|
|
| Elliott Sound Products | Audio Myths |
Main Index
Articles Index
For reasons that may initially seem unclear, the world of audio is rife with myths. Some are harmless enough provided the wallet pain isn't an issue, but some are quite malicious and potentially dangerous. Mains cables fall into the last category - many countries have very strict rules about what may be sold as a mains cable, but the vast majority of 'audio' cables have none of the required approvals. Maybe they are safe, maybe not - the rules are in place for good reasons, and based on the cost of some, paying for approval would be a drop in the bucket.
Because audio 'quality' isn't something tangible for most people, it's an area where the unscrupulous can dive in and make their outrageous claims, with little chance that they can ever be disproved. In many cases, it's only rational thinking and measurements that can really determine if something is going to make an audible difference to your system. In reality, everything that's part of the hi-fi will make a difference, and the only argument is whether it's audible or not. The charlatans will zoom into the fact (and it is a fact) that everything makes a difference, but they conveniently forget to mention that the difference will never be audible to anyone with normal hearing (or in many cases cannot even be measured). Some changes are immeasurable - we know that a difference exists because we can calculate it, but even the best measurement tools are unable to resolve the infinitesimal changes that will be made.
The limits of audibility are somewhat fuzzy - they change from person to person, day to day, and for various reasons. Hearing isn't just about ears - our brain and what we see makes a huge difference to what we hear (or think we hear). This provides a golden opportunity for anyone who is a bit shy of scruples to run rampant. Magic components, rocks, pebbles and holograms, cables that will transform your system ("better than room treatment" I've seen claimed) - the list is seemingly endless. One thing that we do know rather well ... humans cannot easily resolve a level difference of less than 1dB with programme material, yet if the scammers are to be believed, differences of 0.001dB (or even no difference whatsoever) are clearly audible to anyone who doesn't have tin ears. By applying this 'logic', it becomes easy to classify anyone who doesn't hear the 'magic' as being half deaf and not credible as a commentator on the topic. The cult followers will often use this very argument to try to discredit anyone who disagrees with their nonsense.
One of the major problems is that almost zero of these so-called 'improvements' have ever been properly tested. That means a full double-blind test, where no-one knows which component is in circuit at the time of the listening test itself, and the details are revealed and statistically analysed after everyone has finished. Our senses are too easily fooled to allow sighted tests, because there will be preconceptions and (sometimes subconscious) bias towards one test item and against another. This is perfectly normal - we all do it, every day. What is not normal is that these biased views are then claimed as 'fact' and brandished around on the Net.
The result is quite predictable. People with relatively little knowledge look up to reviewers and others who claim to be 'gurus' or (lower-case) gods in the field, and if they tell porkies (lies) most are unable to detect that what is claimed is simply not possible. One only has to look at the number of complete scams that abound for 'energy savers' (as but one example). They are routinely shut down, only to open up again with a different name, but the same old scam. Quack medicos do much the same with 'miracle cure-all' products that may not even contain a single molecule of anything that might have some medicinal value.
The counter-arguments to double blind testing are so trite that they are laughable ... "The extra circuitry in the double-blind switch box (or whatever) adds (or takes away) so much extra detail (or colouration) that it makes the test meaningless." My comment on that ... bollocks! The second argument dragged out regularly is that double-blind testing is 'stressful', and the added stress means that people are unable to discern differences they otherwise might find easy. The arguments (naturally) are intended to deflect attention from their preferred test methods, which are fatally flawed.
There is another counter-argument that's often dragged out, kicking and screaming. You will be told that you must keep an open mind, because some things just work for no apparent reason. They will tell you that 200 years ago, science stated categorically that people would not be able to fly. What they fail to mention is that science has come a very long way indeed since then, and while there are still things to be discovered, they will almost certainly not be anything to do with hi-fi.
Since they have provided what they think is the perfect reason for you not to trust science, it is expected that you should try whatever idiotic 'tweak' is being suggested with a completely open mind. To do otherwise is being closed-minded and not willing to try something new. This occurs in audio probably more than in any other field, and is complete bollocks (again).
If I were to claim that listening whilst having your wedding tackle partially immersed in olive oil improved the sound, would you do it? No, nor would any other sensible person. You wouldn't do it because it makes no sense. There is no connection between your naughty bits and sound quality, and you would rightly dismiss my claim as twaddle. Now, consider that your hearing is directly affected by many emotional triggers, as well as alcohol, many 'illicit substances', whether you had an argument with your partner, etc., etc. Dangling your privates in olive oil might actually make more difference to the perceived sound that you would have imagined, but it's still silly and you wouldn't do it. Why is it different when other equally silly ideas are proposed (such as demagnetising non-magnetic items like CDs or vinyl)? There is no difference - silly ideas are silly ideas regardless of the particular type of silliness.
One of the primary issues with these myths is that they create FUD - fear, uncertainty and doubt. Most people do not have the detailed knowledge needed to be able to determine that the latest craze or tweak is nonsense, implausible or just plain wrong. Reviewers who should be truly impartial are often anything but, and instead of dispelling myths they help propagate them. This is unforgivable in my view.
There are two particular things to which one can easily fall prey - the 'experimenter expectancy (or bias) effect' and the 'placebo effect'. Both are potentially very powerful, and can shape the outcome of a test at the subconscious level. If you 'demagnetise' a nonmagnetic medium and expect to hear a difference, then you probably will. What actually caused the difference will be curdled by your brain (at a subconscious level), and you will be left thinking that demagnetising made the difference, when in fact it was 100% imagination. This is why all proper medical tests are double-blind, to guard against these well known phenomena. It is a BIG mistake to think that you are immune - no-one is immune because we don't even know it's happening.
Not all myths are to do with magic components and indefinable qualities that are imparted by this or that tweak or incomprehensible magic act. Many are just plain nonsense, and although covered elsewhere on the ESP site, they will be referenced here too.
I've also come across some fascinatingly deluded articles, including one that explains why the writer is a subjectivist. This person actually believes that what s/he thinks s/he heard is real, and castigates objectivists with all the same tired old bollocks that we've come to expect. This is denial, plain and simple. Some people seem to be completely unaware of the huge traps they set for themselves with sighted tests - they think that normal reality can't possibly apply to them. "If I heard it, then it's real (to me)" is common enough, and in a sense it's also true enough. However it fails to accept that their 'reality' can be completely imaginary.
