Interesting post on Tubes

[Rajiv]

Hi,

Copying a post from HK tube audio.

Regards
Rajiv

Adrian,

Is this Part 1 or Part 2?

Tim

--- In [email protected], "acfoo2001" <acfoo2001@...> wrote:
>
> INTRODUCTION
> While tubeophiles have tended to focus on which tubes are "good" and which tubes are "bad" there is perhaps a much more advanced way to play the game. In fact, all tubes made after the very early days are "efficient" if the tube type is used correctly. It is also true that no audio tube is perfect. They all have distortions. These distortions are very minor but matter greatly as even small distortions in a transparent system are very apparent.
>
> In short, are there really, truly "bad" tubes or are these just tubes that need to be used a certain way to get great sound?
>
> Part 1 [Note: not published yet] described one clear methodology of using tubes to correct tubes. Part 1's approach may be especially well-suited for systems with vinyl front ends. Part 2's approach is different and would be considerably easier to apply to systems with digital front ends. Note that the word is "easier" but that does not mean "easy" as the methodology requires advanced skills to apply. Whether Part 1 or Part 2 outlines a more effective (better) methodology in terms of sound quality is to be explored in Part 3, which outlines how to combine the two different approaches. Parts 4 to 7 deal with `tube correction' through hardware technologies (e.g. heater technology is one part).
>
> There is some preknowledge required to understand how this method works.
>
> PREKNOWLEDGE A. Tube interaction
> We are used to thinking of the sound of a tube in isolation. However, we can never truly listen to a tube in isolation. Each signal tube is heard with its rectifier, and each rectifier is heard with its signal tube(s). If you have played around with different combinations of signal tube and rectifier and you understand how factory sound works you will have realized that there is a deep and strong interaction between the two. The sound of any one tube will be greatly affected by the other. (Those who use solid state rectification get the same effect ? it is just solid state instead of a tube affecting the signal tube.) The signal tube-rectifier combination is considerably more profound than any signal tube-signal tube combination.
>
> For those who regard listening to a tube as all that is required to assess the worth of a tube this highlights the ultimate impossibility of such an approach. There is NO BASELINE by which merely listening to a tube will teach you to understand it. If you think you understand what a SER signal tube sounds like just by listening to it, the next question should be: "What rectifier did you use with it?" The sound would be dramatically altered depending on which rectifier is used.
>
> Because of tube interaction, the best way to approach assessment is to achieve understanding of what the tube types sound like, what factory sounds are, and how vintage and condition affects sounds. Understanding factory sound allows you to understand how a signal tube and rectifier are interacting because you know which one is additive in the lower treble and which one is subtractive in the lower mids.
>
> Furthermore, truly understanding how these things work allows you to be innovative in solving problem distortions, such as with the method proposed below.
>
> PREKNOWLEDGE B. How to use subtractive factory sounds
> It is useful to consider distortions in terms of amplitude and time distortions. Thus, we have four types: amplitude neutrality (additive or subtractive), amplitude resolution (aka detail), time neutrality (whether or not there are shifts in harmonic structure), and time resolution (to be explained elsewhere).
>
> Most people don't like the sound of subtractive tubes. Because the traditional way of thinking was to consider tubes in isolation, people would generally prefer a "neutral" or "warm" tube. If they have to err on one side or the other they would generally prefer a "warm" tube because it sounds "nicer." That may be true, but imagine if you filled your system up with only "warm" (additive) tubes ? this would make the system sound thick and slow.
>
> Consider some of the subtractive factories: Lansdale (National Union, Philco), Funkwerk Erfurt, Matsushita, ATES Italia, Philips Barcelona (Lampara), GE Schenactady, Tottenham, AWV. These are just a few examples but the general impression is that these are not popular factories. To reiterate, people generally don't like subtractive factories. The price difference between an additive factory and a subtractive factory can be on the order of 10:1. For example, compare the market price for an additive tube like ECC82 Philips Heerlen (say, one of the foil getters) versus a 12AU7/ECC82 from a subtractive factory.
>
> If people use some subtractive factories with their additive factories then this will give better sound and a good subtractive factory can really "save your bacon" in terms of fixing a system's excess.
>
> PREKNOWLEDGE C. Tube factory sound distortions are all frequency-specific
> These distortions do not occur evenly across the 20Hz to 20KHz bandwidth. Typically they are quite narrow band distortions, e.g. Footscray is just high treble and Sylvania is just lower treble.
>
