Talk:Tube sound

Military devices
Regarding Tubes vs Transistor should be written something about military devices because tubes have some advantages vs transistor in military use because tubes are more resistant to nuclear explosion.

--- Here's what I can find:

There is no huge benefit in tube immunity to EMP. Shielding and proper grounding are required for SS, just the same as that used for any device including components used in tube equipment. For example: http://newsgroups.derkeiler.com/Archive/Rec/rec.radio.amateur.policy/2008-11/msg00172.html

EMP that is going to knock out tubes is also going to knock out people who aren't shielded.

There is no modern military requirements for tubes used by the audio industry be used. The only use of tubes in the military is on out of date equipment: http://mae.pennnet.com/display_article/66727/32/ARTCL/none/none/1/DOD%27s-abhorrence-of-microwave-tubes/

FAA reference: http://govinfo.library.unt.edu/npr/library/clinton.htm

Critical military transmitters have been changed to solid-state realizing higher uptime, lower maintenance costs, and less wasted energy: http://www.locusinc.com/library/2004%20Nuts%20Bolts%202.pdf —Preceding unsigned comment added by Kgynkisd (talk • contribs) 16:03, 13 May 2009 (UTC) Kgynkisd (talk) 16:07, 13 May 2009 (UTC)

Tube sound
Although I agree that there are many audiophiles who prefer "tube sound" or "better" sounding audio rather than absolutely accurate audio, I think it's misrepresentation to say that [i]all[/i] audiophile systems do not strive towards accuracy. In fact, I'd say it's a pretty even split between these two grounps. For example, most cable manufacturers, although they tend to advertise the subtle changes in sound their cables produce, in the end their design decisions typically show an attempt to reduce, rather than create, distortion from the cables. This is also true for many other non-tube components, such as DACs, CD transports, etc. .. 69.86.94.231 03:29, 18 August 2007 (UTC)

In my humble opinion all the sections relating to the properties and sources of the "tube sound" are vague and unscientific. They jumble subjective "audiophile" terminology with technical descriptions of supposed causes without leading to any concrete conclusions concerning the objective properties of the "tube sound". The whole article needs tightening up. 212.159.59.5 (talk) 18:45, 1 February 2009 (UTC)

Agreed on the above. Also the phrase "Some audiophiles prefer the sound that is produced by the distortion characteristics of tube-based amplifiers." seems to be contradictory. An audiophile interested in accurate music or instrument reproduction would not want amplification to add distortion. Distortion levels in modern gear, used within specifications, is below the hearing capabilities of people. The only place one might say they prefer one type of device over another is when it's pushed beyond specification and into distortion. Even then there are many different amplifier configuration for tubes and solid state that this couldn't be generalized. Perhaps this article should be renamed "Tube Sound Distortion" ? Kgynkisd (talk) 16:50, 13 May 2009 (UTC)

Tubes vs Transistors article Tubes vs Transistors in Electric Guitar Amps — Preceding unsigned comment added by TubeGod (talk • contribs) 03:19, 9 October 2012 (UTC)


 * No device, no matter how linear and accurate, is completely free from distortion (whether tube or solid state). Linearity is obviously the goal of high end audio amplification components' designers (and listeners), but is never fully achieved.  Distortion can even be present even when the amplification device is not being noticably pushed into overdrive.  Often, transients that come at the onset of the attack of a note (i.e. the beginning of when a string is plucked or the onset of a drum tap) produce harmonic distortion.


 * Vacuum tubes have an advantage in dealing with tranisents, because they produce even-ordered and lower ordered (i.e. 2nd ordered) harmonics (even-ordered and lower ordered harmonics are preferable, even pleasing, because they are in harmony with the music). Whereas, solid state tends to produce odd-ordered and higher ordered distortions, which are not harmonically in synch with the music, and are therefore more jarring and perceptible--and yes, even distortion that we don't think we can hear is subtly perceptible.  All audio amplifiers, whether tube or solid state, have sonic thresholds (clipping) where the peak of certain transients cannot go beyond.  Tubes have the advantage of producing "soft clipping" when the threshold is reached (i.e. a round peak of waveform).  Whereas, solid state tends to produce "hard clipping" (rectangular shaped peak of waveform), which is more noticable.  Some solid state amps have introduced special ciruitry which can mimick the effect of soft slipping, but usually end up veiling the musical signal in a way that tubes don't.  Simplicity in a ciruit's signal path is usually a virtue.


 * Furthermore, there are different kinds of distortion. While tubes tend to produce slightly more harmonic distortion than solid state (but that isn't really a disadvantage, based on what I mentioned above), they produce less of most other kinds of distortion than solid state.  It would also be a mistake to assume that we know everything about all of the different parameters of distortion.  There are types of distortion that we have not yet discovered (or don't yet know how to measure), so there is an element of mystery in all of this.  Sometimes, the ear can be the best guide in evaluating an audio system.  So, I think that audiophiles who prefer tubes are not without justifcation. Garagepunk66 (talk) 06:02, 29 March 2013 (UTC)


 * Well there are some things to keep in mind here in my opinion. With even harmonics, being lower-ordered than odd harmonics means they are more dense. There's "more of them". Also the even harmonic fifths are still out of tune with 12TET intervals and there are higher-ordered harmonics that are very out of tune, still being inharmonic, even if they are less noticeable and sound more like "timbre". Odd harmonics also still have some harmonics that are as in tune as even harmonics. And to point out the obvious, all musical instruments already have a harmonic structure. This means that all additional harmonic distortion adds harmonics that are dissonant. It doesn't matter what kind. Harmonic distortion is always dissonant. Not to mention any preexisting inharmonicity or more complex chord structure than monophony or fifths (power chords..). So in my opinion (and to my ears for that matter) the appearance of harmony in even integers is superficial. Tube distortion is still dissonant and still sounds like distortion. If you are talking about very low levels then it doesn't matter audibly. We really can't hear it.