Something to ponder ... A truly open mind has to be open to the possibility that a new and radical idea, however exciting, may prove to be complete bollocks. [ 7 ].
Loudspeakers are available today that are capable of truly insane amounts of power - or so one may believe from the manufacturers' literature. There's only one small problem - they can't really handle the claimed power at all. The setup used by the maker to determine the power rating is often not disclosed, and in some cases has little or nothing to do with the way the speaker will be used.
The published figures are usually accurate, but only if the loudspeaker is used in much the same way as when it was tested for power handling. If you use a different box design (perhaps bandpass), then all bets are off, because the cone movement is severely restricted so cooling is reduced dramatically. In some cases the maker will cheat too - if the test bandwidth for a woofer is extended to 20kHz, that makes the power handling figure look a lot better because there's a lot more energy in the pink noise test signal.
This inflates the power handling figure, because the high frequencies are incapable of generating current in the voicecoil due to its inductance. Remember that if the maker can claim an extra 3dB, that means that a 300W speaker is suddenly a 600W speaker in the sales blurb. The fact that much of the applied voltage does not cause a corresponding current to flow seems to be immaterial.
For example, one well known 8 ohm loudspeaker has an impedance of over 16 ohms at 1kHz, rising to over 30 ohms at 3kHz. Needless to say there's also bass resonance, so the actual power is considerably less than that claimed. Power level is based on the use of band-limited pink noise, and is calculated using the RMS voltage and minimum impedance [ 1 ]. Already this gives an overly optimistic result, because there is no requirement to measure voicecoil current nor to use that in the calculation. Note too that the minimum impedance (Zmin) is not the rated impedance - for an 8 ohm driver it's typically around 5.5 to 6 ohms. The test method also states that power handling shall be tested in free air, so close to optimal cooling is available.
Because of the speaker's impedance curve, it's likely that the actual power (as opposed to claimed power) may be reduced by half. This means that a 600W driver only really handles perhaps 300W during the test, and if the speaker has a particularly high impedance at resonance it may well be quite a bit less. If the bandwidth is not limited to the speaker's actual frequency range (e.g. extended to 20kHz for a bass driver), the voltage measured during the test will be far greater than that which can be utilised by the speaker. Again, this make it appear that the speaker can handle 600W, but in reality the real power level may be closer to 150W.
To give you an idea of how deceptive it is to extend the noise bandwidth, I ran a simulation so I could see for myself. Extending the noise bandwidth from 2kHz to 20kHz will increase the applied voltage by around 2dB with band-limited pink noise, but the power that the driver actually receives is only increased by 0.2dB because the impedance is too high at the upper frequencies where we gained the extra voltage.
So, real power, based on the applied RMS voltage and RMS current might increase from 300W to 312W, but based on the RMS voltage it appears to increase from 300W to 480W. The apparent power is enhanced yet again by using the minimum speaker impedance rather than the nominal value. Now our 300W driver is rated for 630W based on Zmin of 6.2 ohms. That's impressive - the speaker rating has just been more than doubled by messing with a few numbers, and it didn't cost a cent to develop. Even more impressive, we have complied with the AES standard to the letter and still managed to double the real power handling without a spending a sausage for research.
Should the user think that 600W is the real power (after all, it says that in the brochure) and then adds an extra 3dB for headroom, a speaker that can really only cope with 150W is hooked up to a 1.2kW amplifier. The end result is inevitable failure. For more on this topic, see the Speaker Failure Modes article, and also have a look at Loudspeaker Power Handling Vs. Efficiency.
The whole idea of a loudspeaker being able to handle as much power as a bar radiator and not get so hot that it fails is simply silly. Manufacturers will persist with fudged 'power handling' figures for as long as people buy loudspeaker drivers based on power handling rather than efficiency. Always remember that an extra 3dB of efficiency is like getting double the power for nothing.
Just before this article was published, I came across some (dis)information on a site that really should know a lot better. The following is a direct quote ...
"Many speakers have a 'maximum wattage rating' on the back. Treat this as a 'minimum wattage rating'. You are far more likely to damage a speaker giving it too few watts and trying to play it too loud. High-end amplifier companies make amps with more than 1,000 watts, and you could plug in a $50 speaker into it with no problem."
While the last part of the quote is obviously true, what wasn't mentioned anywhere was that the 1kW amplifier would fry the $50 speaker in seconds flat if someone were to turn up the volume. The same article also claimed that amplifier power ratings are meaningless. While this is true of many cheap HTIAB systems, if a well known (and serious hi-fi/ professional) maker states that the amp can deliver 1kW, then there is usually little reason to doubt it. Note the old chestnut - small amps kill speakers. They don't (see the articles referenced by the above links). If a small amp driven too loud can kill the speaker, a bigger amp driven too loud will kill it faster.
In addition, there wasn't even the smallest mention of speaker efficiency anywhere in the entire article. Remember, if one set of speakers is 3dB more efficient than another, that's exactly the same as getting double the amp power - free. Assuming of course that the more efficient speakers still sound decent. There's no point getting the most efficient speakers you can if they sound like a cat farting into a milk bottle.
The same author as the above quote also claimed that loudspeaker frequency response figures were meaningless and should be ignored. Yes, this is true for the HTIAB systems that all claim 20Hz to 20kHz (but with no graph or dB limit), but for serious speakers, most makers go to some effort to demonstrate that their speakers really do what is claimed. While fudged figures are not uncommon, to make a blanket claim that all are meaningless is going much too far - even for an article aimed at ordinary consumers.
Rather than reiterate what I've already written on this topic, I suggest the reader looks at The Truth About Cables ... first. Despite almost everything you read about this or that cable 'transforming' your system, it won't. Nada, zip, not a sausage. The hype and BS surrounding pieces of wire is astonishing, and there are crooks all over the world ready and more than willing to relieve you of your money.
Should you be so wealthy that $10,000 is small change or pocket money, then you probably don't care one way or another. However, if you are part of the real world then you should resent the fact that thieves and charlatans are relieving others like you of their hard-earned cash. I really dislike crooks, and in my book anyone who claims that their cable is capable of doing anything more than transporting your audio from point A to point B is a both a fraud and a liar.
