Human hearing, hearing loss and audio memory

[yeecn]

This thread relates to how we hear. If we understand the hearing mechanism and process and how our audio memory really works, we are better able to evaluate audio equipment.

THE ANATOMY OF HEARING

The first part of the video contains a very nicely done animation of the anatomy and mechanisms of human hearing. The second part is about the anatomy and symptoms of hearing loss. The third part is about application in hearing aids.

Highly recommended.

[yeecn]

http://en.wikipedia.org/wiki/Echoic_memory

Echoic memory, the auditory version of sensory memory, refers to the phenomenon in which there is a brief mental echo that continues to sound after an auditory stimulus has been heard. In comparison to sensory and iconic memory, echoic memory is thought to last a little longer, for upwards of about three or four seconds. Due to its short span, echoic memory is a type of sensory memory as the echoic memories are temporal and last only for a brief period of time.

The echoic memory acts as a buffer area for raw sound. Form there the sound is interpreted into speech, tunes, pitch and other information like approaching footsteps.

The echoic memory has a retention span of 2-5 seconds. The decay is exponential, and has been measured through brain scanning techniques.

I believe that the echoic memory has to be short, just long enough for us to extract whatever useful information - "Take the garbage out", the French horn playing a beautiful melody, the baby is crying in distress, footsteps approaching in the dark - and fade away as fast as possible so that it will not interfere with the next stimulus.

It is the same with visual memory. Most people only retain the relevant happening within our field of vision, and ignore all the rest. There are people who demonstrate photographic memory. I have seen a documentary where a man can take a helicopter flight over London and draw a bird eye view of London with photographic details. But that person is handicapped severely in other aspects of life. He could not tie his shoelace for example. These people are commonly referred to as the idiot savants.

What does that means in terms of doing sound comparison? I have seem many references to rapid switching as being the golden standard for sound comparison. I believe that this is in reference to the rapid decay nature of echoic memory.

[A.S.]

...Echoic memory, the auditory version of sensory memory, refers to the phenomenon in which there is a brief mental echo that continues to sound after an auditory stimulus has been heard...

I can give you an example of the audio memory at work.

Sometimes my wife is adamant that she said something which I an convinced she didn't say. She is so convincing that at first, I am obliged to believe her and assume that I have misheard her. But I am secretly certain that what she meant to say and what she actually said did not agree. She truly believes that she said what she intended; I equally believe that she became confused. The more she replays here own words in her brain the more certain she is of what she said. True or not.

In these situations I try to replay in my mind the conversation to try and justify my position and get to the objective truth. But try as I might, my recollection in my brain's audio memory of what she said fades within a few seconds and if I'm not careful, I can actually invent words to justify my belief. Relying on audio memory is a very dangerous game.

As the actual memory fades it is replaced with an impression of the event (conversation). In this example, what she actually said disappears and is replaced by a feeling about the conversation. If the result was an argument, then that is the lasting impression and the truth rapidly becomes masked by the feeling of injustice or whatever. That feeling becomes a long-term substitute for the actuality of the conversation and it's that process which builds long-term prejudices and preferences in our mind. Hence, our school teacher praises us, but what remains is not her exact words but the warm feeling that remains with us - possibly for life. We can subsequently invent a conversation to support that feeling.

[EricW]

http://en.wikipedia.org/wiki/Echoic_memory

Echoic memory, the auditory version of sensory memory, refers to the phenomenon in which there is a brief mental echo that continues to sound after an auditory stimulus has been heard. In comparison to sensory and iconic memory, echoic memory is thought to last a little longer, for upwards of about three or four seconds. Due to its short span, echoic memory is a type of sensory memory as the echoic memories are temporal and last only for a brief period of time.

The echoic memory acts as a buffer area for raw sound. Form there the sound was interpreted into speech, tunes, pitch and other information like approaching footsteps. The echoic memory has a retention span of 2-5 seconds. The decay is exponential, and has been measured through brain scanning techniques. What does that means in terms of doing sound comparison? I have seem many references to rapid switching as being the golden standard for sound comparison. I believe that this is in reference to the rapid decay nature of echoic memory.