 * I guess what my points come down to is that tube distortion vs. solid state distortion is really a preference of musical style. Even harmonics aren't actually "better" or "more desirable". They are just preferable if that is the sound you want. The article is about the subjective sound of tube distortion which is really only audibly different at high gain levels yet includes a lengthy discussion of the difference in low-level signal degradation. That's not tube sound. Tubes are also not so objectively special that they deserve their own article. Remember that tubes are only obscure among technology nowadays because they are obsolete, not because their harmonic distortion is different. If we are just talking about music then tubes aren't special because they're used all the time. And on the other hand, sometimes you also want transistors. They are cleaner amplifiers and their high-gain distortion has a different timbre that also sometimes you want instead of tubes. Solid state amps only distort significantly if you are trying to do that or are inattentive so it doesn't even matter if hard clipping sounds bad.


 * Here are my two suggested options:
 * Rename the article to something like Vacuum tube distortion or Tube gain sound and add a section that justifies why it deserves its own article.
 * Merge the content in to one of the other articles about distortion. Here are other places on Wikipedia that are like this article: Distortion (music), Valve amplifier, Valve audio amplifiers


 * Radiodef (talk) 17:48, 29 March 2013 (UTC)


 * I guess the sum of what I think is that tube gain and solid state gain sound different but neither is "better". They just sound different so a comparison of subjectivity and usage is more appropriate. The comparison is covered elsewhere. Also I think the debate is erroneous and not worthy of mention because all good systems have THD that you can't hear. Cranking it is not for audiophiles and is also covered elsewhere. Radiodef (talk) 18:24, 29 March 2013 (UTC)


 * You make some very good points, Radiodef. You have a point that when it comes to harmonic distortion: that tubes can also produce some dissonant harmonics (i.e. even-harmonic fifths).  Of course, harmonic distortion is still only one kind, and not necessarily the worst.  As you are well aware, there are other types of distortion (and we will never completely understand all of the parameters of distortion).  I agree with you that it ultimately comes down to the subjective ear of the listener.  Personally, I use solid state audio equipment, for the sake of convenience, ease, and lesser expense.  I am quite satisfied with it (so either type can be good).  However, the best audio equipment I have ever heard (to my subjective ears) is vacuum tube.  Tubes cannot be considered obsolete, because they are still favored amongst a large percentage of audiophies, not to mention guitarists.  Nor would I say that solid state is necessarily more clean.  I have found that tubes handle midrange and highs with a great deal of transparency and have more three-dimensional imaging.  You are correct that components usually do not go into clipping until pushed hard, but there are certain transients that can max a system out, even when the system is not being noticably overdriven.  In a musical passage that rapidly goes from a quiet passage to a loud one, a system can, and often does, clip.  Garagepunk66 (talk) 07:07, 30 March 2013 (UTC)

A.monteiro2007 (talk) 18:14, 24 September 2014 (UTC):

To be constructive and positive I present what I believe is the kind of text that this article should contain. This means replacing the whole text with the following:

“Tube sound” is an expression used referring to the type of sound produced by tube (or valve) audio amplifiers.

The problem with this expression is that it is very subjective once that the sound produced through a tube amplifier is not easily identifiable as such.

It may be recognizable under very specific circumstances like a small degree of overdrive of the amplifier. But a sound produced by a tube amplifier can be achieved with other technologies like transistors / operational amplifiers or digital signal processing (DSP) amplifiers, leaving us without a scientific definition.

Amplification types

For the purpose of this article we can split the amplification in two types: hifi and instrument amplification.

Hifi amplification

For hifi amplification it is assumed that the objective is to reproduce an input signal without distortion.

In this field, both tube and solid state (transistor) technologies can produce indistinguishable results. Only in the case of overdrive producing a clipping situation a difference may be noticeable (see below).

Intrument amplification

Instrument amplification, namely electric guitar, is probably where the expression is more widely used.

Here, the expression reffers to a distorted sound which breaks up gradually when the signal level increases, as oposed to a case where the clipping is abrupt like in the old transistor guitar amplifiers.

It is important to say that today (2014) all types of technologies, tubes, transistors / operational amplifiers and DSP amplifiers are able, with adequate design, to produce the gradual clipping as in old times only tube amplifiers did.

Nevertheless the expression is frequently used generally to mean a good more or less distorted guitar sound.

monotonically decaying
At risk of appearing slow I venture that although I've a fairly good education, it includes three levels/years of university electrical engineering, the term "monotonically decaying" used repeatedly in this article has me nonplussed. Yet I CAN handle Laplace and Fourier transforms - so I doubt the general reader will fathom the description given. FWIW I think this article was much better a year ago than it currently is (it was then known as "valve sound" and it was less subjective and seemed better referenced). RichardJ Christie 10:47, 2 September 2007 (UTC)

Monotonic function should explain--TreeSmiler (talk) 04:12, 29 November 2007 (UTC)

perhaps. but it doesn't. like much of the article, the use of this phrase is careless of the reader, lacks context & clarification, & hints at a poor grasp of the science behind a rigorous analysis of the properties of the various sorts of amplifier design. no offence to the unknown writer, but if you've got a good grasp of your subject, you should be able to explain it to a layman, i.e. without recourse to terms that send the reader off to another page to decode them.

we're no closer, after all of these years, to understanding what- if any- contributes to the nuanced differences (in the domain of high fidelity, at least) between valve & solid-state approaches to audio amplification.