In essence, that is all a cable ever does, and if it's designed and built properly it will do exactly that ... transport your audio from point A to point B. Nothing more and nothing less. Yes, there are losses - always. These are utterly unimportant for signal leads provided some common sense is used. 5km signal leads using cheap shielded wire is not sensible, but the vast majority of interconnects are perfectly alright for the job. You don't need to spend more than perhaps $20 or so to get decent signal leads, and those costing $hundreds will not do anything differently - despite all claims.
Speaker leads can be more of a challenge, but that's all about keeping resistance and inductance low. Resistance means that power is lost, and inductance means that high frequencies can be affected. The simple answer is to keep speaker leads as short as possible, and preferably make your own. I've seen speaker leads selling for not $hundreds, but $thousands, and that defies all logic. There isn't a speaker cable made anywhere, by anyone, that is worth that kind of money. In my book, anything over $20 or so for 3 metres of unterminated wire is looking suspiciously like a scam. Terminations can be fairly expensive (particularly for those with a gold plating to prevent corrosion), but even these shouldn't cost more than perhaps $10-20 a pair.
To make matters worse, some cables (especially those with a low characteristic impedance) can cause amplifiers to oscillate - definitely not something anyone wants. The fix is easy - add a terminator to the far end, using a series network of a 10 ohm resistor and 100nF capacitor, wired across the speaker terminals. Some of the charlatans will offer to charge you serious (additional) money for a terminator, which should be included as a matter of course. IMO, this Zobel network should be included on speakers - it adds almost nothing to the cost, and ensures that cables with excessive capacitance don't harm the amplifier. A Zobel network will not influence the sound, regardless of claims you may hear.
In reality of course, all cables make a difference that might be measurable - especially at super-audible frequencies or radio frequencies. Unless you compare bell-wire with 5mm² cable of sensible construction, the difference will rarely (if ever) be audible in a double-blind test. Sensible speaker cables are readily available for a few dollars per metre, and anything else should be treated with suspicion.
These are another matter entirely. Most are a blatant rip-off, that much is predictable, but a great many (probably most) are also likely to be illegal. They often have no fire rating certification, and/or are otherwise ill-advised additions to your system. In Australia for example, it is mandatory that all mains cable designs are tested and certified for safety. In the US, some insurance companies may deny a claim if it's thought that a non UL-certified cable started a fire. Elsewhere in the world other regulations apply, but the mains cord sharks don't give a toss!
Who cares if little Johnny is electrocuted, as long as the hi-fi system sounds great? Well, I do, and so do the authorities. It has to be considered that anyone who cares so little for your wellbeing that they will tell you (with a straight face) that 1 metre of their 'magic' crap will undo the alleged damage caused by possibly hundreds of kilometres of perfectly ordinary wire simply cannot be believed. Good grief! Many electricity suppliers use aluminium cables for high voltage transmission, and that sounds dreadful (so they say). No-one seems to be concerned about aluminium voicecoils in loudspeakers though. Ditto for ribbon tweeters, which almost all use a thin aluminium ribbon as the drive membrane - as do many ribbon microphones in the recording studio. Strange, that.
The gall and audacity of these sharks to claim that 1 metre of their magic mains cable will make an audible difference! These claims are simply deluded. Read the introduction section again - the only way to be certain is to perform a double-blind test, and anyone who claims otherwise is lying - straight and simple. Good quality connectors that make firm and positive contact are worthwhile, but the rest is horse-feathers.
If you have a problem with mains noise, a quality mains filter might help to reduce interference. You don't need to spend $hundreds to get one.
"Second harmonic distortion is pleasing to the ear" say those who enjoy their little distortion boxes called SET amplifiers. If only that were true, we could all relax and stop worrying about all the intermodulation products that these pointless atrocities produce and just enjoy the music.
Strangely, every high quality VCA (voltage controlled amplifier/attenuator) and anyone who uses FETs as the active element in limiters and compressors, will include distortion cancelling circuitry. Such circuits only reduce even harmonics (second, fourth, etc.), but are unable to reduce odd harmonics at all. We are left with a distortion cancelled circuit that only produces odd harmonics, because it's difficult to reduce all distortion to zero, and a bit of odd harmonic distortion is far less intrusive than a lot of even harmonics. The odd order harmonics that remain cannot be removed by distortion cancelling circuits, but removing even harmonics reduces overall distortion to (usually) acceptable levels.
When valve amplifiers were all we had, not one amplifier that had any pretense to quality used single-ended triodes. Most mantel radios of the day used a single-ended pentode output stage, but that was for a simple (AM) radio with no pretense to hi-fi. When more power and/or higher quality was needed the output stage was invariably push-pull. Push-pull output stages cancel even harmonics, and also allow the full use of the laminated iron transformer core (which is dramatically less effective in a single-ended stage).
The remaining distortion consists of predominantly odd harmonics - there may be a small residual second harmonic content, but quality designs came close to eliminating it altogether. Feedback was used (albeit in moderation because of the output transformer) to reduce distortion as far as practicable. The very best amps of the day (at the end of the valve era) came very close to equalling a decent transistor amp.
So why do we have this myth that second harmonic distortion sounds 'nice'. I wish I knew. It's possible that it was started by a SET fanatic somewhere along the path to nirvana, but other silly explanations that I can't think of are probably equally plausible. Regardless of the origin, it's complete nonsense, and distortion of all kinds should be below 0.1% (system wide) to qualify as hi-fi. Less is better, and easy to achieve until you reach the loudspeakers. Speaker distortion is typically at least an order of magnitude greater than that from most competent amplifiers!
So, do we need opamps and power amps with distortion that's virtually immeasurable? No, not at all - however aiming for extremely low distortion doesn't hurt anything provided nothing else is sacrificed to get there. Almost always, very low distortion systems usually have wide bandwidth too - it's comparatively easy to get flat response from 1Hz to 50kHz or more.
In some areas there is a serious prejudice against opamps. I don't have a problem with DIY people wanting to build a discrete opamp - indeed, I even have a PCB for one. This is an excellent way to learn about circuits and how they work, and a discrete opamp can perform very well, within limits. For anyone to claim that traditional (IC) opamps are grossly inferior in some way is just silly - there are opamps available that beat anything you can build hands down. Still, there is often fierce debate about which opamp sounds better, but almost always with no reference whatsoever to a blind test to prove it one way or another.
A claim that I find interesting is that many or all opamps sound 'better' if biased into Class-A. With few exceptions the reverse is true, because the additional loading presented by the bias circuit loads the opamp's output stage and increases distortion. The increase may be quite pronounced in some cases, and it's an idea that doesn't tally with reality.
