Working with the human ear: the proper evaluation of audio equipment incl. amplifiers

[STHLS5]

I was having trouble with Pitch 2 and 3. I couldn’t tell. I was using the same PC as Pitch 1 but just couldn’t distinguish any difference. I tried again in more relaxed environment and with my laptop and I hear the subtle difference.

Could an old soundcard or a speaker (built in Mono speakers) make the difference less obvious? I am not sure how a speaker functions to produce a 440Hz, but my guess is it needs to vibrate 440 times per second. If that is the case, then the speaker needs to vibrate or oscillates 443 times per second to produce the 443Hz. If this is a correct assumption, then how accurate can a mechanical cone that depends on magnetic force to create push and pull function against the surround and air accurate keep count of these numbers? Can we be sure that it actually would vibrate 443 or 440 times and not a plus minus of few beats?

ST

{Moderator's comment: ... because you would have created a time machine.}

[keithwwk]

I not sure what I should I say after listened pitch 3, anyway, I just write my feedback here.

I tried to listen this pitch 3...I do not know how to say it..If only just the A and B separately I do not think I can detect the different. But if listen them side by side and keep repeating for few time, it is clearly that, both sing in different pitch. And by resting my ear for 30s and listen again for the 1st time, I am confused, then I must listen for the 2nd time and repeat for few time to be able to detect the difference. By listening the clip 1st few second, pause it, jump direct to the last few second, the difference is undetectable by me.

[A.S.]

... If only just the A and B separately I do not think I can detect the different. But if listen them side by side and keep repeating for few time, it is clearly that, both sing in different pitch. And by resting my ear for 30s and listen again for the 1st time, I am confused, then I must listen for the 2nd time and repeat for few time to be able to detect the difference. By listening the clip 1st few second, pause it, jump direct to the last few second, the difference is undetectable by me.

Surely what you're saying is exactly what this experiment is designed to illustrate - or do I misunderstand you?

The point I am attempting to convey is that only when the two events (A and B) are presented 'adjacent' to each other in time with a negligible gap, can our extremely limited (but highly predictable) sense of hearing effectively tackle the analytical task of determining if A and B are the same or similar, or quite different. If I understand you, you have found that at around the switch between A and B you are able to readily detect that they are indeed different frequencies, but not easily or at all when they are not compared directly against each other. Is that correct?

If it is, you have proved the very point: our ears are expert at making comparisons but only when we have the two stimuli directly set against each other.

[keithwwk]

..... If I understand you, you have found that at around the switch between A and B you are able to readily detect that they are indeed different frequencies, but not easily or at all when they are not compared directly against each other. Is that correct?

If it is, you have proved the very point: our ears are expert at making comparisons but only when we have the two stimuli directly set against each other.

Yes Alan, you are right. That's exactly how I feel.

[A.S.]

Yes Alan, you are right. That's exactly how I feel.

Well that's good. It proves that your ears and my ears (and indeed everyone else's ears!) work the same way. Very good at instanteous side-by-side comparisons, not very good at all at time-separated comparisons as I've shown throughout this thread.

A vision comes to mind of bartering for fruit in a far-away street market. The seller has no scales but he does have a pre-made weight. It doesn't matter if it's in pounds or kg or even his own made-up weight system. He places this weight in his left hand and then the fruit you select in his right hand and does a little joggle with his hands. Which one is heavier and how much to charge? If you don't agree with him, he'll hand you his reference weight and the fruit and you can confirm for yourself. Humans are universally good at that sort of comparative measurement and can quite accurately resolve differences in temperature, sound, weight, texture, bitterness or sweeteness, frequency, colour etc. providing that we have something 'in the hand' to compare against.

Thanks to everyone who contributed feedback. If nothing else, despite the wholly misguided armchair-expect criticism I have had over the amplifier comparison, I hope that I've illustrated the point that it is a very brave man indeed who would put his total trust in judgements about audio separated by minutes, hours, days, months or years. There is just one final observation to make, tomorrow.

[A.S.]

Thirty days ago, we laid down a challenge (from post 58) to finally put to rest the all-too-frequent and unanswerable question here on HUG 'should I upgrade my amplifier' .....

...The objective of this thread on A-B amplifier comparison is to attempt to confirm (or not) that the millions of words written on the subject of 'amplifier sonics' are backed up by the reality of being able to hear the difference when cosmetics, features, brand name and variation in loudness are removed from the equation. Our position is that Harbeth speakers will work well with any credible amplifier performing to original specification. You need not be concerned that once you buy Harbeths you are on an unwelcome, expensive merry-go-round of electronic upgrades: you are not.

A.S. ... let me repeat yet again: for reasons that we do not understand, it would appear that under instantaneous A-B switching driving Harbeth speakers, that amplifiers of broadly comparable technical performance and matched for gain are not readily distinguishable. That is not to say that 'all amps sound the same' under random, uncontrolled listening.