I'm more than familiar with the audible behaviours of both when they are deliberately configured to exaggerate their respective characteristics for the amplification of musical instruments.. but the page fails to successfully explain the differences between valve & transistor behaviour when the context is amplification with the aim of "high fidelity", with or without intentionally-introduced artifacts. there is a huge market predicated on these differences, so why is it such a mystery?

& it will remain a mystery when contributors to this page use esoteric (& possibly, in this context, meaningless) mathematical functions to explain the different distortions introduced by each.

duncanrmi (talk) 16:30, 19 June 2017 (UTC)

Software emulation
I'm not qualified to write it but there needs to be a heading detailing the history and current art of software tube emulation. We've got solid state emulation listed but not software. Any takers? Binksternet 14:40, 2 September 2007 (UTC)

nfb and Miller effect
"Triodes always have a property called Miller Capacitance which couples the output plate to the input grid, effectively acting as a negative feedback path. Because this feedback path is extremely short, it performs very well."

well, not in any meaningful sense. It simply reduces gain at high frequencies, typically above audio. This is quite different to useful nfb.

FWIW... Tabby 01:10, 2 December 2007 (UTC)
 * the physical length of the nfb path is immaterial
 * the most effective nfb by far is nfb around the transformer, which is a significant contributor of distortion.
 * Miller effect exists in transistors too.

Guitar amplifiers
I think the article now gives the impression, that most people use solid state amplifiers and only audiophiles use tubes. However, I wouldn't say it's exaggerating to say that every professional guitarist uses tube amplifiers nowadays. I'm sure there are some who don't (especially since some solid states, like the Roland Jazz Chorus are great), but almost everyone uses tube amplifiers. Amateur guitarists might use more solid states (tubes are very popular in there too), but usually because good tube amps cost so much. I think this should be reflected in the article. 212.213.90.10 (talk) 09:59, 6 December 2007 (UTC)


 * To quote the article: "If sensitivity to small signals is a significant goal, then tube designs will typically[dubious – discuss] outperform transistor designs." This is perhaps the subject of the most audible aspects of "the tube sound." When guitarists who play through a tube amplifier speak of "picking dynamics," they are usually referring to the tube amplifier's responsiveness (sensitivity) to the subtle input signals compared to a transistor amplifier's ability to "pick up" these small signals.  I do not claim to know, however, whether the human ear is sensitive to these subtle, small signal dynamics that many tube aficionados claim to be a differentiating factor between the tube's sound and the transistor's sound.--12.179.188.5 (talk) 19:56, 9 June 2009 (UTC)


 * So... it's reliable sources we need, not personal observations, however intriguing. Binksternet (talk) 20:07, 9 June 2009 (UTC)


 * And where no reliable source exists, we should leave similarly unjustified, but incorrect information in the page? 94.192.83.59 (talk) 14:10, 26 January 2010 (UTC)

Vacuum tubes are the vastly preferred amplification devices for most professional guitarists. They are definitely more sensitive to the touch (i.e. the notes when one plucks on the guitar). This is probably due to the different way the current flows through the circuits in a vacuum tube amp. Furthermore, tube amps handle distortion a lot better (for the very reasons I stated in a thread "Tube sound" above). Garagepunk66 (talk) 06:17, 29 March 2013 (UTC)


 * Vacuum tubes are vastly preferred by professional guitarists that like tubes. They are just different. I like tape distortion in my mixes a lot but I'm not going to make an article about how tape is better for ____ reasons that I like it. Radiodef (talk) 18:36, 29 March 2013 (UTC)

Redundant text
I'm 66.92.144.65. I removed a number of sections which I thought to be redundant to other sections later in the lengthy article. I noticed that these changes were reverted. I am adding this comment to explain what I was thinking when I made the edits. —Preceding unsigned comment added by 66.92.144.65 (talk) 05:10, 14 December 2007 (UTC)

Class B p-p distortion "like" that of SE class A ?
"Push-pull tube amplifiers can be run in class A, AB, or B, When in the class A region the distortion will be as described above, but while in class B the distortion will be SE like."

This statement has been in the article for a long time. As balanced class B push-pull doesn't generate even-order harmonic distortion how can its distortion spectrum be "SE like" (SE can generate even order harmonic distortion). I have never read a text that states that class B p-p operation is "SE like" (which in any case is a vague statement). Also, the sentence that follows it "Also, a class AB amplifier will have crossover distortion that will be typically inharmonic and thus sonically very undesirable indeed" is in this context inaccuarate as it implies that crossover distortion is greater in class AB than it is in class B operation. User R.christie 121.72.32.183 (talk) —Preceding comment was added at 04:15, 7 March 2008 (UTC)


 * I saw that too and took it out, it's wrong. Be bold.  Articles don't fix themselves. Atomota (talk) 01:22, 30 December 2008 (UTC)

All systems distort?
"All practical amplifiers distort to some degree; some audiophiles prefer the sound that is produced by the distortion characteristics of tube-based amplifiers."