As you are no doubt aware, reality and fantasy are worlds apart, and people making claims should be willing to back up their story with proof. Sadly, few even attempt to do so. It really doesn't matter if 1 person or 10,000 people think there's a difference - if there's no proof then it has to be likely that there is no difference. Proof is defined here as statistically significant results based on double-blind tests, or measurements that show the difference is measurable and within the limits of audibility as we currently understand them. Don't expect to hear of some miraculous discovery that changes what we know about human hearing - there are exceptions (primarily government standards based, and often quite wrong), but that's not at issue.
Despite claims that there are hearing mechanisms that are not well understood (which may or (more likely) may not be true), it has been established over many, many years that normal people with normal hearing are completely unable to pick most differences that a reviewer might claim is "astounding" (or some other superlative). Nor can most people hear the 'veil' over the high frequencies, or be convinced that the bass has the 'authority' claimed - beware of words that attempt to convey emotions, as they are the emotions of the reviewer, but usually no-one else.
There are things that we hear that are not thought to be audible. Countless people listen to MP3 audio, but most don't seem to have noticed that the stuff that we allegedly can't hear is the very stuff that provides the stereo image. Listen to the same track direct from CD and then as an MP3 - imaging is gone, and you are left with a mostly mono signal with some left and right highlights every so often. Digital radio is the same - for ages I thought my DAB+ digital radio had only a mono output, until one day I heard something that was panned hard left.
The above notwithstanding, most well-engineered music (if you can find any these days) sounds extremely good. On my system, this is despite the fact that I haven't used a single magic component anywhere. 
TID (aka TIM) was proposed by Matti Otala in 1972, and the basic concept is 100% true. Unfortunately for the proponents of TIM/TID, it doesn't actually happen with real music in any reasonably competent amplifier (which is almost all modern amps, including IC types). Many have tried to demonstrate its existence with programme material, but to my knowledge no-one has ever managed to succeed. The information supplied in Wikipedia [ 4 ] is untrue - no known amplifier shows the problem with normal programme material.
Virtually any amplifier can be made to show the 'problem' - all you need is a low-level high-frequency sinewave superimposed on a fast risetime low-frequency squarewave. If the squarewave's rise and fall times are fast enough, no audio amp ever made (including those with no feedback, valve amps, etc.) is fast enough to prevent some loss of the high frequency sinewave signal. A fast squarewave may easily demand a bandwidth of several MHz - well beyond any realistic expectations for an audio amp. Once the squarewave is filtered so its bandwidth is more in line with a typical full range audio signal, TIM/TID simply disappears.
Of far greater concern is the amplifier clipping - even for an instant! All frequencies other than the one that caused the amp to clip are eliminated once the amplifier is no longer operating within its linear range. Despite this, occasional transient clipping is generally considered to be inaudible under most conditions.
The range of things that one can buy that will allegedly improve their hi-fi system is mind-boggling. There are rocks, pebbles, holograms, feet, weights, springs, 'special' capacitors, 'special' lacquers that match the human body's carbon content (I truly wish I could say I made that up, but it's true), audiophool knobs (yes, that's true too - over $200 for a timber knob!), carbon composition resistors, 'demagnetisers', little 'towers' to keep your speaker leads off the floor ... the list goes on and on. Pretty much without exception, these are scams. If someone wants to believe that a rock on top of their speaker makes it sound 'better' then fine - put as many rocks as you like on the speakers.
It's when others insist that the rock is 'special' (and costs hundreds of dollars) that the claims become criminally fraudulent. Other parts that supposedly have magical properties are often simple passive components. While there are certainly differences, if an appropriate part is used in a circuit (rather than something completely unsuited to the job), the likelihood of audibility is usually nil. The same applies to cables of course - they have already been covered in several ESP articles as well as above (albeit briefly).
These are an especially good target for the scammers, because most people don't actually know much about them. They have an air of mystery, so it's easy to claim that polyester caps sound 'bad', damaging the 'air' around instruments for example. When you see the claim, try to get a meaningful explanation of exactly what is meant by 'air'. Don't expect anything that makes sense.
Polypropylene and other mildly exotic dielectrics don't seem to have attracted the wrath of the magic parts proponents, but they are comparatively large compared to a polyester or Mylar cap of the same value, and don't fit on many PCBs (such as those I sell). There is no credible evidence that any film cap is aurally different from any other. Tests have been run and various distortion products measured and identified, but in all cases the results are extremely difficult to measure because they are below the noise floor of most test equipment. If even purpose-built test equipment can't identify a significant difference, then it is folly to imagine that we can hear it. Consider that if a capacitor is (much) physically larger than another type of the same value then it may act as a relatively large section of unshielded wiring, and may pick up noise - hardly an improvement.
There are certainly differences between plastic film capacitor dielectrics, but nothing that need concern us for audio. If you happen to be making a high resolution sample and hold circuit then the choice is critical, but none of the effects are relevant to dealing with audio signals. I have seen it claimed that ceramic caps shouldn't be used for supply bypass because they somehow affect the audio quality - this is utter nonsense, and anyone who claims this to be true is either a fool or a liar. Multilayer ceramic caps are specifically designed for bypass applications!
Even electrolytic caps are perfectly usable in the signal path of most audio gear, and if large enough their limitations will never cause any problems. There is a long-standing myth that you have to bypass electros with a small film cap, but this is also nonsense. The inductance of any capacitor is simply a factor of its physical size, and small sized caps have low self-inductance. If you are working with RF equipment, then yes - add the bypass cap, otherwise it's optional. It won't hurt anything though, so if it makes you feel better that's fine.
The exception for electrolytic caps is their use in any kind of filter. When the AC voltage across a cap is significant, it is able to distort that signal if there are internal non-linearities. Electrolytic and many ceramic caps certainly have non-linear behaviour (usually both voltage and temperature dependent), but if the voltage across the cap is close to zero, then distortion is also close to zero. Using bipolar electrolytic caps in passive crossover networks is a bad idea, because they degrade with time - especially if the system is pushed hard and the caps carry significant current. Add to this the voltage dependent distortion characteristics and you have a non-linear system that can't be relied on. High 'k' ceramic caps have no place in the audio path, but are perfect for supply bypass (and that's what they are made for).