Okay, I am going to play devil's advocate here. What you say is only true if the only valid comparison that can be made is between something still in the "buffer" (echoic memory) and the sound one is listening to at the moment. But that's so, doesn't it imply that the only valid period of comparison is the first two or three seconds after the switchover? After that, the original has faded from echoic memory and there's nothing left to compare.

Surely that can't be right. If echoic memory is analogous to a buffer, what happens afterwards? Doesn't sound get imprinted on longer-term memory in some way? Surely it must, or we wouldn't be able to discriminate voices, tones, sounds, etc. So isn't there a question as to how reliable that imprinting and recall process is, and doesn't that potentially vary somewhat by individual. And doesn't education and training perhaps play a role? I'm thinking for example of the auditory memory of a trained and skilled musician (for example) versus a non-musician - isn't the musician going to file more and different types of information into longterm memory, and isn't at least some of that information going to be available for recall?

It may be that the only reliable form of comparison at the end of the day is instantaneous switchover, but if that's true, I don't think echoic memory provides a complete explanation.

[EricW]

I can give you an example of the audio memory at work.

Sometimes my wife is adamant that she said something which I an convinced she didn't say. She is so convincing that at first, I am obliged to believe her and assume that I have misheard her. But I am secretly certain that what she meant to say and what she actually said did not agree. She truly believes that she said what she intended; I equally believe that she became confused. The more she replays here own words in her brain the more certain she is of what she said. True or not.
...

Devil's advocate again. Surely this is different as well from the comparison of two audio components. What you are saying is really two things: (1) memory can be unreliable, and (2) memory can be strongly influenced by emotion (i.e. what you want to believe is true, as opposed to what is true). But what you are describing is a situation that is unrepeatable and unverifiable (assuming no objective external recording device or listener). You can't go back and have the conversation with your wife over again. You can't do it over and over until you're sure who's right. But you can compare Amp A to Amp B as long and as often as you like. Surely this makes a difference.

That memory can be unreliable is undeniably true, but it's not universally true. Some have poor memories, some have great memories, some are in between. Memory can also be trained, and with training, staggering feats are possible. People used to memorize the entire Bible, for example.

[A.S.]

That memory can be unreliable is undeniably true... memory can also be trained, and with training, staggering feats are possible.

But remembering pages of numbers is not at all the same as making very fine subjective (or even objective) judgements. If we could observe and then precisely remember for subsequent and accurate further comparison what exact shade of blue your shirt is, or what colour you painted your walls, then man would not have developed the great need to create fixed, unambiguous reference colour charts. References so that we all have a comparative framework, globally.

We are comparative animals. There is not one sensory input that we detect that is an absolute. Even the detection of motion is dependent upon our location (in the universe) relative to the moving object. And our total inability to make absolute measurements about anything is the root of the problem when comparing, as a trivial example, the nuances of audio equipment A some moments later with B. This really is so self-evident that I don't need to restate the obvious again. No professional photographer would attempt to reliably grade different camera lens distortion from peeking at two photographs minutes or hours apart. That would be nonsensical. But if both photographs were put in front of him, after flicking instantaneously backwards and forwards between them for perhaps many minutes or even hours, those tell-tale optical distortions may slowly become apparent. And it is precisely the same comparative process that we need to employ to grade any sensory input, including of course, our hearing. Then we give our ear/brain a fighting chance to tease-out the differences, reliably.

What we are good at is relative, side by side 'instantaneous' comparisons. Then the subtle differences are magnified in our brain so that we can begin to grade them. Yes, side by side I can now see that my blue shirt compared with yours is slightly more green-blue. And critically, we must be able to compare side by side - not to rely on memory.

We have to work with the limitations of our senses not against them. To get the best from any of them we must have a comparative framework and the ability to switch between the reference and the new example. Of course, it may not matter a jot whether our shirts are this or that shade of blue. But when investing hard-earned money in technical equipment - a camera, TV or audio equipment - than surely we want to make the best informed decision that we can. Because if that camera has unacceptable lens distortion that we didn't initially notice, we might well have to dump it and buy all over again.

I'm keen to emphasise the importance of comparative listening because under AB testing Harbeth speakers really do show up the colorations and problems latent in others. The quick switch-over is devastatingly revealing and highlights many sonic defects not obvious to the casual listener relying upon his memory.

I'll just look out and prepare a spy-cam video I made when developing the M401 using my foot-switch change-over comparator to make instantaneous evaluations with the earlier M40.