What the fuck? Distortion happens when you stress out the amplifier; a 4558 silicon op amp clips when overdriven, while a tube simply distorts softly and THEN clips. In a tube the plate can STEADILY supply some amount of power, and then when your signal pulls enough power off the grid (which blocks flow) it CAN'T supply that much steady power and distorts. If you turn down the power on the grid enough (big signal) the plate is free to dump everything it's got, and pretty much flails at it.. --John Moser (talk) 22:40, 4 April 2008 (UTC)


 * I read that section as simply stating the truism that no amplifier is absolutely perfect. There is some harmonic distortion in any realizable amplifier.  A perfect amplifier would have a mathematically flat, zero dB response across infinite bandwidth.  A good amplifier provides a very low degree of distortion, and the article is discussing the subtle differences in distortion between tube amplifiers and other kinds.  Not that I could tell the difference if you held a gun to my head!  :)  Another take on this is that sometimes we apply distortion on purpose, in the form of a graphic equalizer.  CosineKitty (talk) 19:51, 28 June 2008 (UTC)


 * The "distortion" applied via equalizer is on purpose, that is, the change in frequency-related amplitude response is not considered distortion—it is intentional. The quoted, questioned phrase above does indeed intend to convey that all amplifier designs include a degree of distortion. It may only be a few hundredths of a percentage point, though. Ideally, it would be impossible to hear the difference in the most linear tube designs and the most linear solid state designs, as long as they were kept within their linear performance envelopes, not overdriven. Binksternet (talk) 21:46, 28 June 2008 (UTC)

Cleanup tag
I agree with this assessment. What we need I think is a better selection on sources. Can anyone suggest good article which are pro-tube and/or neutral in their technical presentation. Given the right sources, we should be able to balance the article. Neutron Nick (talk) 14:16, 12 June 2008 (UTC)

tubes > transistors
my uncle builds his own amplifiers using vacuum tubes over transistors. he told me vacuum tubes are better and something about the signal on an oscilloscope. i will try and get more information and a source. —Preceding unsigned comment added by 68.188.195.13 (talk) 20:37, 16 July 2008 (UTC)
 * So, did you get more information? --195.148.75.83 (talk) 03:21, 8 December 2008 (UTC)

SR & D Rockman
I've been interested in the design of the Rockman_%28amplifier%29 developed by the venerable Tom_Scholz. It's widely regarded as the Gold standard of guitar distortion, at least for solid state effects. Having studied the internals of a Rockman decades ago, it is a bit more complicated than a couple of op-amps and diodes. I don't know of any publically available schematics, but I happened to stumble upon one of several patents filed by Mr. Scholz in the '80s and '90s. Particularly, patent 5133015 - Method and apparatus for processing an audio signal.

It's an interesting read. I am of the opinion that the Rockman inspired this patent. What appears to make this design interesting is the use of a compression stage and significant filtering prior to deliberate signal distortion. The distortion stage itself is quite simple - an op-amp driving a couple of diodes. What I find curious is the patent schematics show two diode in parallel cathode to cathode - not back to back. It's clearly asymmetric distortion, which many suggest is a cleaner sound. It's not clear to me why two diodes are used, though. Could it be an error in the drawing?

Anyone have any thoughts on this? Madhu (talk) 01:53, 31 July 2008 (UTC)

This article is a mess.
This article needs either a massive and strictly methodological re-write, with solid citations -- or simply to be discarded. Folks need to understand that an effort like Wikipedia is not the sort of place where one can just pontificate off the top of one's head without citation.

24.182.34.116 (talk) 18:05, 18 January 2009 (UTC)


 * Yes it is a mess. What should be done first? How about tagging all suspicious sentences, including time stamps? Uikku (talk) 15:59, 20 January 2009 (UTC)


 * I agree. This article was clearly written by individuals with limited audio background, overwrought opinions and mediocre writing skills.


 * For example, comments about negative feedback are completely off-base. The overwhelming majority of high fidelity tube amplifiers produced from the 1950s onward incorporated negative feedback (even those using Williamson ultra-linear output circuitry).  In fact, the term "high fidelity" had to do with the absence of distortion in the early amplification stages, which was achieved both with high quality tubes and negative feedback.  Negative feedback per se has nothing to do with "tube sound" whatsoever  "Tube sound"" is the result of the greater degree of linearity achieved in electron tubes as compared to equivalent solid state devices.


 * Regarding clipping and the resulting distortion, that is a completely separate topic. It is true that the onset of clipping in a tube circuit is gradual and produces mostly even-order harmonics, which tend to be less grating on the ear.  Conversely, the transition from linear reproduction to clipping in a solid state circuit is very abrupt and produces a sizable odd-order harmonic content, which is usually objectionable.


 * Be that as it may, an amplifier intended for use in audio reproduction (as opposed to a musical instrument amplifier) is never intentionally operated in a way that will produce clipping. Unfortunately, this article gives the reader the erroneous impression that intentional distortion is part of the "tube sound."  No audiophile would consider clipping to be acceptable sound, regardless of the type of amplifier.  A musician may consider distortion to be acceptable under certain conditions, but would still expect the amplifier to stay linear when operated below its maximum power rating.


 * This article should be rewritten from scratch by someone who knows the subject or should be deleted. In my opinion, the current article is a mishmash of opinion, bullshit and unsupported assertions, and is not up to encyclopedic standards.