The challenge offered an attractive gift for a person able to distinguish amplifiers under A-B testing:

In anticipation that I may be surprised and for whatever reason the two amplifiers are readily and repeatedly distinguishable under the instantaneous A-B switchover, we have decided to 'provision' for the gifting of a pair of M40.1s in our financial accounts for the year end 31 March 2012. That means the will be no hard feelings at all from this side in the event we are proven wrong - which we hope we are because if one 'star performing' amplifier is revealed it will have been worth the cost. Other than actually drawing-out the circuit for the relay switch-over comparator I believe that I've done all I can and need to at this stage and await proposals.

I provided a list of all the parts needed to make a relay comparator box (assuming the participant would not have confidence in my switcher) in this post here. This 'challenge' has been widely reported and discussed in specialist audio forums and has apparently had over 50,000 viewing hits. Scanning the threads shows that having started out with extreme hostility and incredulity, that led to personal attacks on my motives and integrity and finally the subject died long before the 31 day closing date.

As this thread has shown, when humans are able to compare two sensory events with only the briefest interruption between them, our brain is capable of making reliable judgements. When a delay of more than one or two seconds is introduced, we find it much more difficult to make objective comparisons. So it is with sound. The essence of my comparison test is to reduce the time delay between switching from A-B to just that of the relays flicking over - as illustrated here, about fifteen thousandths of a second. Under these rapid-fire change-over conditions, differences (or not) should be audible between equipment such as loudspeakers, electronics, cables amplifiers etc..

To my dismay, this simple, cheap and effective A-B comparative test suggestion caused a fire storm of outrage amongst certain audiophiles. Every line of attack was tried to discredit the concept before it was even tried-out: I am deaf; Harbeth speakers must be so poor that they are incapable of distinguising the self evident differences between the amps; the test was biased in Harbeth's favour; the relay switches were imperfect and would corrupt the signal somehow; the participant would be deliberately fatigued by the suggested 100 switch-overs*; the concept of A-B was a denial of human sensory nature and so on. We became aware that despite the dedicated audiophile's eager willingness to regurgitate manufacturer's techno-marketing babble about the technical minutia of this capacitor type or that, this transistor type or that, this cable type or that (all of which are electro-chemical specifics of the laboratory and PhD experts) rolling up the sleeves and actually assembling the switch-over box was admitted to be far beyond their constructional ability. Yet, in truth, a battery, a switch and some relays could be constructed to a satisfactory standard by an average high school physics student. Nobody who is unwilling or incapable of soldering a few bits of wire to a battery should be able to pontificate about electronic design. This disconnect from hands-on (across most/all areas of consumer electronics) sets-up the unpaid consumer as a marketing conduit for the upgrade-inspired manufacturer and has confirmed my worst fears about the impossibility of the average consumer separating technical fact from marketing BS. We've already covered that here. It's going to destroy this industry.

We were extremely relaxed about almost every aspect of the comparison, and had a clean, blank sheet of paper with just two mandated requirements: the participant must drive the speakers through his (or my) relay change over box and the amplifiers must be level matched beforehand. How they were to be matched (what method? what frequency?) was negotiable. I assumed that the participant would wish to observe that measurement process and that I'd patiently walk him through it to allay any misgivings he may have about test equipment. In fact, everything except the relay arrangement was negotiable - the music, the location, the amplifiers to be used as A and B, which Harbeth speakers, cables, stands, music source, how many observers, scoring method .... absolutely every detail could be worked-through and agreed beforehand. I would have helped the candidate solder together his switcher if he so requested and I even considered a legal agreement (obviously at my expense) so that the outcome was binding on both parties - we handing a pair of free M40.1s to the lucky 'winner' or he graciously accepting that he could not tell the amps apart. I assumed that I would have laid-on and paid for a nice lunch and/or evening meal for participants and that the whole process would be undertaken in my normal relaxed, friendly way complete with a tour of this historic area. I did not expect the participant to pay a penny piece, except building the relay box (about GBP20-30) and I budgeted about GBP 250 for food/drink plus the cost of the M40.1s if they were handed over.

What did we anticipate would happen, and what actually did happened?

We anticipated that we would have to arrange a number of sessions of perhaps 2-3 hours spread over a week or so as participants brought forward their amplifiers claiming 'night and day' differences in sound. We'd measure them technically and then carefully adjust the replay levels so that each one was playing at the same loudness. Some internet chit-chat said that it would be easy to walk away with a free pair of M40.1s (any veneer, incl. rosewood) and then next day they'd be listed on Ebay, but as we didn't preclude that we could not have objected. We were told that there was absolute certainty amongst the die-hard audiophiles that they could, under sighted conditions, easily tell this amp from that. As the arguments raged, some were not so sure what to expect if the amps were directly compared under A-B conditions. Slowly the voices of objective, technically trained commentators came to the fore and reminded readers that all known public A-B comparisons had reached the same conclusions during the past thirty years: sonic differences diminish or disappear when the participant cannot see the electronic equipment just as I suggested at the outset. And so, the raging torrent of fury and disbelief slowly dried-up and then ceased.

So, how many presented themselves as participants within the rules of the game, which mandated the use of the relay box? Not one single, solitary sole. Not one. I am so surprised. And furthermore, not one phone call, email, fax or letter from anyone connected with Harbeth sales or distribution. Not a peep. Business continued as usual.