[A.S.]

...What we [humans] are good at is relative, side by side 'instantaneous' comparisons. Then the subtle differences are magnified in our brain so that we can begin to grade them...

To seriously evaluate loudspeakers so that you can work with your senses and take maximum advantage of their excellent comparative ability, but non-existent absolute ability, some sort of instantaneous switch-over is necessary. In my case I made a box containing some relays, operated by a foot switch, which route the amplifiers speaker output to one or other speaker (or speakers) under test.

Here is a very simple video I made a couple of years ago showing the concept. The sound is not great because it's from the camera's microphone on the other side of the room. But if you listen carefully you can just about hear the difference in sound character when I press the foot switch (slight audible mechanical click picked-up by the microphone as the change-over relays operate).

[EricW]

...

We are comparative animals. There is not one sensory input that we detect that is an absolute. Even the detection of motion is dependent upon our location (in the universe) relative to the moving object. And our total inability to make absolute measurements about anything is the root of the problem when comparing, as a trivial example, the nuances of audio equipment A some moments later with B. This really is so self-evident that I don't need to restate the obvious again. No professional photographer would attempt to reliably grade different camera lens distortion from peeking at two photographs minutes or hours apart. That would be nonsensical. But if both photographs were put in front of him, after flicking instantaneously backwards and forwards between them for perhaps many minutes or even hours, those tell-tale optical distortions may slowly become apparent. And it is precisely the same comparative process that we need to employ to grade any sensory input, including of course, our hearing. Then we give our ear/brain a fighting chance to tease-out the differences, reliably.

What we are good at is relative, side by side 'instantaneous' comparisons. Then the subtle differences are magnified in our brain so that we can begin to grade them. Yes, side by side I can now see that my blue shirt compared with yours is slightly more green-blue. And critically, we must be able to compare side by side - not to rely on memory.

We have to work with the limitations of our senses not against them. To get the best from any of them we must have a comparative framework and the ability to switch between the reference and the new example. .....

Yes, fine, but surely you must accept that there's not one audio dealer in the world who is set up to do instantanteous speaker switchovers? Does the consumer's process of selecting a Harbeth by normal means then have no validity whatsoever? Doesn't that follow from what you're saying?

[A.S.]

...there's not one audio dealer in the world who is set up to do instantaneous speaker switchovers?

Maybe so, but that's only because they have unlearned how to do that. Back in the 1950's the dealer comparator was the norm. It may have sometimes produced incorrect results in the hands of a manipulative salesman but it gave the sceptical consumer at least some baseline.

I am not asking that the dealers should change their modus operandi - far too late for that - what I'm pleading is that you, the consumer need to be far, far, far more cautious in the absence of any baseline whatever. Otherwise, you are just setting yourself up to make purchases that don't give long term satisfaction. And that 'churn' is the reason this industry has such a poor image with wives, girlfriends and bank managers. It's not the producers/sellers fault - they're merely taking advantage of mass confusion to their commercial advantage as they are required to do by their shareholders.

One of the advantages of a long memory is being able to draw comparisons with industry practices long forgotten. Maybe many of our younger readers would be surprised how audio equipment was demonstrated by specialist dealers fifty years ago. The norm, at the start of the industry, was to give the consumer the opportunity to hear equipment (amplifiers, tuners, speakers, cartridges and turntables) side by side and in almost infinite combination using a comparator. It took me just a minute or two to find and scan pictures of perhaps the most famous Largs of Holborn unit: "Take one thousand switches ... three miles of wire ... and the ingenuity of our Mr. Barrows ...".

The design of such a massive instantaneous switch-over system was a huge engineering challenge. It must have cost a fortune. It was provided as a means of letting the consumer make the best possible informed decision of any combination of pickup, amp, tuner and speakers. He could hear the entire system and a few clicks of the switch later, hear an entirely different combination. Yes, doubtless the cabling and switching degraded the signal (a little) but it probably disadvantaged all equipment equally so could be ignored. But the essence of the comparator was to avoid the listener having to remember - he could switch back and forth to his hearts content. All day. Until he was ready to make a purchasing decision based on a proper ABCDEFGHIJLKMNOPQRSTUVWXYZ comparison.