 * Bigdumbdinosaur (talk) 20:50, 18 January 2009 (UTC)


 * I have to agree. Tagging all the suspicious parts is going to be quite a chore. Kgynkisd (talk) 17:53, 13 May 2009 (UTC)


 * Citations to an external authority would be great. Someone has just taken out a lot of POV-ish comments . But this, too, is a POV.  Nobody writes about "audiophile" beliefs because the academic world is content to let the audiophiles believe what they want, and people in the business of course have a vested interest in keeping the mythos alive, so it's hard to get "fair and balanced" coverage. It's a bit like writing about wine criticism...there's no academic studies. --Wtshymanski (talk) 13:59, 24 August 2010 (UTC)


 * I think the whole audiophile thing is unnecessary. IMO, this article should just have a technical explanation (with citations) of what is responsible for the tonal characteristics associated with tube musical equipment. As well as perhaps some quotes from musicians/producers/engineers regarding their thoughts on working with tubes. Drsmoo (talk) 00:44, 25 August 2010 (UTC)


 * Tagging is an easy pie. Generously tag all suspicious claims and let the tags be there some time. If appropriate citations have not appeared after few (perhaps four) months waiting, then delete the suspicious claims. That is a polite method and in harmony with Wikipedia rules. Of course, if the claim is plain nonsense or trash, delete immediately. I think that I am rather forbearing user, because I have left some suspicious and tagged claims be in some articles more than a year, but I don't suggest that other users should be as forgiving. However few months would be moderate waiting time for source material. Uikku (talk) 15:39, 25 August 2010 (UTC)

A.monteiro2007 (talk) 16:13, 24 September 2014 (UTC)
 * I agree that this article should be rewritten. It contains a series of pseudo scientific statements including sources, misleading the readers.

"Tube Sound" is a subjective concept. But what people usually mean when the expression is used is something that can be explained. And the best way to achieve this is with some audio examples. This would be a very short article totally avoiding quality comparisons.

Tube transient response.
You should probably mention somewhere that tubes have the ability to have transient power peaks far exceeding similarly rated transistors because of design topology. Transients can cause some tube amplifiers to output double or more of their rated power for a very brief moment because of the large amount of energy present on the B+ rail and saturation delay in transformers, whereas the transistor counterpart would either be unable to produce the waveform, or clip when asked to.


 * True enough, but a full power waveform of the worst highest amplitude and shortest transition time possible itself has a limited slew rate. Any amplifier, either tube or transistor, can be designed to handle this worst case condition by simply having a slightly faster slew rate than the musical waveform in all stages.  The fastest possibe transients with a potential of causing slew rate distortion in an amplifier comes from the best vinyl records and best phono cartridges.  The compact disk slew rate (codes per second, which converts to volts per second in an amplier) is a known absolute maximum.  As long as the worst case musical waveform's slew rate doesn't exceed the slew rate of any stage in the amplifier, slew rate distortion (a.k.a. transient intermodulation distortion) becomes impossible, even if that amplifier happens to be transistorized.  How many transistor amplifiers, or even tube amplifiers for that matter, are actually designed to meet this standard is another question entirely.
 * To categorize one type of device as better by saying a given device is unable to perform what the other can is dubious. Since it is the overall circuit that is the actual limiting factor, not the device type. To say, "the transistor counterpart would either be unable to produce the waveform, or clip when asked to" is fundamentally incorrect, since power supply plays a large, if not the most important role, in transistor amplification. — Preceding unsigned comment added by 96.250.190.179 (talk) 01:50, 10 October 2013 (UTC)

Ohgddfp (talk) 14:29, 25 September 2011 (UTC)

High impedance inputs: advantage?
I am taking out the section that supposedly describes how high impedance inputs that are generally found on tube amps are better than 15k Ohm impedance inputs on solid-state amps. All the references in that paragraph establish the impedance value of this or that amplifier, but they don't say why a 100k or 1 megohm impedance might be better than 15k. There was no talk of the 10:1 rule of thumb at which the source is assumed to be a pure voltage source, not a current source. If the source's output impedance is as high as 1500 Ohms (unusually high for a CD player or preamp!), then a 15k input impedance is perfectly fine. There's no advantage to having a higher input impedance if the driver of the source device can handle the very small current load without distortion.

The paragraph also failed to describe how 1M inputs are subject to greater RF interference. Binksternet (talk) 15:44, 4 July 2009 (UTC)


 * The new position of that text doesn't help establish why a high impedance might be better. The offline Rotel and Sony sources are opaque to others, and most likely represent input specifications specific to those particular models. If so, the references do not support the general statements in the text. A general source, like from an engineering textbook, would be better. Binksternet (talk) 18:16, 4 July 2009 (UTC)


 * About the idea that high impedance inputs are better. An engineering text book would talk about a voltage dividing effect of say for example, 1500 ohm output impedance preamp driving a power amplifier with 15000 ohm input impedance.  If the above-mentioned impedances are fully resistive, then all that happens is a gain change, which is not a problem.  If, on the other hand one or both of those impedances has a high reactive component, like a very high input capacitance, then the system frequency response will change by the voltage dividing effect, problably from being correct to being incorrect.  A very low preamp output impedance, combined with a very high power amplifier input impedance will minimize these possible impedance reactance problems. Ohgddfp (talk) 15:32, 25 September 2011 (UTC)