I have permission to present, unedited, the one email we did receive from the public but as this gentleman says, he would not permit the use of the relay switcher (and therefore he cannot be considered a prospective candidate) but he did have the courage to step forward. What he says summarises the audiophiles position:

..."I don't care what method other people use, my method that works for me is as described below; A-B-A-B with the whole track played each time.

In my view 2-30 second swaps are a complete waste of time and only show up tonal differences, not difference in timbre, image or detail. I don't see the reason for limiting it to 30 seconds, as this proves nothing. For info, I tried my method out last night at a friends house using 4 amps and 100% was able to tell which one was which, as usual (3 valve 1 transistor). Now I am sure they don't all measure the same, but that is another point, all the amps were current and good quality (£2K to £6K cost).

I listen to rock and techno fairly loudly.

Lets be clear, I am not making or buying anything and I am not getting involved in me having any costs (other than travel) in this whole thing. I can supply amplifiers and cabling and if needed a decent source (either vinyl or CD). I don't believe in switch over boxes as they add extra masking, simple banana plugs on the speaker cables should be fine."

Where do we go from here?

As we said many posts ago, we respect and support all those who design, make and sell electronics. But we are not at all comfortable that this loudspeaker forum is used to provide unsubstantiated and unsubstantiatable opinions for electronic 'upgrades' none of which are necessary for getting the best from Harbeth's easy-to-drive speakers. Other speaker brands may well be amplifier sensitive. In our opinions that is sheer bad speaker design. Harbeth speakers are an easy, amp-friendly load. Full stop. End of story.

So, we state yet again our view about the purchase of electronics ....

'... the selection of an amplifier should, in our view, be based primarily on pragmatic issues of its durability and longevity, styling, features and facilities, brand reputation, technical specification (power, hiss etc.) and after-care service back-up until such times as the claimed night-and-day 'sonic differences' can be validated as real, not illusory under non-sighted, controlled listening tests. If 'sonic differences' can be positively identified under controlled conditions then, obviously, they should be added to the list somewhere'.

Please do not breed discontent amongst Harbeth users by claiming that changing electronics will give them a 'night and day' sonic experience. It won't, but it may give them prettier, more reliable electronics with better features all of which may indeed be worth paying for. Visit your dealer: he'll tell you the inside story behind an electronic product/brand, the one you don't read about in forums. Base your purchase decision on hard face to face facts from your dealer. Go listen for yourself but be sure that you are listening at exactly the same level, matched to 0.1dB to accurately compare sonics or just buy what you like the look of and are assured has a proper service backup. But please don't buy based on unsupportable claims about wonderful, earth shattering sound differences.

Conclusion: I am shaken after experiencing first-hand the knowledge vacuum between how audio/sound/business/human nature actually works and how certain dedicated audiophiles think it works and how loud they shout to defend their beliefs. It now wouldn't surprise me if there are internet chat forums discussing the ins and outs of neural surgery amongst people who have never held a scalpel. This is madness. To make technical purchasing decisions, to discuss intricate technical parameters and to influence others, you have to have to have baseline technical knowledge by whatever means you acquire it, formally or informally. Plus a good deal of common sense. It's a denial of reality to believe that you can comment meaningfully on neuroscience because you once visited a butchers shop. That's the state of affairs in high end audio though.

I remain dedicated to doing whatever I can to illuminate how audio really works for the benefit of great music at home (example today here) whether the facts are palatable or not. You can always look away.



*Very few commentators bothered to check what I actually wrote here. I did not define exactly how long the participant could listen to A before switch to B (and vice versa). I listen in short burst of perhaps 10 or 20 seconds then switch; I do not listen for more than 30 minutes or so at a session to avoid fatigue. 100 switch-overs could take as little or as long as the participant wanted.

[denjo]

Hi Alan

Indeed, the offer of your challenge has lapsed with not a single soul taking up the gauntlet! Like many a curious bystander, I was hoping someone would actually accept your challenge so that there would have been an experiential anecdote that would have doubted or validated your strong belief that under controlled circumstances the selection of amplifiers of broadly comparable technical performance and matched for gain are not readily distinguishable, one from the other.

Like many, such a proposition seemed almost insanely preposterous to be believable and evoked reactions ranging from mild indifference to heretical swearing! Yet not a soul stepped forward to prove you wrong. While it would be wrong to deduce or infer from this inaction that your proposition is therefore right, I have deep respect for your courage in holding fast to your beliefs! While I have not always agreed with everything that you have said in my many years as a forum member, I am increasingly inclining to the view from my own experience (almost one score years) that you are largely correct in evangelising such a view, contrarian as it might at first blush sound.

It is interesting to note from your many posts that you have candidly admitted that you don't fully understand the rationes behind this view which might empirically explain why this is so, perhaps the answer might lie in the complex field of pyscho-acoustics and the cerebral hemispheres of that grey matter which lies between our ears. Perhaps someday, somewhere and somehow we would have the science (and art) to finally put to rest this phenomenon, proving beyond reasonable doubt that the electronics, so long as of competent design and quality, does not make such a big difference as audio manufacturers and audiophiles make it out to be. When that day of pragmatism arrives (and I hope my heart will still be beating and my ears, capable of hearing), I think we will all be closer to the main objective of really enjoying music instead of chasing our tails for that elusive amplifier that powers our beloved Harbeths!