If you look carefully, you can see highly respected brands - QUAD, Leak, Wharfdale, Rogers and others - who were seemingly happy for their equipment to be ruthlessly compared with all signals in and out through the comparator. So when I'm using my changeover foot switch (see video above) to compare speakers, that's not a procedure I've dreamed up out of thin air - that's how even the consumer made reliable comparisons half a century ago.

Pictures from 1957 ('Our 80th birthday') and 1965 ('Take one thousand...')

[yeecn]

The notion of fast switching as the only reliable basis for sound comparison can be counter intuitive to a lot of people, as everybody knows that we can recognize voices, and can tell people apart form the voices alone.

It was confusing for me - as after spending a few months soaked in the sound of Harbeth, when I have the chance to hear other speakers I can usually sense that something is different in the sound presentation, and will maybe make a comment or two like the mid is too accentuated, the bass is not integrated etc. So something about the sound signature of Harbeth has made an imprint in my mind somewhere.

I like to investigate what constitute our memory of sound/voices - and whether it can be made a basis for meaningful sound comparisons.

I like to relate an incident that Alan mentioned, where he was given a blind demonstrated of a speaker without being told that it was. He commented that he was impressed how good the speakers sounded, and was really delighted when he found out that it was SHL5.

When I heard that story it impressed upon me that Alan did not recognize the sound of his own creation! Later I found out that he does not have a hi-fi set at home - and sometimes he would go without listening to any hi-fi for months on end, in order to maintain his objectivity. When I put these two together it dawned on me that Alan was deliberately not allowing the sound of his previous creations to take hold on his mind - so that he will have a free mind to create the next masterpiece.

It was quite a remarkable revelation for me. One would imagine that a great speaker maker would be somebody with a deep passion for music. But now we are seeing Alan making a deliberate sacrifice so that all of us can enjoy his masterpieces.

So what constitute our memory of voice/sound? I have some notions of what it could be. While I am formulating my thoughts I would like to hear what other have to say about this subject.

[EricW]

But the question remains: given that no retailer uses comparators today, valuable as they may have been, on what rational basis does a customer choose a pair of Harbeths over any other loudspeaker if auditory memory is truly that unreliable?

Is it possible that, even if accurate comparisons from memory are impossible (i.e. comparing A and B to each other), both A and B can separately be compared with C, where C is defined as the sense of pleasure and enjoyment (say on a subjective scale of 1 to 100) that the listener gets from a particular loudspeaker?

If there's not at least some validity to this, then why does anyone prefer a Stradivarius to a student violin made of plywood, or a Harbeth to a mass-produced $50 speaker in a plastic box? Does there not have to some constant frame of reference with at least a degree of reliability, even if not perfect reliability?

[A.S.]

Yes, we have a very handy constant frame of reference right under our nose - literally.

Most of the artifacts (or coloration) that we associate with loudspeaker reproduction cannot be attributed to the live human voice box. So speech, reproduced over those speakers switched via the comparator, is invaluable for exposing what simply could not be present in the original voice but is (sadly) present to one degree or another in the reproduced voice. The real-live human voice doesn't 'do' spitty, wiry, gritty, grainy, pinched, peaky, barking, biting etc. - so if you hear those (and other) characteristics brought to the fore by instantaneous cross-comparison you have educated yourself. And self-education is hugely valuable when selecting audio equipment. It doesn't matter a damn what I think of this speaker or that: what matters is your objective ability to grade them and hear the problems. Problems which might drive you mad after a week or two. You don't need test equipment. What you do need is the ability to directly and quickly compare one with another.

The point I'm trying to make (badly?) is that you, the consumer should take more responsibility to select the right equipment for your needs - Harbeth or not - after making the best possible product comparisons. It's not that difficult to train yourself to evaluate hi-fi equipment, and by far the best way is to start by making fast, switched comparisons without moving from your listening chair - hence the comparator or relay change-over box.

The comparator is only one part of the 'kit' needed by the curious listener. It is not the entire solution. But when the listener grades speech by making instantaneous comparisons of how the voice is reproduced over a selection of loudspeakers, then he is beginning a journey into decoding and interpreting what he hears. He's unconsciously building a mental model of the characteristics of the speakers based on how they behave on speech, which humans innately know very well, and when he turns to music, may help him be more analytical and hence make the right purchase for his needs - Harbeth or not.