High fidelity and negative feedback
The article says now: "The majority of commercial Hi-fi amplifier designs are Class AB, in order to deliver greater power and efficiency, typically 12–25 watts and higher. Such designs will invariably use at least some NFB." That gives to believe that Hi-fi amplifiers have always included intentional negative feedback. However that is not true, since High fidelity was found on 1927 by H. A. Hartley and negative feedback for amplifiers was patented by Harold Stephen Black on 1937. Obviously High fidelity existed ten years before the first negative feedback amplifiers. I think that something should be done to that paragraph. Uikku (talk) 00:37, 11 August 2009 (UTC)
 * I will correct that. Uikku (talk) 18:45, 29 August 2010 (UTC)

Some audiophiles prefer the sound that is produced by the distortion characteristics of tube-based amplifiers.
That claim is in the opening paragraph. Now there are two references added. However verification is difficult mostly because there is no relevant text quotation and also because the referred article is not freely available (on Internet) or so it seems. An exact and relevant quotation from the source would be most welcome. Uikku (talk) 20:38, 23 August 2009 (UTC)
 * I have read this said (and explained) a lot in audio magazines, and I'm sure that sources can be easily found. Generally, in high-end audio amplification, whether it be vacuum tube (i.e. valve) or solid state, the aim is always to be as accurate and as free of distortion as possible--but no audio system, whether tube or transistor, is completely without distortion, particularly when pushed to the limit, so the question, then, becomes how does the system deal with whatever distortion that is there?  Adherents of solid state point out that tubes are higher in certain kinds of distortion, such as harmonic distortion.  But, what they fail to understand is that harmonic distortion is only one kind of distortion (amongst many--there are many kinds, some that solid state is higher in).  Regarding harmonic distortion, adherents of vacuum tubes mention that, when pushed hard, tubes tend to create even-ordered harmonics, which are more musical sounding to the ear, than solid state, which tends to create odd-ordered harmonics that give a less pleasant sound.  On a vacuum tube amp, harmonic distortion, though slightly higher, is less of a liability than on solid state.  Also, tube adherents feel that tubes handle transients better, because peak signals tend to max out in a more rounded (i.e. natural) waveform ("soft clipping"), rather than in solid state, where the transients max out in a square (cuts off abruptly at the top) waveform ("hard clipping")--transistor amps often have circuitry to create a soft-clipping effect, but it tends to veil or muddle the sound, unlike in tubes, where the sound remains open and extended.  Tube enthusiasts are sometimes accused of being subjective, because they do go as strictly by stats, but also go for "feel."  Tube enthusiasts believe that that the sound of tubes is more organic, natural, and true-to-life.  Keep in mind that there are many different types of distortion--we cannot pretend to know them all.  The types of distortion we currently classify are only in certain isolated parameters.  For instance, when digital came out in the 80s, analog was criticized, because it measured higher (than digital),  in most distortion stats used at that time--yet audiophiles were complaining that digital sounded harsh and sterile.  But, later on, new types of distortion were discovered that applied only to digital (that analogue didn't even possess, or was very low in).  So, the people who criticized analogue on strictly technical grounds were proven to be short-sighted--and the "golden-eared" audiophiles were proven right.  The same applies to now: maybe our ears and our brains tell us something that cannot be measured by any piece of test equipment--an audio signal is too complex.  That is not to say the solid state is bad, or that digital is bad, because both have come a long way over the years.  A lot of the problems of early digital have been much improved on with more recent advances.  It comes down to what people like or prefer.  But, still to this day, many audiophiles (though not all) prefer tubes, often for the reasons (stated above) so often explained in audio magazines.Garagepunk66 (talk) 00:50, 14 May 2015 (UTC)

Weasel words
What do you think about this recent edit? Uikku (talk) 14:37, 24 November 2009 (UTC)


 * To my eyes, that's vandalism: Removing appropriate maintenance templates without fixing the indicated problem. Binksternet (talk) 16:49, 24 November 2009 (UTC)


 * Reverted 24.85.131.247 (talk) 06:25, 18 February 2013 (UTC)

Shorter's 1950 work on higher-order harmonic distortion
I removed some new text added, text which meant to explain that higher order harmonics are more audible, and should be given greater importance in testing and analysis. The text referenced D.E.L. Shorter's paper "The Influence of High Order Products in Non-Linear Distortion" which appeared in Electrical Engineering in April 1950, pages 152–153. Shorter, a BBC researcher, wrote that "subjective assessments of non-linear distortions does not always agree with the assessment based on measurement".

I removed the Shorter bit because his prescriptive proposal never made it into regular testing procedures. As well, attention must be paid to the frequency at which the non-linear distortion artifact occurs: as the distortion rises above about 4kHz, it will be heard less well by the listener because of the equal loudness contour of human hearing. Binksternet (talk) 08:32, 20 June 2010 (UTC)


 * So you just plain removed the well sourced text bit that I added yesterday because, as you said, Shorter's method "... never made it into regular testing procedures ...". Is that some Wikipedia rule I haven't heard about? The article appeared in the journal Electronic Engineering, not "Electrical Engineering" as you said. Obviously you have not read Shorter's article. The one sentence that you quoted is in the opening paragraph of the paper and can be found also in Daniel Cheever's 2001 thesis and I guess that is your source, because Cheever had the same spelling mistake and his thesis is so easily available. Shorter's article is often cited in literature and well known and the core discovery described in his article has not been questioned, as far as I know. If you know some scientific sources that really question it, show them. I read Shorter's article and quite effectively summarized the relevant information in the text that you deleted. Here is a challenge for you: You'd better have sources ready if you are so willing to delete. Don't you deletionists have better things to do? Uikku (talk) 10:58, 20 June 2010 (UTC)