Listen to this clip which was recorded using two microphones positioned approximately where our ears might be to experience the wonderful miracle of the human mind in how we hear, anlayse and perceive sound in order to create the sense of ambience and depth and consequently, the realism of a musical event. Although earphones are recommended, you can nevertheless experience the holographic effect with your Harbeths!

http://www.youtube.com/watch?v=IUDTlvagjJA

[Kumar Kane]

Indeed, the offer of your challenge has lapsed with not a single soul taking up the gauntlet!

It is interesting to note from your many posts that you have candidly admitted that you don't fully understand the rationes behind this view which might empirically explain why this is so, perhaps the answer might lie in the complex field of pyscho-acoustics and the cerebral hemispheres of that grey matter which lies between our ears. Perhaps someday, somewhere and somehow we would have the science (and art) to finally put to rest this phenomenon, proving beyond reasonable doubt that the electronics, so long as of competent design and quality, does not make such a big difference as audio manufacturers and audiophiles make it out to be.

It is indeed surprising - even more so given that any challenger could get the prize, but with no price to pay for losing, other than travel costs, I guess.

On the cerebral thing - I am convinced that this is what is "messing up" things. My system sounds best to me, when the lights are down low, and after a couple of glasses of wine or a good single malt. Nothing has changed in "real" terms, but what happens inside the brain is more real to me than what is "real" outside in the room.

And when I change electronics to a highly recommended (touted?) and more expensive unit, I "hear" better sound.

I now have realized this, and prefer the lights down low/glass of wine route.

[A.S.]

It is indeed surprising - even more so given that any challenger could get the prize, but with no price to pay for losing, other than travel costs, I guess...

It is surely the most disappointing (failed) 'experiment' of my career. Yes, nobody stepped forward except the one I mentioned and as he would not play by the rules (didn't accept the switch-over box), he debarred himself. I heard recently that a UK amplifier manufacturer was much aggrieved by the proposed comparison. I can only imagine that he fears (or anticipates, or knows) that the outcome would not be good for sales if, as I postulate, when amps are compared under controlled conditions the big reported sonic differences "diminish or disappear", to quote myself.

That is a barmy, short sighted way of looking at the marketing of quality audio products. As I said, what I would do as an amplifier marketeer would be to promote the real, tangible benefits of the brand: quality components, careful consideration of longevity in design and assembly, company reputation for after-care, a buy-and-forget (almost) forever approach .... these are all real benefits and worth paying good money for. Fabled 'sound quality' is a completely nebulous, intellectual dead end. As dead an end as the sort of utter cobblers that is used to sell anti-wrinkle cream to women. No marketing campaign - no viable business - should pivot its entire existence upon such a wiffly-waffly, easily disproved (under controlled tests) promotional strategy. I'm so sure of this that if I were tasked with bringing a quality audio amplifier to market, I'd set myself the challenge to never introduce 'sound quality' into its promotion and I'm confident that it would be a sales success solely by concentrating on the real-world user's benefits. Promoting superior sound quality (or even strong sonic personality) of electronics is nothing more than the marketeer scraping the very bottom of the barrel for something new to say to a consumer daft enough to be seduced by sweet words that he wants to hear.

Sure as night follows day, when this subjectivist road if fully walked, (because there is nothing more to say, the would-be consumer has lost interest) there will eventually be a complete reversal in the media and it will be fashionable again to say ... 'You know what? We were all kidding ourselves about that subjectivist stuff. Let's get down to basics again ....'

Bank on it.

[Kumar Kane]

to a consumer daft enough to be seduced by sweet words that he wants to hear.

Sure as night follows day, when this subjectivist road if fully walked, (because there is nothing more to say, the would-be consumer has lost interest) there will eventually be a complete reversal in the media and it will be fashionable again to say ... 'You know what? We were all kidding ourselves about that subjectivist stuff. Let's get down to basics again ....'

Bank on it.

I am not so sure I would bank on it, Alan. Human nature being what it is, there are so many people out there that are waiting for someone to tell them what they want to hear, so they can indulge in retail therapy. Marketers build careers doing just that in the consumer products and services industries. Global GDP would crash if people stopped listening to their pitches. Maybe it would be a one time correction, and be a lot better for the planet in the long run, but I don't see it happening.

[Euler]

It is surely the most disappointing (failed) 'experiment' of my career.

If you think of the challenge and the response as part of the whole experiment, then the (non)response of the audiophile community to your challenge, Alan, makes your experiment a success. Suppose some audiophile (in England, to minimize travel costs) really and truly, in his heart of hearts, believed with great certainty that he could hear differences between two amplifiers (under controlled conditions). Would he not have taken up your challenge? Why wouldn't he? So (argument by contraposition), from the total nonresponse to your challenge (to your experiment) we may infer that no audiophile (at least in England) really and truly, way deep down, believes with any certainty the garbage about amps. In my opinion, that's one very, very successful experiment.

Bruce

[STHLS5]

.... we may infer that no audiophile (at least in England) really and truly, way deep down, believes with any certainty the garbage about amps. .