Capacitors are much maligned by many in the audio field, but I know of no double-blind test where listeners have been able to pick them apart. I specifically exclude multilayer ceramic and electrolytic caps from this because it's far too easy to make the difference audible by using nonsense circuits that are designed to reveal any flaws, but are not used in normal circuits.
Most capacitors have some distortion, but for almost all film caps it is so far below audibility that you don't need to worry about it. Even electrolytic caps are fine as long as the AC voltage across them is minimal. It stands to reason that if there is no significant voltage across any capacitor, then it can contribute no significant distortion.
Carbon composition resistors are useful for one thing - the rubbish bin! They have no place in audio equipment because they are noisy and unstable with time. As with most of the other scams, I can't imagine how this started - it just doesn't make any sense. Way back when electronics just started, carbon comp resistors were the mainstay of low cost resistors, but they are now outdated and stupidly expensive for what is really a pretty crap component. Metal film resistors are far cheaper, are much more stable, and have better tolerance (typically 1%). Carbon film resistors are better than composition types, but not as good as metal film. Audibility for any of the above? Almost zero provided the noise of each sample is not audible.
Wirewound resistors are used where high power is needed, and contrary to popular belief they generally have a very low inductance compared to resistance. While it is possible to see a measurable change in performance due to the inductance, it is unlikely that the difference will ever be audible because the inductance is so small. Some 'non inductive' wirewound resistors are just the standard version with a different marking - they don't use a non-inductive winding at all (but they do cost more).
A few years ago, a bunch of lunatics launched a speaker crossover that completely did away with evil capacitors, and used only nice, friendly inductors for the whole network. I managed to obtain the schematic and was able to simulate it, and to say the results were dreadful would be high praise. The results were worse than dreadful - it was an abomination.
What the 'designers' missed completely is that inductors are the worst electrical component of all - inductors have self-resonance that's much lower than any sensible capacitor, and they are lossy because of the wire resistance. So-called inductors are only a useful inductor over a relatively limited frequency range, and their internal resistance ruins damping factor ... if you happen to think that's an important parameter. Naturally, all resistive losses result in heat (usually inside the speaker cabinet) and wasted power.
Like so many other silly fads, the all-inductor crossover seems to have mercifully passed away, but not before it created much controversy and instilled FUD in some sectors of the hi-fi fraternity.
If you doubt that inductors are as bad as I say they are, run a passive crossover at reasonable power into dummy loads for a while, with a full bandwidth signal. Feel the capacitors - they should be at room temperature. Now the inductors. They will be far hotter than the caps. The heat is wasted power, and indicates that the inductor also has significant resistance.
There are countless myths that fall into the 'miscellaneous' category, and those shown below are just a sample. These are some of the more popular distortions of reality, but the number continues to grow, with 'new' BS 'products' being introduced all the time. It's clearly impossible to keep track of them all (and who would want to?), but those shown here have been around for a while.
Residual magnetism in component leads causes either distortion or something completely unexplainable (but apparently bad anyway), according to some. This is unmitigated drivel - it doesn't happen in any competently assembled equipment, and is just an idiotic claim designed to separate you from your money. If this were true, then that would be a known distortion mechanism of loudspeakers - they have extremely powerful magnets. So do magnetic phono pickups, guitar pickups, and various others used for electric piano and other instruments. Surprisingly perhaps, there is a magnetic distortion component in loudspeakers, and it could be eliminated entirely by removing the magnet completely. This would naturally mean that the speaker would no longer work, but surely this is a small price to pay. 
Even worse than the original claim is the range of products that are allegedly designed to demagnetise the leads in question. These products do not work at all - they can't, because it's impossible to get enough current through the components to do anything even remotely useful. Demagnetisation goes a lot further though ... if you demagnetise a CD or vinyl disc (which both have close to exactly zero magnetic material) they will sound better. It's hard to even waste time on claims like that, because they are so obviously and blatantly false. A non-magnetic material cannot be magnetised (because it's non-magnetic) and therefore, it cannot be de-magnetised because it was never magnetised in the first place.
The magnetism bogey-man seems to be fairly popular at the moment. I find it fascinating that depending on what you read you will discover that (electro) magnetic fields are either the most efficacious miracle cure-all known, or are evil and will cause your internal organs to collapse into amorphous cancerous jelly (I may have exaggerated the latter claim a wee bit ). I even saw an advertisement for a CD (yes, a CD) with 'special' demagnetising tones recorded on it (at least that's what I surmise) that will (astonishingly!) improve "transparency, dynamics, details, soundstage, and all other parametres" (sic). As an Aussie comedian was heard to say often ... "I see it, but I dooon't believe it."
Another one that's guaranteed to get the lunatic fringe-dwellers on their soap-boxes, shouting loudly, is break-in. Special boxes that produce an equally special signal will break in your leads faster than just listening to music ... we are told. Again, this is almost completely bollocks too. There are a few components that do change characteristics over time (such as speakers), but it happens so slowly that we will never actually hear the difference.
Our audio memory is notoriously short, and it is simply impossible to hear a change that takes weeks to occur. What really happens is that we become 'acclimatised' to the sound - there is rarely any significant change at all. This is doubly true of cables - there isn't any reason whatsoever to break-in an interconnect or speaker cable, because they don't change enough to create a measurable change, let alone one that's audible.
With no exceptions that I can think of, electronic equipment only needs to reach normal operating temperature for everything to work as it should. In most cases, the temperature doesn't even matter. After sitting in warehouses and/or on the dealer's shelf for a few months, electrolytic capacitors might need a minute or two to polarise themselves properly. It takes a short while before the leakage falls to its normal value. The sound doesn't change during this process.
The claim is made in countless forum arguments that "sinewaves are too simple to get a useful measurement". It is true that a sinewave is simple - it is a mathematically pure tone, containing exactly zero harmonics. Because of this, it makes it relatively easy to measure tiny amounts of non-linear distortion in any audio product. Sinewave testing also shows audible distortion, well before it can be heard in most music. With a pure sinewave, it's possible to hear 0.5% THD (total harmonic distortion) or less, depending on the speaker used to monitor the results and the room acoustics.
The reason that sinewaves are used for testing is that the waveform is so clean that any modification to the original signal is easy to measure. Some will protest (often vociferously), but quite frankly, they are wrong. Supposedly 'simple' sinewave testing is still far and away the easiest way to quantify and qualify distortion - the exact nature of the distortion is revealed to anyone who knows how to conduct tests properly.