[yeecn]

Music and the Human Brain

Historically, much of the knowledge of specialized functions of the brain was deduced from the failure of the functions following a stroke or an accident. Recently magnetic resonance imaging (MRI) has shed a lot of additional light on how the brain operates. Nerve impulses from the cochlea arrive at the brain and are first processed to extract specific categories of information. Additional levels of processing eventually result in our emotional reaction to music.

Levitin lists eight "dimensions" of music, meaning attributes that can be individually varied without affecting the other dimensions. Among the most important dimensions are: perceived pitch; rhythm; timbre; melody; and reverberation. It seems that the brain contains specialized modules for extracting each of these attributes. The evidence for this is that brain damage can cause a loss of one of the functions without affecting the others. (There is similar evidence in the case of speech, where it is possible to lose an ability to speak verbs, or to speak nouns, for example). The brain appears to be a collection of highly specialized modules that are seamlessly integrated by an hierarchy of higher order modules.

Consider the perception of timbre, the attributes that distinguish a saxophone from a trumpet. When both are playing the same note A4 each instrument creates a fundamental tone at 440 Hz, and the same spectrum of harmonics at frequencies of 880, 1320, 1760 Hz and so on. The relative amplitudes of the harmonics, and their variations over time, are what give the instrument its characteristic sound. Nerves from the regions of the cochlea excited by these frequencies send signals to the brain, and presumably a brain module recognizes the different pitches, similar to the way we recognize different colors - although at this level the recognition is probably not conscious. Now what? How on earth does the brain disentangle the overlapping saxophone and trumpet harmonic series, and re-assemble them, so we hear two distinct instruments, each playing a single note, rather than a mishmash of pitches? Levitin suggests that a difference of a few milliseconds in time between the arrival of the two harmonic series is the basis for this amazing feat. Directional clues might be used as well. I also suspect that our memory of what each of these instruments sounds like when played solo helps in this process, . The brain module that does this will even fix the bass response of a inferior stereo system. If an instrument plays a note that produces tones at 39, 78, 117, 156 Hz and so on, but your stereo system can't produce a 39 Hz tone, your brain will fill in the gap and you will hear a 39 Hz pitch! This phenomenon is called "restoration of the missing fundamental." All of this processing occurs automatically, largely in parallel, and without any conscious effort.

Different pitches heard at the same time are processed by a module that extracts harmony. Different pitches heard at different times are processed by another module that extracts melody. Again we know this because Sacks describes patients who have lost one capability without affecting the other. Sacks describes a gifted musician who had a stroke. Suddenly he was unable to recognize a tune as simple as "happy birthday." Yet his perception of pitch and rhythm was intact, and he could read music and hum a melody. So the problem was specifically an inability of auditory processing of a sequence of pitches.

Emotional response to music is arguably the highest response level - although the most primitive areas of the brain appear to be heavily involved. It also apparently arises from a specialized part of the brain. Sacks gives examples of people who have the full tool kit for music processing, including a strong emotional response, and then suffer an accident. They retain the ability to perceive all of the structural characteristics of music, but completely lose the emotional response. He quotes one patient: "[music] had always been the primary unfailing source that nourished my spirit. Now it just didn't mean anything."

There are indeed a lot of complex cognitive processes happening inside our brain. There are 'modules' that allows us to perceive/interpret various aspects of the sound stimuli - but I have not come across any evidence that our brain is able to 'remember' raw, uninterpreted sound.

This means that when we hear that something is different, for example saying that a speaker has a congested mid range - what we are saying is that we have to adjust the internal circuits to interpreted the new sound signature. Research has shown that the human brain is extremely adaptable. Even with a pretty bad sound system it will sound pretty good within 20 minutes!

Sound comparison is not what it seems on the surface!

[STHLS5]

......
There are indeed a lot of complex cognitive processes happening inside our brain. There are 'modules' that allows us to perceive/interpret various aspects of the sound stimuli - but I have not come across any evidence that our brain is able to 'remember' raw, uninterpreted sound
.