On many occasions, after carefully level matching between different amplifiers I am unable to tell any difference. Having said that, I am inclined to prefer the sound of some amplifiers played under the normal listening level.

If anyone spewing garbage about amplifiers sounding difference, then Norman H. Crowshurst should be the first one to be blamed for making such a claim.

Two amplifiers of conventional design were taken and modification made to bring their design into the line of mathematical theory….AB checks were conducted between the amplifiers using the original circuits and the revised feedback circuit. A difference was quite noticeable …particularly..when wind instruments… or string instruments are played. These experiments certainly seemed to have uncovered at least some of the major differences that can exist between amplifiers with equally good specifications – differences that do not show up, at any rate, in the standard method of specification. These are, in fact, defects that are not in the books!
- Norman H. Crowhurst.

ST

[Kumar Kane]

In my opinion, that's one very, very successful experiment.

I agree.

What will happen though is that it will be buried in new threads/posts with the passage of time, and the same tedious discussions (about the love for this amp or that) will continue, even on this forum.

In other fora of course, the debate will continue to rage, the population at large has no clue about this initiative from Alan.

Oh well, the internet is a free medium, and at least, no trees are cut down to make paper:-)

A few years ago I decided to simplify my system, and settle down with a now ten year old Quad 909. I love it and I hope it will last at least another ten years, if not more. I got my C7s just last year because I was moving into a larger house that needed their heft, and the discussions on the forum, and the C7 capabilities underline how sensible the decision was.

At 140wpc the 909 is more power than required, but that isn't reason enough to change amplifiers. And the ten year old Quad 99 preamp that feeds the 909 is also brilliant for its feature set - tone controls, and a phone stage that can take moving coil as well as moving magnet inputs. Even today, it isn't easy to find one with the combination of features it has.

[A.S.]

I agree. What will happen though is that it will be buried in new threads/posts with the passage of time, and the same tedious discussions (about the love for this amp or that) will continue, even on this forum....

Whilst my (failed?) experiment has been a disappointment* what has been even more noteworthy is, as you suggest, the continuous stream of those eulogising about this amp over that amp - even here on the HUG. We simply cannot understand the force that drives people to utterly ignore the thrust of our considerable input on this debate (that amps must be compared under controlled conditions) and then tell the world, here, that under uncontrolled conditions that they've just had a night and day experience substituting amplifiers. Of course they will hear a difference! That's a given! But that difference is not likely to be the amp itself, it's the result of comparing two uncontrolled events. How many millions of words and visual/musical examples do I have to make to hammer home once and for all the concept that the human brain is very, very, very easily duped?

I'm beginning to ask myself why I actually care about this. It certainly isn't because I want to dissuade folks from visiting their hi-fi dealer and spending wads of cash - quite the opposite. It's because this nonsense is endemic of a breakdown in society of rational thought, the ability and willingness to use one's own common sense. This failure of independent thinking, and the willingness to be led by the nose by persuasive opinionated leaders has cost, and continues to cost, human lives.

* Disappointment because what was in it for me was the potential to discover an outstanding amplifier which truly released the sonic potential of the Harbeth RADIAL cone. So I've been deprived of that discovery - or not maybe.

[A.S.]

Clearly there is no comprehension at all of what goes on inside an amplifier's case. So let's compare the variables that the speaker designer (or car designer, microphone designer: all mechanical systems) have to play with and balance into some sort of uneasy compromise, and bring that product to market. As no two design briefs are the same, no two cars are the same, nor speakers nor microphones. Not only will there be individual piece-by-piece variations due to small but uncontrollable variations in parts used and their assembly but the overall balance of mechanical parameters will vary according to the designer's target. So you can have a car that glides over the road (a Rolls Royce) or a car that grips the corners like glue ans where you can feel every bump. But the designer cannot deliver a car that simultaneously glides and grips like glue. The design of any mechanical system is about making choices and prioritising them as the market requires.

How about electronic systems? Tuners, amplifiers, CD players, DACs, sound cards and so on. Let's compare. How much flexibility does their designer have?

A quick first-thing-that-comes-to-mind list of variables that (electro) mechanical (speakers, microphones etc.) has to play with and optimise for the market need:

1. Electro-acoustic efficiency from sound in/out to electricity
2. Mass of the moving parts
3. Diameter of the moving parts
4. Edge termination of the moving parts to allow them to move yet return to a neutral rest position aka spring compliance
5. Diaphragm thickness
6. Diaphragm shape
7. Diaphragm material
8. Acoustic properties of diaphragm materials versus frequency
9. Temperature stability of all moving parts
10. Optimum adhesives to bond and stay bonded over time
11. Acoustic properties related to adhesive rigidity
12. Amount of and exact location of glues and how to terminate a glue thread without blobbing (adding weight)
13. Fundamental resonance of the mass of the diaphragms bouncing on the air stiffness in the cabinet
14. Port tuning for amplitude, frequency and Q
15. Magnetic linearity with increasing power
16. Distortion minimisation
17. Effect of ambient temperature esp. on bass performance
18. Flatness of frequency response from individual drivers
19. Out-of-band unwanted peaks and troughs - avoidance or minimisation
20. Crossover, selection of optimum crossover frequency
21. Shape of crossover slopes below, at and above crossover frequency
22. Physical position of crossover components on the PCB to minimise magnetic interaction
23. Power durability in use with the customer over years
24. Cabinet damping on panel surfaces
25. Cabinet damping in the air cavity in-box
26. Method of attachment of drive units to cabinet and considerations of opposing forces and resonance avoidance

and so on. That's just for starters. There are probably over one hundred understood physical parameters which have to be simultaneously optimised by the designer, to fulfil his market needs even in the most humble, inexpensive speaker.