An amplifier doesn't actually care if the input signal is a sinewave or something more 'complex'. While the claim that sinewaves are 'simple' is true up to a point, amplifiers don't actually have any idea what they are amplifying. An instantaneous value of voltage is amplified by the amp's gain, to produce an amplified version of the signal for that moment in time. If a signal happens to change too fast, the amp cannot keep up and some information is lost. If the input signal is too high, the amplifier will clip and some information is lost. Distortion is generated in both cases.
In reality, no normal audio signal can change fast enough to trick any competent amplifier, but this has never stopped the pundits from claiming it happens anyway. The idea of TID has been bandied around for ages (see above for more), yet no-one has ever been able to name an amplifier that suffers from it. It's easy to demonstrate with (perish the thought) test equipment, but the test has to be modified to account for real-world audio signals. This isn't done, but there are those who still claim that the results are relevant. Sorry, but they are not!
There are many people who insist that phase shift is evil, and it must be eliminated from a system for it to sound any good. The actual complaints vary, but there is general agreement amongst those who think it's bad that it really is bad, in any number of ways. In reality, like many other things, phase shift is a fact of life. It's actually generally harmless unless the amount of phase shift varies cyclically - this is only ever found in effects pedals used by guitarists and the like, and never happens in any amp or preamp. Phase shift may also create audible artifacts if it's different between two channels of an amplifier. This is theoretically possible, but extremely unlikely unless the amp/preamp (etc.) has been modified by someone incompetent.
It's easily demonstrated (and used as 'proof' by those who think it matters) that inverting a signal changes the sound. With instruments that produce an asymmetrical waveform (many woodwind and brass instruments, human voice, etc.), inverting the signal often makes it sound different. The problem then becomes "which is right?". The unexpected answer is "both" and "neither". The first answer is because there simply isn't a 'correct' polarity. No-one knows how many inversions the signal may have been subjected to before you hear it. The answer is also "neither" because no reproduction can ever return the original performance, and to imagine it can is pure folly.
Some generated waveforms (a sawtooth for example) usually sounds 'different' depending on its polarity. As far as I'm aware, no-one knows exactly why, but it's very common phenomenon. There is no 'right' or 'wrong' polarity, because it's an electronically generated waveform. The real test is to listen, then leave the room while someone else either changes or doesn't change the polarity, then return and listen again. In the vast majority of cases, you will be unable to determine whether the polarity was changed or not.
The complete rubbish that you'll find about the alleged superiority of 'valve sound' over evil little transistors is astonishing. In almost all cases the reverse is true - a competent transistor amplifier will usually murder even the best valve amps for overall quality. There are some extremely good valve amps (mostly from the end of the valve era), but none can be recommended any more because the available valves are now generally well below the quality that was routinely produced by mainstream manufacturers of yesteryear.
The standard explanation for the so-called superiority of valves is that they produce predominantly second harmonic distortion, but this is simply untrue for any competent design. The best of the late 70s valve amps had very low distortion - not as good as a decent transistor amp, but much lower than most of the more recent attempts. As described above, having great gobs of second harmonic distortion is nothing to crow about - it's a good reason for the 'designer' to hang his head in shame though.
Left to their own, transistors will also show predominantly second harmonic distortion. The job of the designer is to remove as much of this (and indeed, all types of distortion) as possible, while keeping the final circuit manageable in terms of cost and complexity. There are actually many factors, and distortion IMO is not something that should be considered a virtue. I suggest that anyone who wants to look at this more closely should read Amplifier Sound and also look through the valves section.
There have been countless claims that measurements don't cover all contingencies, or we don't know how to measure certain things, and therefore (all) measurements are pointless and don't give us the full story. This logic then goes on to conclude that since we have rendered measurements useless, we can therefore state with authority that only subjective tests are of benefit. This is, of course, nonsense.
Along similar lines you may hear claims that hearing resolution (of a select group of people) is better than any test instruments, and can pick up the details that measuring instruments cannot. Proponents of this school of 'thought' never consider the experimenter expectancy or placebo effects, and many in this group will claim to be 'immune' to these effects because they have done it for years and know how to avoid the traps. Utter garbage ... no-one is immune from these effects, and to claim immunity is to be in denial of reality.
Measuring instruments have come a very long way over the past 60 years, and they can resolve details that are completely inaudible to anyone - regardless of their 'golden ears'. There are techniques that simply subtract the original signal from the amplified version, so any difference can easily be detected. If the amplifier or preamplifier has any problems, the output from the subtraction circuit will be non-zero and easily identified. This process was first described by Peter Baxandall in 1979 or thereabouts (see 'Null Testing' below).
Measurements don't cover everything though, this much is true. We have no way to measure sound-stage (the apparent placement of instruments in front of and between the speakers), but we don't actually need to. Provided the signal path is clean (minimal distortion), has a flat frequency response across the audio band and both channels have equal phase shift, we know that it's not 'damaging' the signal. If the signal can get through our systems properly and without significant modification, then there is no reason that the soundstage will be better or worse than what was recorded.
This latter point is missed by most reviewers and most of the magic component cult followers. There seems to be an understanding that in order to get 'good' sound, your system needs to be made up from the most inconvenient and expensive parts available. An amplifier that won't burn your fingers isn't worth listening to, and all internal components must be physically much larger than whatever you used before, and at least 5 times the price. The same 'logic' affects everything else. No matter how good your CD player might be, there is always a modification (that uses expensive, large and inconvenient parts) that will make it so much better, and the same applies to everything else in the system.
There is a measurement technique that shoots down all complaints that "sinewave testing doesn't show what an amplifier does with a complex signal". As mentioned briefly above, the original and amplified signals are summed, with one inverted and scaled so it is exactly equal and opposite the other. The result is a null - one of the easiest things to verify. The smallest difference between input and output is immediately audible (or visible if an oscilloscope is used), and this technique demonstrates that most amplifiers can handle any audio signal that comes along. Remember, the smallest difference between the signals shows up clearly, and null testing can be used with any normal full-range signal.
There is a version of just this that I called the SIM (sound impairment monitor). The version that I've used several times simply looks at the signals on the input and feedback nodes of the circuit (typically the bases of the input long-tailed-pair). Should be amplifier be unable to handle the rate of change of the input signal, it shows immediately. Should the amplifier even approach clipping or show any non-linearity, again it shows immediately. Noise, power supply ripple, slew rate limiting - all show up very clearly, and are an instant indicator that the amp can't cope.