> To show you how those who use THD (total harmonic distortion) as a claim to science are in fact using pseudo-science, consider this. It is very bad science to rely on "total" harmonic distortion and claim that this "total" summation allows you to understand what is going on. It is very good science to apply your brain and understand that there are different types of distortion and that each different distortion should be understood in its own right and solved in its own way. THD is an invention of the latter 20th century and it allowed engineer-types who don't approach matters scientifically to claim to be scientific just because they think they can measure something worthwhile. If you read the vacuum tube literature written by geniuses ? true scientists ? from the early days to around 1950 you find that these truly smart and knowledgeable people had an entirely different scientific approach from the "measure THD" crowd. The level of mathematics applied to vacuum tube technology in the old days is astounding. There were also many types of distortion discussed in the old days (not THD). For example, there was something called "shot noise" to which I saw endless pages of mathematics applied. "Shot noise" doesn't exist in today's tube lexicon ? could it be that, although it can be mathematically described, no one knows how to measure it?
>
> The gravest flaw of hiding behind THD is that THD, as it is commonly used, assumes that frequency-specific distortion does not matter. Tube distortions are frequency-specific. A tube may be a great performer at most frequencies but suffer distortion in the upper bass.
>
> The approach of a true scientist is to study, study, study and retest, retest, retest your hypotheses and theories. Like an old dog that cannot learn new tricks, forever stuck in the past, those who cannot learn are not scientists; "engineer" may be a better term.
>
>
> RATIONALE FOR THIS METHOD
> Engineers think that we listen to music with our ears. A true scientist would study the matter correctly and understand that we hear with our ears and listen with our brains. There are things that the brain puts together that forms our understanding of sound and music. It is well known that we can trick the brain at times into thinking that something is heard when it is not actually heard by the brain. This is psychoacoustics.
>
> The human brain abhors "gaps" in the musical information. This can be due to gaps in resolution (lack of information). More commonly, the brain abhors gaps in tonality. In order for something to sound natural, the harmonics (and the distortion) needs to be of a type the brain finds natural.
>
> Those who prefer the sound of solid state perpetually accuse tube lovers of loving the sound of distortion. There is an element of truth to this. The whole truth is that tube lovers prefer that their distortion be NATURAL-SOUNDING distortion as opposed to the unnatural-sounding distortion of solid state. There are types of harmonic structure that the brain finds natural and types that it does not.
>
> Specific to vacuum tube hi-fi systems, what most of these systems demonstrate is a `lumpy' tonality where some tubes have created some focal frequency-specific additions and other areas are left either neutral or subtracted. Human brains find this unnatural ? we often call this sound "hi-fi" sound. This is often the sound that audiophiles go for because it spotlights certain areas that they feel give benefit (e.g. providing the illusion of more detail).
>
> What if we did not allow this `lumpy' subtractive tonality? How about if we did a very even addition across the entire frequency bandwidth? Because no one factory sound can provide addition to the whole frequency bandwidth, we would have to use a variety of tubes to fill in the whole spectrum. The sound would be tonally even, possibly beautiful, but it would be slow because of the additive nature. Thus, we also need to use some subtractive tubes to take the addition down a little bit.
>
> In other words, we lift everything up so that it better to smooth everything down better, leaving no subtractive areas. (Remember that people hate subtractive tubes. More specifically they hate tubes that are subtractive in a region of the audio bandwidth. They probably won't hate a tube that is subtractive across the entire audio bandwidth ? people like solid state, after all.)
>
> Now here's the neat trick to this. We don't have to make all the additions of the same amplitude. As an example, if the high treble is +1dB, the mid treble is +0.4dB, the lower treble is +0.6dB, etc. then subtracted down a bit, it shouldn't disturb the brain. The brain is likely to still think it sounds tonally natural even though it is lumpy because all the frequencies are still a bit additive (or at very least neutral) and none are subtractive. After all, play a piano in different environments and you find that the frequency responses are very lumpy and different ? but live piano sounds natural in all these environments without lumps of subtraction. The human brain is thus accepting of amplitude non-neutrality of the additive type (time non-neutrality is an entirely different matter and the brain is very sensitive to this). This is applying psychoacoustics.
>
> Another way to understand this issue is to consider CD versus vinyl playback.
>
> CD playback is very low in amplitude distortion, but suffers from considerable problems in the time domain. (This time problem gives the so-called "digititis", which I propose is due to digital power supply issues and not due to all that clocking jitter that the engineering types have been so fixated on. I further propose that I can cure "digititis" through digital power supply technologies that no one has tried to implement before.)
>
> In contrast, vinyl playback has a lot more amplitude distortion (remember that it is a mechanically vibrating information system) but does not suffer from CD playback's time problems. That is why timing and rhythm are so good on vinyl playback systems. This has long been understood by many people. Tonearms tend to have quite a broad additive sound. Cartridges tend to have the most focally additive sound. What is fresh knowledge is that additive amplitude distortion across very broad audio frequencies from vinyl playback may be giving the same sort of effect as the vacuum tube broad additive method described here. In other words, it may help vinyl playback sound more natural and liquid. Use solid state with vinyl and some subtraction will occur (this can be useful). Use too much additive tube sound and it gets thick and slow.