Now let's compare the levers of influence that the electronics designer has access to, and can apply to his design. It's a very short list indeed. The reason is that, unlike the speaker transducer which is an open-ended system (meaning, electricity is transformed to sound to one degree of efficiency or another) the amplifier is a closed-loop system. There is no conversion from one form of power to another (electrical energy to sound energy or sound energy to electrical energy): the amplifier works only in the electrical domain. The function of the amplifier is to a-m-p-l-i-f-y whatever voltage appears at the input socket into a bigger voltage at the output socket. That's it. That's all it has to do. Nothing more, and hopefully nothing much less. It is dumb. It does not know whether music or test tones are at the input terminals. It does not care whether it is Pink Floyd or pink noise it has to amplify. It is a brainless, obedient, robotic servant.

So what variables does the amp designer have to play around with? Well, he is chronically handicapped by two absolutely paramount, essential, life or death requirement:

1. Circuit stability over hours, days, years i.e if the amp fails it will destroy the speakers and/or catch fire
2. That the output must be directly correlated with the input i.e. if Mozart is appearing at the input terminals, Led Zeppelin is not appearing at the output terminals

So, with those overarching mandates the design begins. Presumably the prudent designer would start with a sketch of a simple block diagram that would increase the signal in two or three steps from the very small input voltage to a suitable one to drive a speaker - the so called 'gain structure' of the amp. Let's suppose that we needed 50v at the speaker terminals from a phono input of 50mV ... we'd need to increase the gain by (50/0.05) = 1000 times, also known as 60dB. We'd probably have a low-hiss input stage that increased the gain by 50 times and then a heavy duty output stage with a gain of 20 times, or whatever the designer wished for. The primary variable that the amp designer has available to him is the gain of those two stages (50x and 20x) can be set by the relationship between two resistors in the feedback loop, at a cost of about $0.02. This is the amplifier designers orbit of influence because whilst he may have dreams of exotic circuitry (because, as a creative person he's curious about new design approaches) he has to work within the two paramount requirements of circuit stability and input-output correlation. And we all know that the more complex a design becomes the more parts there are waiting to fail.

These paramount requirements mean that every component added to the design must be justified. If one capacitor will do the job, will two degrade long term reliability? Can adding components beyond those absolutely needed to perform the gain-up function be expected to improve the 'sound' of the amplifier? Fantasically unlikely because if they did, the amp had not been initially designed within the constraint of 2, the mandate of input-output correlation. Remember that unlike the speaker (or mic) where there is a transformation from sound into electricity making measurement tricky and hence introducing a big subjective element, the input and output of the amp can be directly compared since they are both electrical. They are both in the same domain. They can be both measured with the exact same equipment and directly compared, volts-in compared with volts-out.

Let us repeat the speaker designers design variables list again and remove from it those elements that the amp designer doesn't have and see what's left ....

1. Temperature stability of all (moving) parts
2. Distortion minimisation
3. Effect of ambient temperature
4. Physical position of (crossover) components on the PCB to minimise magnetic interaction
5. Power durability in use with the customer over years

We must add to that concerns about ventilation and electrical safety.

But the essential point is this: because electronic equipment such as amplifiers have no moving parts, their designer is not concerned with balancing the physical properties of moving parts such as mass, springiness and damping. And the 'personality' and sonics of every electro-mechanical system is directly and inextricably bound up with weights and compliance of the moving parts. Any system without moving parts is most likely to be free of the coloration that those moving components inevitably introduce.

As amplifiers have no moving parts and their gain is set by tiny resistors, is it really credible that there can be night and day differences in sound quality? No. Subtle ones surely, but they must be exceedingly small indeed or the test equipment could reveal them in the output when not present at the input.

[denjo]

How about electronic systems? Tuners, amplifiers, CD players, DACs, sound cards and so on. Let's compare. How much flexibility does their designer have?

Alan,

Are you saying that the proposition that all amplifiers sound broadly indistinguishable applies to tuners and DACs as well (electro with basically no moving parts) [how about CDPs since they have moving parts?]? By your line of reasoning, it would seem that most items with electro parts should broadly sound the same if subjected to a similar A-B comparo under controlled circumstances? From what you seem to be saying, the single piece of the audio chain which has the greatest impact on the sound are the speakers, am I right?

Best regards
Dennis

[Kumar Kane]

That's just for starters. There are probably over one hundred understood physical parameters which have to be simultaneously optimised by the designer, to fulfil his market needs even in the most humble, inexpensive speaker.

Now let's compare the levers of influence that the electronics designer has access to, and can apply to his design. It's a very short list indeed. [/U]

Good comparison. Consider though the plight of the amp maker/designer, given the demand for differentiation in the market place to sell against the competition, as well as to persuade existing customers to buy his latest creation.