Tests I've done show that every amp I've tried it with is perfectly happy amplifying any audio signal I can send its way. Likewise, I can push any amplifier with a my squarewave generator and see that none can handle the extremely fast rise and fall times. This doesn't mean that every amplifier is flawed, it just means that a squarewave test is inappropriate for determining audio performance. Experienced technicians will use squarewave testing for other purposes though, and it's a very quick and easy way to check tone controls and equalisers (for example).
Subjectivists seem to abhor all measurements, and signal null testing is a measurement. Therefore it cannot be used to prove a point, because it's a measurement and thus is automatically inferior to a sighted listening test.
At one stage, you could buy (for an insane price of course) a wooden knob that would supposedly transform your hi-fi. Yes, you read that correctly - a knob. Not a high quality replacement pot (potentiometer/ volume control), just a knob. I have no idea if anyone fell for this scam, but I expect there would have been a few takers, even at $485 - I kid you not. Needless to say, changing a knob from plastic or metal to wood will make absolutely no difference to the sound, but that obviously didn't disturb the criminals selling and promoting it.
Along similar lines, there is an alligator/crocodile clip that you clip onto leads and such - again, this will supposedly work wonders. This is no ordinary clip though - it's a Quantum Clip, and "is capable of manipulating certain inanimate material into a condition that mimics the quantum state of our living senses". WTF!! What insufferable, unbelievable crap!. The purveyors of this garbage belong in prison for fraud. Personally, I don't care whether they believe this shite of not, they are common criminals and nothing more. Everything (and I do mean everything) they sell is nothing short of fraudulent. The cost of the supposedly 'quantum" clip - apparently it's £500 (British Pounds)! This for perhaps 50c in materials.
As for so-called audio review sites and 'independent writers' who support this drivel - anyone who gives anything other than a big thumbs-down to the frauds cannot be trusted to review a soiled baby's nappy (diaper), let alone hi-fi equipment. I'm almost ashamed that I live on the same physical planet 
.
The normal fuse supplied with your system can't possibly sound any good, but that's easily fixed. Yes, you can buy true 'audiophool' fuses to prevent the inevitable congestion as the current has to flow through that tiny little wire. A bargain at only $60 each (give or take). Mind you, you'd expect to pay that for a 'hi-fi tuning fuse', because it's so much more than just a fuse. It's also a ... ahhh ... hmmm ... no, my mistake, it's just a fuse.
Audiophile power outlets? I'm kidding, right? Sadly, no I'm not, and as if that wasn't bad enough you should see the price - almost US$150 each. Mind you, they do appear to be a cut above the average in terms of build quality (so are so-called 'hospital grade' outlets in the US), but the price is just outrageous. In addition, you can even buy a set of outlet caps (special ones of course) for a mere $99 for a set of four. I'm unsure how they improve the electrical supply, but apparently they stop nasty EMI from sneaking in through the little holes where the plugs go, when no plug is inserted. They claim to be gold plated solid copper - perhaps they short all the pins together? That should make a nice bang. .
Blocking the little holes predictably does diddly-squat - EMI doesn't sneak in through the holes - it doesn't need to because it can get to the wiring so easily through most interior walls and anywhere else where wiring is not completely shielded, not to mention the wires out in the street and all the way back to the power station.
But not your ordinary boring room treatment that actually works. No, you don't need to do any of that when you can go off and buy a few little bowls (with wooden stands of course) that will allegedly convert a $200 HTIAB into respectable hi-fi (no, I'm not kidding - a reviewer claimed pretty much exactly that). A movie intro showed "grandiose differences" and the sound became "voluptuous".
Who wouldn't drop what they were doing and rush out to spend around $3k for a collection of little ornaments? You can get the bowls, balls, pebbles (with or without glass jar), rocks and all manner of accessories for less than the cost of a small car! These things actually treat your room, better than any conventional proper room treatment, and if the cat doesn't decide they are actually cat toys you should be in for a real treat ... or perhaps not 
.
There is nothing at all wrong with using balanced connections, but some take it to extremes. A balanced connection is designed to reduce common-mode noise, whether injected into the cable by nearby power cables or due to earth/ ground loops between separate pieces of equipment. There seems to be a school of thought that balanced connections sound better in some way. If using balanced cables and inputs/outputs removes hum or other noise then yes, the system will sound better. However, in most cases with a hi-fi system it makes little or no difference. There are exceptions, and if you find that you need balanced interconnects to remove hum, then that's exactly what you should use.
I have even had enquiries about using Project 09 in fully balanced mode. In other words, two P09 boards, with one used for the pin 2 (hot) lead of an XLR, and the other for the pin 3 (cold) lead. The opportunities for things to go seriously wrong are many and varied, and every passive part needs to me matched to better than 0.1% or serious CMRR errors will result. In addition, there will be more noise (opamp and resistor thermal noise in particular), and no 'improvement' to sound quality.
Professional/ studio mixers all have balanced inputs and outputs, but all internal circuitry is unbalanced with the possible exception of the mix busses. No-one has ever considered that each and every module within the mixer should be duplicated to maintain the balanced connection right through the mixer. Apart from anything else, the cost would be prohibitive in the extreme.
Balanced connections are used for long mic cables and interconnects between different pieces of equipment. Cable runs in studios are often very long, going to and from patch bays and other gear that might be physically separated by some distance.
For a home hi-fi system, if you cannot hear any loop-induced hum or buzz, there is no reason to use balanced connections. Contrary to what seems to be common belief, a balanced connection does not sound 'better'. Floating (unearthed) signal sources such as microphones don't actually need to be balanced, but they are almost invariably balanced for historical reasons. Many other sources (CD & DVD players, etc.) are floating because they are double-insulated, but are earthed as a matter of course via the interconnects. Again, a balanced connection is only needed if there is a hum problem when the device is connected to a preamp.
There is absolutely no need for speaker signals to be balanced, as the signal is low-impedance, high-level and the speaker is floating with respect to mains earth/ground. Using a BTL amplifier is only worth consideration if you need the power, but not to 'improve' sound quality.
