by making them WANT to believe you in the first place then the job is already done.  I won't go there.  Evidence is all that matters and that must be independent, repeatable and outside of human error and influence.  That means "engineering"!  Nowadays, if you genuinely CAN hear it you CAN measure it.  If you can't measure it then you have either discovered a new scientific phenomenon or you are fooling yourself in amongst the existing ones and you can't really hear it at all.  The ear and brain are one of the easiest senses to fool.  You just give them a very short time to forget and tell them they can hear what you say.  The Emperor then puts on his new clothes.

Have you swapped preamp tubes in and out of an amp before? I find the best way to hear the differences is to turn all your preamp parameters all the way up. I think some of the more noticeable differences com from different tube manufacturers. I think you won't find much difference in sound of powers tubes unless you to a completely different type of tube, say El34 to KT88. I made that power be switch on my Coreblade and it was pretty drastic mostly because I have a Port City 2x12 OS Wave. I realized just a couple of days ago that one of the contributing factors in my excessive thickened bass guitar like sound was the fact that I was using a very bright Jensen BlackBird Speaker with a less bright ET 90 Speaker which when I dialed back the Jensen high-end frequencies it would dial back (normally) bright ET 90 to sound more like a bass speaker, which seemed to thicken up the overall tone. I even check out how low the cab was going frequency wise with a makeshift Pioneer car sub speaker that I hung from a kick drum practice pad, that I turned into a sub kick mic, and it was picking up frequencies down in the 40 Hz range that was as nearly as loud the high-end frequencies guitars normally generate.

I'd been wondering my sound sometimes would be so thick when 2x12's were normally not. I figured it out once I started testing out the speakers on their own going back and forth from the same patch. I should have realized this long ago but just thought it was the ported cab or that the speaker was bassist than the other one. When I used two Identical ET 90 speakers the sound was nearly as bright or as thick with bass.

Yes I certainly have tried swapping out valves in preamps.  That was the start of my search for proof that it was all a myth.  When I tried different makes, despite what every one else was saying, I couldn't hear any difference.  And I do have good ears which have a long history of being able to detect very minute and more subtle differences in hifi reproduction when improvements are made.  Differences in speakers are immediately obvious and they DO make a significant difference to your overall tone.  It took me a long time to find the actual living engineering proof of whether I was right or wrong to believe there are no real differences in valves and, when I did find it, it was handed to me by a company trying to actually prove they did sound different to sell more valves.  Their final results, to an engineer, show that there are no differences worth even worrying about.  However, that didn't stop them frigging the results into an artificial and meaningless way of displaying any differences to make the models look different.  In their way of displaying the data each grade in their scale of 1-10 represents less than 0.2dB and which is at the level of the random ripple in the testing.  I can point you to audio tests online which will show you that you have a hard time accurately identifying relative differences of 1dB!

That's the link I posted above.  To make it easier for anyone I'll post the actual graph here.  These valves have all been tested by plugging into the same Fender Blackface style preamp circuit custom built for the test and measured with the same test equipment so there are absolutely no other differences from other sources to take into account.  Have a close look at the frequency response of those multiple 12AX7 valve makes and tell me where the difference in shape lies?  They actually show that every make of valve has the exact same frequency response curve.  The make of valve tone idea is a myth and is down to things like people performing tests on new v old valves or just down to the placebo effect where they hear what they desperately want to hear.

I have also been a part of blind listening tests on a few occasions over the years and they show quite clearly that when the golden eared ones don't know which valve they are listening to they simply can't identify it.  Sorry, but all of the evidence that I have been able to turn up in a lot of years looking in this field points to there being no major tonal differences between valve makes.  If I'm wrong then please show me how that is displayed in the graph attached which are not my own.  I'm not taking a position on this then looking for results to prove it.  I always look at the results and then go with what they show is the truth.  I'm interested in learning not in being right.