Feature set? How many features can he offer anyway - an amp isn't a Swiss army knife at the end of the day. Reliability and after sales support too is a given these days, and not enough of a differentiator, although it isn't everyone who offers this of course.

So what is left is the eye candy of the case work, glowing dials, analog meters, buttons and knobs, backed up by the skill of industrial designers, where he can play around to create apparent differentiation, along with what seems to be obscenely high power outputs in some cases - 1000wpc?. And coat that with all the language in the product brochure that is amplified ( if I may use the word!) by the reviewers and magazines. And the merry go round keeps moving in what must be a diminishing market place. To add to this, by the same justification, the revival of tube amps. And all of this at spiraling prices, in the absence of volume of sales, to remain profitable.

Even in the case of speakers where from your list it seems there is enough to optimize to give headaches to a designer, outrageous and bloated design seems to be the driver these days, if I see the photos from the Munich show. Some of the speakers on display there were horrifying. Things with 50 drivers or some such fantastic number?

[Kumar Kane]

the single piece of the audio chain which has the greatest impact on the sound are the speakers, am I right?

I would agree. And how well they are positioned along with the room acoustics. At the other end, the quality of the performance, and how well it has been recorded/mastered.

Solid state and digital technology has matured such that the old wisdom of spending most of the money on the source, or a thirds each on the source, amplification and speakers isn't valid any more, I suggest. As long as the source isn't a turntable of course. I would first get the best speakers I could afford, the rest of a solid state system could come in for 25% of their price. Give or take a few percentage points.

Upstream, I would also suggest that the greatest advances in the decade gone by have been in the area of convenience by digital tech, doing away with the need for CDs and CD players. Along with wifi/internet radio and internet music services, this will probably transform the way music is heard at home for the large majority of people.

[A.S.]

Alan, Are you saying that the proposition that all amplifiers sound broadly indistinguishable applies to tuners and DACs as well (electro with basically no moving parts) [how about CDPs since they have moving parts?]? By your line of reasoning, it would seem that most items with electro parts should broadly sound the same if subjected to a similar A-B comparo under controlled circumstances? From what you seem to be saying, the single piece of the audio chain which has the greatest impact on the sound are the speakers, am I right?

Absolutely so. Speakers are by far the weakest link in the chain. When they are replaced with electrodes directly into the brain, fidelity will take a big leap forward by eliminating the speaker/room problem.

Just to recap: what gives equipment a 'personality' — be it a car and its drive performance on the road or a speaker or a turntable or a microphone — is the balancing of mechanical forces. The car's drive is directly the consequence of a certain car mass bouncing on suspension of a certain stiffness which comes to rest (aka damping) a certain time after the pot hole in the road (a perturbation of a mechanical system). The speaker's bass performance is the result of a certain mass of cone bouncing on a certain stiffness of air in the box and coming to rest a certain time after the note ceases (aka damping), another perturbation of a mechanical system.

Is there a method to perfect the smoothness of the ride simultaneously with and high speed cornering ability? No - can't be done. The designer must prioritise one over the other. The Rolls Royce designer would be aware that his users are typically pootling around at 50mph and the smoothness of the ride is all critical to them. The fact that at 150mph, a speed which the car is certainly capable of travelling at, it maybe incapable of high speed cornering without the risk of turning over is a consequence of the primary design decision to optimise the very high mass with a certain suspension stiffness. Thus, a speaker designer has that long list of mechanical parameters to juggle as he thinks best - or is told to by his marketing department who understand the customer's expectations.

But the poor electronic designer just doesn't have those tools. There is almost nothing to tweak or tune. Nothing moves, rattles or shakes. There are no forces of any sort at work, other than the flow of electrons. If the output of the amp is merely a bigger version of the input, there are no "spanner" in his design toolbox to bring to bear on the design. He is really cornered by just working with electrons, which being massless stop and start instantaneously so unlike fluids or cars or speakers cannot exhibit the consequences or personality of moving systems which do have mass.

Of course, the amp designer could (and probably often does) shape the measurable frequency response to provide a little more gain in this frequency band or that. But if he takes that approach, he is not adhering to the primary requirement of a good amp that the output of the amp should be nothing more or less that a bigger version of what appeared at the input terminal. It is easy to overtly or covertly tailor the frequency response of an electronic system to give it a 'tone control like' enhanced sound just with a handful of $0.05 capacitors and resistors, if that is what the designer wants. But again, the electronic test equipment can easily measure the deviation from flat at the amp's output. As so-called audiophile amplifiers are deprived of user tone controls, the implication must surely be that the designer's target is to provide a measurably flat output that is merely a bigger version of the input with exactly the same overall response characteristic, 20Hz to 20kHz.

And this is where the illogicality steps in. If the audiophile amp designer (which I define as one who attests to the 'straight wire with gain' concept that the amp output is merely a bigger version of the amp input) has taken pains to make sure that the amps output is linear and flat he has intentionally turned his back on all the little circuitry tricks that a non-audiophile designer could use to shape the frequency response. How than can that audiophile amp then be proclaimed as 'more musical', 'more revealing', 'more open and transparent'? It is simply impossible to add personality at circuit level whilst maintaining the overarching objective of what comes out is merely a bigger version of what goes in; the ideal amp.