In a report published on 27 May 2010:-

"The researchers discovered that repeated exposure induced learning for totally unpredictable and meaningless sounds. The listeners were better at detecting repetitions within noise samples that had been presented several times than new noise samples, showing that a new auditory memory had been created. "The sound memories were formed rapidly, with performance becoming abruptly near-perfect, and multiple noises were remembered for several weeks," reports Dr. Agus. "

Source

ST

[yeecn]

In a report published on 27 May 2010:-

"The researchers discovered that repeated exposure induced learning for totally unpredictable and meaningless sounds. The listeners were better at detecting repetitions within noise samples that had been presented several times than new noise samples, showing that a new auditory memory had been created. "The sound memories were formed rapidly, with performance becoming abruptly near-perfect, and multiple noises were remembered for several weeks," reports Dr. Agus. "

Source

ST

I pondered about that report a few days ago. I believe that it is pointing to the pattern recognition capability of the human brain.

This is how the human brain works. When presented with a stimulus the some neuron pathways will get stimulated. The neurons along the pathway will remain in an excited state for some days. They will have a higher chance of being stimulated again against other stimuli. If the stimulus pattern is repeated the neuron pathway became permanent, and became a recognition module in our brain.

As the above experiment shows even the stimulus was only a meaningless pattern embedded in random noises, the mind is able to form a recognition circuit to recognize the pattern after some repetitions. It is a memory of some sort - but it is interpretive in nature. It is several steps remote from the actual sound of the pattern.

This is how behavior patterns - good and bad - are 'learned'. This is how propaganda works as well. If you repeat some bullshit to a person, after awhile the person will begin to crap out the same bullshit. Falsehood is thus propagated.

Some memory labs are able to implant all sorts of nonsense in your brain - without you realizing how it was done. Marketing has mastered this arts long time ago. Just look around and see the amount of unnecessary stuffs in your house, you will know what I mean.

[Supersnake]

Yeecn,
You just might be the person who can answer a question for which I have never found the answer.

Why is that a musical melody played in a minor key most likely evokes a sad feeling (as opposed to a major key)? (e.g. Satie's Gymnopedie No. 1 and other compositions).

I presume that it is related to the particular combination or progression of frequencies that are within the minor key, but why?
Granted, one can embellish the melody to increase the likelihood it will evoke sadness by playing it slowly instead of rapidly, or play it at a low volume instead of high - but at the basis of it all is that it was played in a minor key.

[A.S.]

...what I'm pleading is that you, the consumer need to be far, far, far more cautious in the absence of any baseline whatever. Otherwise, you are just setting yourself up to make purchases that don't give long term satisfaction...

One of my pet subjects is the fallibility of the human senses including vision and hearing. Anyone who places undue confidence in any of our senses is just daft - so the world of science and measurement evolved to augment the limitations of our senses, enabling us to make better informed decisions about the environment around us.

Ignoring hearing - the main topic of this HUG, how about our visual sense. How reliable is it? How reliable even when we make the equivalent of an AB switchover, flicking our eyes between adjacent images?

Judge the shocking facts for yourself. It was so hard to accept the result that I opened and checked the image in Photoshop as you'll see on the last page. My point is that even presented with the means of making an immediate, instantaneous side by side direct comparison, not relying on memory, we still cannot arrive at a reliable conclusion. So 'comparisons' - i.e the ghost of a vague memory - separated by time - minutes, hours, weeks are completely worthless in my opinion.

[Supersnake]

I too had to check the RGB for the two squares and they are exactly the same. That is scary.

Alan, during your designing stage - when you are listening to the playback of musical sounds and tweaking various components within the speaker to produce the sound that you are seeking - do you take into consideration that you may be affected by the phenomena of fallibility in what you think you are hearing and defer to making the adjustment based upon what some sound meter denotes? Or is there a place for welcoming and incorporating that inherent fallibility because you are designing your speakers for human ears to hear.

Are there conditions in which you would tweak using results from an instrument over what you hear and vice-versa?

[A.S.]

Yes, it's really concerning that, in a courtroom situation, I think we would both swear on oath that one square was unquestionably darker than the other. But as we've seen by simple meansurement, they have equal colour characteristics. What has swung our perception is the presence of a pattern and the larger object casting a shadow. So, we have been hugely influenced by not the core subject (A v. B) but peripheral issues. I'm sure our ears are equally susceptible to being tricked - yet we would say, with unshakable certainty - that we truly believed A sounded better than B. And we'd mean it. But we could be completely wrong, as we are with the colour squares.