>
> It was the late Kondo-san who said that CD saved the vacuum tube. This may be one of the reasons (there are other potential ones).
>
>
> THE METHOD
> 1. Select a collection of tubes for use in your system that each has a specific frequency addition. Together these tubes/factories need to cover roughly the whole audio bandwidth.
>
> We have worked out additive amplitude and subtractive amplitude distortions for at least 58 factories. (We need to go back and work out some of those we may have forgotten about.) It is very difficult to work out precisely where each frequency distortion is. Unless you too have a synethescope, this is the first very difficult task you will have.
>
> To make your life easier, I recommend you stick only to the ones that are relatively time accurate. Strong time distortions can be interpreted as affecting tonality.
>
> Amongst the remaining factories, divide them into additive and subtractive. Subtractives vary in which areas they are subtractive in. Additives vary in which frequencies they are additive in, but generally the band is relatively narrow.
>
> I will give everyone a little start on the additives.
> (a) Brimar Footscray is additive in the high treble. Philips Sittard, MOV Dover and one of the Ediswan factories are also additive in the high treble, but not as strongly so as Footscray so you may have to use these twice to achieve the same effect as Footscray.
> (b) M.B.L.E. has mid-treble addition. Brimar Rochester also has mid-treble addition, extending up into the high treble (but watch its limited treble extension).
> (c) There are quite a few choices for the lower treble. Sylvania Emporium, PA; Philips Heerlen, Holland; EI; Lorenz Esslingen, Germany; L. M. Ericsson Stockholm, Sweden; and RT Suresnes, France are all choices. MOV Hammersmith also has addition here but its upper bass time non-neutrality will pose big problems.
> (d) The rest you will have to work out for yourself. (I don't give up all my knowledge and technology so freely.)
>
> Because each factory only gives a relatively narrow band of addition, you will need about 7 additive factories to get the job done.
>
> 2. Then select 2-3 subtractive factories/tubes. This will be the right number to bring the additive effect down to where you want it (or else accept a slower sound).
>
> Since subtractions are also frequency-specific (but tend to be broader than additions) you will want to balance these as well. Try not to play only subtractives that are more subtractive in the bass than the treble, or vice versa.
>
> 3. You can fill the rest of your tube complement with some neutrals.
>
> 4. This works out to be about 9-10 tubes plus neutrals (this includes both signal tubes and rectifiers). If you cannot play at least 9-10 tubes then you cannot properly try this method (you can only try it in part).
>
>
> RESULTS
> This works. Tonally balanced, it makes you realize how the hi-fi systems you used to play were tonally unbalanced with lumpy subtraction that sounded unnatural. This is not "hi-fi" sound; this is "natural" sound.
>
> The sound is liquid. Remember how I described that one of several hallmarks of WE factory sound was midrange liquidity? I can now confirm that liquidity is a consequence of "filling in the gaps in tonality". When you use this method you achieve one of the hallmarks of WE without using WE. (Ken-Rad Owensboro also has midrange liquidity although its time non-neutrality means that the harmonic structure is shifted off so that it sounds thinner.) Moreover, WE only achieves liquidity over a broad span of the frequency spectrum around the midrange ? the effect is not present at the treble and bass extremes. This method allows you to achieve liquidity at the extremes as well ? as far as you can push your factory additions.
>
> Does this method allow you to achieve WE sound? WE also has the hallmarks of natural scale dynamics (this requires time neutrality) and high amplitude and time resolution. We will solve these problems another day, but this method allows you to get part of the way there (and better).
>
> Can you use any factory for a specific addition? For example, for the lower treble you can use any of the factories listed above. What may affect your choice is how much resolution a factory/tube has. You generally want higher resolution as you cannot get it back if you choose a lower resolution tube. You may also want high bandwidth as that is something else you cannot get back if you sacrifice it.
>
> For example, for mids to upper mids you could choose to use Mullard Blackburn. But Blackburn is not the most resolving factory sound. Perhaps you could find something more resolving for the same job.
>
>
> We tried this method in a system with a great DAC (this is nowhere close to my ultimate DAC but still a great DAC). There are no relevant questions of CD versus vinyl. Vinyl is not even in the game as this system has all the timing and rhythm and naturalness of sound people attribute to vinyl with none of vinyl's flaws. The only valid comparison is with live music. It is like stepping into another dimension and allows the scientific mind to ask the next question: What more should we do to fully achieve live sound? Perhaps some other technologies (long waveform / deep bass recorded room ambience through a special subwoofer system; general ambience through a method I don't wish to reveal right now). The problems certainly don't lie in the tubes anymore.
>
>
> Adrian
>



__._,_.___
Reply to sender | Reply to group | Reply via web post | Start a New Topic
Messages in this topic (2)
Recent Activity:
Visit Your Group
MARKETPLACE
Stay on top of your group activity without leaving the page you're on - Get the Yahoo! Toolbar now.

Switch to: Text-Only, Daily Digest Unsubscribe Terms of Use.

__,_._,___