 * I am not seeing how Shorter's article relates specifically to tube sound. It relates to all amplifiers, all audio devices. The section does not belong in this article. Binksternet (talk) 14:46, 20 June 2010 (UTC)


 * Your explanations for your actions seem to vary greatly. First you said that Shorter's "prescriptive proposal never made it into regular testing procedures" but when I pointed that indeed Shorter's discovery was correct, holds true and has been often cited, you changed your explanation to "The section does not belong in this article." What actually happened right before you did anything to that paragraph? I corrected an error | that had been in the article since 2 February 2007, which means that the error had been there over three years but you had not done anything to it until | yesterday I corrected it and then | you quickly deleted what I had written. Do you understand how your actions look like? If you don't know what I mean, may I show you more examples | here and | here. Uikku (talk) 16:17, 20 June 2010 (UTC)


 * Without a bearing on this article, the paragraph on the audibility of higher distortion artifacts should be taken out and applied, with modifications, to the article psychoacoustics. Your arguments about past edits of mine, and past of yours, do not affect the discussion of how the higher harmonics paragraph does not fit here. Binksternet (talk) 17:54, 20 June 2010 (UTC)

Power Supplies
There's a section which claims most amps use regulated power supplies. In my experience that is incorrect. Unregulated power supplies are probably MUCH more common. For example, most receivers likely use unregulated power supplies for cost reasons. Multi-channel home theater amps using a shared power supply probably use unregulated power supplies more often than not. This makes sense, as all the amps can share from the central power supply. Pro amps seem to be moving to switching power supplies to save weight. Seems that some expensive "audiophile" amps such as Mark Levinson use regulated power supplies. But near as I can, they are the exception, not the rule.

p.s. This article is one of the worst I have seen. There's so many objections asking for citations, it's unreadable. Maybe someone should just put a blanket statement that the article needs much more citations at the beginning. At least it would be more readable, if not totally accurate. —Preceding unsigned comment added by 75.71.200.117 (talk) 05:26, 2 August 2010 (UTC)


 * As far as power supplies go, they are wide in design type and are designed for the specific criteria used in the design. In amplifiers that require high linearity, the added expense of regulation of some form is required. In amplifiers where reaching a specific price point is required, power supply may not include these components (and their cost), with the resulting non-linearity as being of less importance. As a designer for power supplies for Mark Levinson many years ago, I understand the design aspects of power supplies, particularly in amplifiers.


 * However I do agree the statement you talk about is a generalization and too broad in scope. It should say "Some amplifiers use regulation and other don't" or something to that effect.


 * In general this entire page is wrought with facts, misinformation, contradictions, and in some cases purely subjective opinions being stated as facts. I think the entire page should be deleted as a technical document and should just describe the "tube sound" as "a subjective topic of philosophical debate", describe the use of the term in different arenas, describe the various arguments within those arenas, and nothing further. No attempt should be made in this article to prove or disprove any particular views, which is where this article has stretched towards. Neither by technical or scientific analysis, since using fact (or fiction) to support opinion is meaningless; that's the nature of opinion. You can't prove an opinion, you can only prove a theory. And that is not easy, since theoretical proofs require peer review in order to be validated as fact. Opinion is only stated. It needs no proof.

Sound samples & Graphics required (oscilloscope traces??)
The article could be radically improved if someone were to actually add some graphics showing examples of how the various audio-clips effects effect the sound, plus some sound samples allowing readers to hear the differences. (plus it needs a very big TIDY UP.)

As it stands it is a "discussion" between those with differing opinions, and imparts little useful information to someone who has little understanding of the subject matter. —Preceding unsigned comment added by 85.210.109.65 (talk) 17:22, 26 December 2010 (UTC)


 * That's impossible. The only way to hear what a tube amplifier sounds like would be to use one. Even if recorded, when compressed and played over your cheap sound card and tinny computer speakers it would sound like something compressed and played using a cheap sound card and tinny computer speakers. Think about it...can you tell what a good audio system sounds like by listening to it over your mobile phone? No, you can't, not even over a land line. You have to be there in person. As for oscilloscope traces, the only ones that would be informative to the average reader would be a comparison of soft vs. hard clipping. Yworo (talk) 00:43, 27 December 2010 (UTC)


 * Suppose that you have found an excellent scientific article that reveals something important about tube sound and there is a very good graph that explains more and better than thousand words. So, there is a graph that should be used in Wikipedia. But how do you do that? Use a scanner! Ok? But is that a good idea? It depends: If you got rights, go ahead and scan it. If you don't have rights, do something else. Redraw it by hand. Do a computer simulation. Measure by your own lab equipment and take a picture from the oscilloscope. But what ever you do, remember to refer to that original article because it is important. It is scientific, not your own original research, so it is suitable for use in Wikipedia. Uikku (talk) 12:33, 27 December 2010 (UTC)

Some sonic qualities are easy to explain
Under the subtitle Audible differences there is a paragraph that begins with this unsourced claim: "Some sonic qualities are easy to explain objectively based on an analysis of the distortion characteristics of the gain device and/or the circuit topology." I am not willing to delete that sentence because it is pretty much true. Class B devices produce crossover distortion, while Class AB devices mostly do not, and Class A absolutely do not produce crossover distortion at all. That is about circuit topology. But then there is also the claim about gain devices. A simple Class A tube amplifier is different compared to a small Class A transistor amplifier. Without any negative feedbcak they both will produce harmonic distortion. The gain devices are different and that causes differences in harmonic distortion. That is probably perceptible. I would like to see a good source that covers the whole sentence. If I don't find it, I hope someone else will. Uikku (talk) 17:00, 3 September 2011 (UTC)


 * I think the flaw in that statements is in the first part: "Some sonic qualities are easy to explain objectively ..." with the word "objectively" being the limiting adverb. I also have a problem with the use of the phrase "easy to explain". These are complex occurrences and never "easy" to describe, since they include psycho-acoustics, perception, and presumption, in addition to the known distortion characteristics. Basically ones mans distortion is another mans sonic bliss. Which leads me to the use of the term "sonic qualities", which has apparent underlying presumptions.