We are designing speakers to be listened to by fellow humans. The trick is to use the measuring and designing tools to keep the technical, objective measurements good (for example, flat-ish) yet to make the speaker sound natural. I've illustrated two aspects of the design process here and here in videos. In fact, now I look at these I can see that in reality, the way I work is actually on a daily basis a combination of the two ... that is, I am measuring as I listen as I tweak as I re-simulate. A four-way design loop. Which is why it takes so long to balance the hard objective data with an acceptable subjective sound. After all, when being peer reviewed, good measurements are expected.

The turning point for me was about twenty years ago at, as I recall it, the London Penta hi-fi show. I had hurriedly assembled a pre-production pair of speakers the day before the show and hadn't checked them carefully. The sound at the show to my ears and the public visitors was beautiful. I was completely seduced by my own creation. Imagine the sweat that broke out a few days later, back at base, when I discovered that I'd wired both tweeters out of phase. I should have heard that, but I was so romantically wrapped-up in the experience I dropped my objective guard. Since then, test equipment has always been a few steps away. You can hear what you want to hear.

It's to be expected that the public, unfamiliar with the speakers room or music would not hear my stupid mistake - but it was shocking indeed that I, purporting to be a professional, after three days exposure didn't.

P.S. There is a small (and only small) justification in my favour. I was standing continuously at the rear of the room, well above tweeter axis. The actual phase of the tweeter so far above axis would probably not have been material. But for the public, seated more or less on axis, the effect would have been magnified. Since then, I've not completely trusted myself sans test equipment.

[EricW]

Yes, it's really concerning that, in a courtroom situation, I think we would both swear on oath that one square was unquestionably darker than the other. But as we've seen by simple meansurement, they have equal colour characteristics. What has swung our perception is the presence of a pattern and the larger object casting a shadow. So, we have been hugely influenced by not the core subject (A v. B) but peripheral issues. ....

But not just any "peripheral issue" in this case, but rather, one which clearly been carefully designed to fool the human perceptual system. The shadow is obviously the key: the brain continues to see the contrast between light and dark squares under the shadow, and generalizes that contrast across the board. In other words, the brain creates a perceptual map of what it "knows" is there, even if it's not what actually is there. What no doubt aids in this illusion is the fact that, in the real world, we have a very strong bias in favour of seeing chessboards with just two alternating tones or colours, and we really don't expect to see any variation.

I've been thinking about this a lot recently as I've been contemplating the purchase of a new pair of Harbeths. Again, I've had the opportunity to compare my P3ESRs to the Monitor 30s, the Compact 7s, and the Super HL5s. I love the articulation of the P3ESRs but I want something bigger. I almost pulled the trigger on a pair of Monitor 30s but then reconsidered. Perhaps wrongly, I don't know.

My ultimate yardstick was not so much a comparison of the speakers to each other - you know, "this one has more detail, that one has better bass, etc. etc.", the usual hi-fi, audiophile stuff. Instead, I tried to focus more on how I felt when listening to each speaker. Was I involved? Removed? Tense? Relaxed? How much did the music grab me? How much did I want to listen to the next track? The next disk? And so on.

And by that yardstick, the comparisons became a little easier. The P3ESR is a wonderful speaker, but it's a "head" speaker - it doesn't have the bass necessary to let you feel the music below the neck. It's analytical: you hear it, but you don't feel it. I guess it's a kind of illusion.

The M30, to me, is similar, though with more body to the sound. It's a wonderfully detailed and articulate speaker, and I have no interest in critiquing it in any "audiophile" way - it just wasn't quite as relaxing to listening to as I would like. I don't know why, but there it is. The Compact 7, which has gotten universal raves, didn't involve me as much as the M30, and I wasn't crazy about it. I couldn't point to a single fault with it, however. It's just how I felt.

But the Super HL5, ah, that's a wonderful, warm relaxing sound. Even in my small study, everything sounded great: Mozart, jazz, rock, you name it. I enjoyed it all. Don't get me wrong: it has all the clarity and detail one could ever want, but somehow it all adds up to a sense of relaxed, sensual involvement in the music: it's not as much an "in the head", analytical kind of experience as the others.

Note, however, that I am described only my own subjective reaction to the various speakers, in a sighted unscientific test. However, that's how I actually would use them, so I think that's okay. For me.