In my opinion we should entirely cease to talk about nebulous amplifier sonics and literally take the lid off the amp under consideration and have look at how well it is made. How the hot does it run? How easy is to disassemble? Is it well laid out inside? Is the transformer big and meaty or puny? Does it use lots of ICs which will become obsolete or discrete components which will be around forever and readily available from general trade supplies like RS Components? Is the design modular? Is it overly complex? How neat is the wiring loom? Is there a background hum or buzz up close to the speakers? Is there digital breakthrough onto the audio tracks? Are there enough inputs/outputs? How well does the remote work? Does it have tone controls (a very good idea)? What is the quality of the connectors like? In my experience, such scrutiny would rule-out most far-east made AV amplifiers on several counts. Perhaps most important characteristic of all is - do you like the look of it?

We're back to a willingness to pay more for a better made amp with better facilities, using simpler readily available parts with a proper service backup that looks good. What's wrong with that? To hell with claimed 'musicality'.

As I said a few posts ago, there is so much more to say about the design and use of speakers or amplifiers than concentration on unprovable, totally subjective, potentially misleading, valueless marketing waffle about 'musicality' and the similar non-specifics. EDITED FOR CLARITY: It is only for reasons of brevity that we/I would resort to audio jargon, so why doesn't everyone make that extra effort to properly explain what they hear without audiophile jargon?

This forum is a testament to just how much of real scientific weight there is to explore in audio and psychoacoustics without euphemisms and jargon and where appropriate, with the use of actual playable audio examples — and we have hardly scratched the surface of the subject. What I would like to hear by means of audio clips is how the electronics designer selects one circuit/construction topography over another, since if he can hear the difference, should the audio signal passing through those alternative circuits be recorded, we too could hear the difference.

[A.S.]

I have been castigated for the (rare) but unavoidable use of occasional audiophile marketing shorthand simply to convey a point, such as 'microtonal detail'* etc.. As you will know from the vast amount I've contributed here, I would rather deconstruct and explain in detail some point to avoid using a less specific generally understood marketing euphemism. But a detailed explanation is not always appropriate for a wider audience, and it takes (printed) space, an attention span that may not always be available and is information overload and not expected by the consumer in brochures and primary contact.

If you look through my comments here which represent the core of our thinking, I don't think you'll find much evidence of the use of euphemisms in my writing. I go out of my way to avoid them if I believe that, with effort on my part, we can get to the truth.

I have a business to run and have very limited time available to contribute here so you'll have to forgive the occasional personal and corporate slip into shorthand which, yes, I authorise through absolute necessity.

Back to the subject of loudspeakers for me.

*At least with this jargon a little thought reveals an engineering core. Let's break down 'microtonal detail' into its component parts. See how many words we have to use as a substitute for those two ...

1) 'Micro....' - we must be talking of something small
2) '....tonal...' - we must be talking of something relating to sound
3) '... detail' - we must be talking about some nuance of the small (1) sound (2)
4) 'Microtonal detail' - taken together: some small character of sound that is perhaps not immediately obvious on casual listening and may go unnoticed except to the trained ear?

That's converted 'microtonal detail' into common language. But we can take that a stage further and convert 'microtonal detail' in to the underlying engineering language something like ....

5) Since the microtonality is by definition related to small events in the sound, those events must have low amplitude and hence low energy related to the main sounds. We would expect that those micro sounds be significantly or totally masked by the louder dominant sounds. If 'microtonality' is audible to the careful listener, assuming that their amplitude remains low relative to the primary maskers, then can we assume that the masking threshold must at least temporarily shift such that the low level micro-sound is exposed? This threshold shift may be in frequency relation to the frequencies of the small sounds or in amplitude threshold or both. There may also be a temporal shift in the masking threshold. We can postulate from prior listening experience and hence estimation of the frequency bands in which the microtonal detail becomes audible (when the masking threshold exposes it) and the sort of music that is best at revealing the effect, and the type of cone materials that reveal/do not reveal those micro sounds, that this microtonality must be related to mechanical or acoustic damping in the speaker unit itself and/or the electrical network and/or the amplifier that drives the speaker. If this is so, can we conclude that there if the damping is excessive - example: the cone is made from sponge - that all the low level microtonal sounds (with their very low energy) will be fully absorbed into the structure of the cone (sponge) and never be converted into sound ... hence the fundamental limiting factor of the speaker's ability to reproduce micro-sounds is limited by the mechanical properties of the sound generating cone?

And there are several layers below that general engineering introduction. A chemical engineer would look to the molecules in the cone and consider the energy bonds between those molecules with all the specialist terms associated with the branch of science. The mechanical engineer would consider the forces in the cone and how those compressive shock waves through the cone would eventually radiate as sound (or a proportion would, anyway) using his specialist language. And still we have barely explored the issue of 'microtonal detail'. Can I be permitted to occasionally use a snapshot term (which I can and willingly will illuminate if asked) or do I have to tap it out every time in full, boring you (and me) to tears?

Now try and turn 'musicality' into some sort of engineering language. I can't.