 * The statement should read: "The way the human ear perceives some types of distortion can only be described objectively when based on an analysis of the distortion characteristics of the gain device and/or the circuit topology."


 * This statement says essentially the same thing; however, is less ambiguous. The addition of the word "only" is essential in correcting the ambiguity. Something can only be objective if it is quantifiable, which what analysis affords. So, "something can only be described objectively if something else is done to quantify it". Jrlaudio (talk) 03:05, 10 October 2013 (UTC)

Class A
I have doubts about two edits in this article. The first is this one: "(which is substituted with non-linear higher harmonic distortion, q.v.)" The second is "push-pull". Also single-ended Class A amplifiers are free of crossover distortion, not only push-pull amps. As far as I know, biasing an amplifier to Class A is not a worse source of harmonic distortion when compared to Class AB. A hot amplifier is a source of waste heat and noise, and they are the natural byproducts of Class A, but not necessarily distortion. Nikolas Ojala (talk) 21:56, 8 October 2012 (UTC)


 * I agree with you. I have removed the misleading parenthetical statement about non-linear distortion, and I removed "push-pull" which is addressed in the next section down. Binksternet (talk) 23:02, 8 October 2012 (UTC)

Ambiguity Abounds - Major Revise Needed!
I just when through this entire article. If a person was looking up what is meant by the "Tube Sound" they would not find it. What they would find is a sounding board by people who have preferences and strong views and want to discuss the "Tube Sound", not describe or define it. This had lead to an article fraught with real facts, misinformation, supposition based on limited understanding, outright guesses, and most outrageous, citations that are used out of context supporting opinions. The debate itself.

The "tube sound" is a philosophical debate, a cultural phenomenon, nothing more. Not a fact, nor a fiction. It is a debate. The article should be more about the debate itself, than the merits of the various points of views within the debate.

I have big issues with entire paragraphs and sections which are complete exaggerations or simply supposition, based on other things either not understood completely or simply used out of context intentionally. I have a huge problem with the entire section called "Output Impedance". It gives the impression that tube amplifiers are all trans-conductance amplifiers, which is simply not true. Some have current capacity up to an impedance as high as 17Ω, however that is nowhere near infinity which defines what is classified as an trans-conductance amplifier. Then it goes on to imply all solid-state amplifiers as voltage amplifiers, and exclude saying the same about tube amplifiers, making an implication about solid-state amplifiers that appears to attribute some negative behavior.

It is this type of ambiguity that makes me believe that the article should not include the debate details, basically all the circuit descriptions, the opinions about perception and all the "proofs" everyone is giving. The "tube sound" is a subjective argument and cannot be objectively quantified. So remove all the opinion, proofs, facts and science. And simply describe the phenomenon and mention the debate; however, the article should avoid engaging in the debate. Leave that for the audio message boards, not Wikipedia.Jrlaudio (talk) 04:01, 10 October 2013 (UTC)


 * Thanks for your input. I did few corrections recently after I had read your comment. Especially I tried to improve the opening paragraph of the article and the "Output impedance" section. Nikolas Ojala (talk) 14:51, 13 October 2013 (UTC)

107.77.193.168's templating, explanations
I attempted to add reason entries but some did not show up so here are those that did not: dubious 2: amplifier characteristics don't really matter? dubious 3: contradictory with the point in this section's second sentence, particularly if one gives audiophiles' opinions a lot of weight. Unreliable source 2: Importance of accuracy appears to be very variable in terms of the listener and the situation; many examples of people preferring inaccurate recording and reproduction, like microphone presence peaks and exaggerated bass -- or just not minding inaccurate reproduction because the sound is considered good enough even if not very accurate. Many high-end and pro audio articles also make dubious or overly-broad claims, because snake oil, casual (not scientifically-verified) thought, and unchecked human cognitive biases are major factors in audio marketing and the industry's general approach with the public. Need scientific evidence instead. This source literally says things like "many designs are not particularly musical sounding". Musical-sounding? Really? That is the sort of vagueness that is not good enough for sourcing. — Preceding unsigned comment added by 107.77.193.168 (talk) 23:27, 7 March 2018 (UTC) "Large amounts of global negative feedback are not available in tube circuits, due to phase shift in the output transformer, and lack of sufficient gain without large numbers of tubes". I feel I have to disagree with the first part of the quoted statement. Up to 20 dBs of global negative feedback can be applied to most tube amplifiers if it is properly compensated. I have done it with no problems. Proper compensation is very important in preventing oscillations, but it is achievable with good measurements and choice of compensation component values. Please see Max Robinson's excellent article here: — Preceding unsigned comment added by Photogoddle (talk • contribs) 16:06, 29 September 2019 (UTC)