Only very recently I was asked about thermal crosstalk in dual operational amplifiers (opamps). This (amongst other things) is very real, but it has to be understood that limitations such as this are only relevant for precision designs where the opamp circuit has very high gain, and DC offset is critical. Just like capacitor dielectric absorption (aka 'soakage'), there is no need whatsoever to consider this for audio. It's simply not relevant with the relatively low gain and bandwidth needed to transfer an audio signal in typical hi-fi applications.
Where thermal crosstalk and other electrical cross-coupling effects become important is in measurement systems (yes, the very systems that so many audiophools abhor), where very high gain, exceptionally low distortion and wide bandwidth are critical to ensure the measurement is accurate. With most audio circuits, the current and power demands on opamps are very low, and the effects mentioned are completely irrelevant. Despite claims to the contrary, small temperature variations across the die don't produce audible artifacts - they can be measured easily enough sometimes, but don't cause significant non-linearity. Unfortunately, the audiophools will sometimes pick up on very technical articles that they usually don't understand, and extrapolate this to define the reasons for certain opamps sounding 'bad'.
You see, only the most expensive and difficult to get opamps are suitable for audio in their opinion. More pedestrian types are obviously inferior, because ordinary people can get them easily and cheaply - that can't be good. In reality, some of the types that are claimed to be so obviously better in all circumstances may not really be suitable at all. Some will sacrifice noise for incredibly low input current for example, and while this may be an important consideration for scientific or laboratory equipment, it does not translate that it's therefore better for audio.
The same logic applies to many other opamp functions - there is a huge range of specified parameters, and the rules of design indicate that the designer should choose those that are important for the application and ignore those that are irrelevant. What is irrelevant in one design may be highly relevant in another, one of the reasons that there is a mind-boggling number of opamps on the market. While a particular opamp may be ideally suited to precision sample-and-hold applications for example, it does not follow that the same device is suited to a phono preamp or other audio applications.
As noted earlier, there is a belief that some opamps introduce colouration, despite the fact that measured response is ruler-flat and distortion is immeasurable with normal equipment. It's alleged that somehow these measurements miss the subtle effects that stand out like dog's nuts to those blessed with golden ears. A friend claims that he can hear a TL072 in a system instantly - said he heard one in mine, despite the fact that there aren't any. No-one has hearing so good that they can hear the difference between competent opamps, regardless of their claims to the contrary. If test instruments have difficulty detecting differences between opamps, you can rest assured that you will generally not be able to hear anything of interest. Claims that some opamps have better bass than others are just silly - all opamps can give their designed gain down to DC if allowed to do so, and no-one can hear that!
There is one condition to all of the above though - the noise floor of all opamps auditioned has to be well below audibility. Noise is often a clue, and in some cases the noisy part might be preferred as it can appear to have better top end. The noise may add a tiny bit of 'sparkle' that the listener prefers, without necessarily noticing that there is a background hiss (carbon composition resistors, anyone?).
I coined the term "Black Knights" to describe the cult followers - see the Monty Python sketch of the same name and you can work out the explanation for this yourselves (it's from "Monty Python and The Holy Grail"). They are in complete denial - science must be fatally flawed if it disagrees with their listening experiences, and therefore they have a propensity for throwing out the baby with the bathwater. They will never admit that they may have been tricked by the experimenter expectancy or placebo effects - what they think they heard is reality, and anyone who disagrees is just wrong. End of story!
Unfortunately for everyone, these off-the-wall opinions are touted as fact all over the Net, where they are picked up by others who use the fatally flawed arguments as backup for their own (equally fatally flawed) opinions. While we might hope that they would simply run in ever-diminishing circles and disappear up their own exhaust-pipes, they seem to gather mass and keep growing. This isn't helped one bit when formerly credible engineers apparently succumb to Alzheimer's and fall into the dark side.
Once someone starts sprouting utter BS about the "audibility of capacitors" they are no longer credible. In many cases apples are cheerfully compared with oranges and the 'comparison' is touted as reality - sometimes with parts stressed beyond their normal working limits. Unfortunately, most don't understand the reality behind these claims, and they gain acceptance as being real. Ultimately, it makes the world a poorer place, because proper investigation is derailed by the nonsense. The same goes for other audio nonsense, from green pens for the edges of CDs to silver cable in interconnects (or even signal transformers!). There is no credible evidence that any of the major or minor 'tweaks' will have any effect at all, let alone transform your system.
It's easy to dismiss most of the nonsense as harmless, but in reality it's no such thing. Countless people are duped into thinking that the rubbish posted is real, and once duped it's likely that they too will fall victim to the placebo effect. After all, no-one likes to admit that they have been conned, so will often (albeit inadvertently) jump onto the bandwagon as well. This perpetuates the belief that this or that tweak, rock, hologram or whatever has some benefit, when in reality it has achieved exactly nothing useful to the buyer.
A lot of the scams are enabled by the simple fact that no-one can actually define what 'perfect sound' really is. Innumerable speaker, headphone and amplifier makers claim to give you just that, but everything you listen to has already been tweaked and messed with in the studio or during mastering. The only way anyone can hear the sound exactly as it was recorded is to be in the studio or mastering suite, listening to it at the same volume and through the same equipment that was used when the tracks were finalised prior to CD, vinyl, SACD or Blu-Ray disc production. This is likely to be somewhat inconvenient, even assuming it's possible.
Over the years there have been countless attempts to convince buyers that someone has finally created the 'perfect' system. It's generally considered by audiophiles that most of the mass-market 'perfect' systems are anything but perfect, and in many cases they are probably right. However, there is no amount of tweaking or modification by adding magic components that will make any difference. I won't name names here, but most readers will be able to guess who has been responsible for some atrocious systems that continue to this day, and their owners are generally perfectly happy. They too have been influenced by advertising and consumer reviews that claim they are getting the best of the best, when the systems are best described as overpriced toys.
We have no defence against this kind of onslaught, and vast numbers of people now think that MP3 encoded music sounds 'good'. They have possibly never heard a decent sound system, and would most likely dislike it if they did because it sounds so different from what they expect. As a direct result, it's now extremely difficult to get decent CDs (of artists you actually like listening to). Most have been compressed so heavily that everything is at the same volume (loud), and they sound like crap. No tweak, cable, rock or 'magic' capacitor can fix that - it's ruined forever.
There are countless places in this article where I could have named names and products, but that would only serve to bring them up in searches when people are looking for some sanity. I flatly refuse to link or provide information that can be used in a search or improve page rankings in search engines, where the 'product' or 'service' is fraudulent.
Main Index
Articles Index