No real reference

April 18, 2019
 by Paul McGowan

Wandering the halls of an audio show it’s always stunning to me how different the same music sounds.

How can we say there exists a reference?

If we’re so bold as to suggest there is a reference, like that of the sound of live unamplified music in acoustic space, how do we justify the major differences in audio system reproduction?

Could it be that as a group we’re so far away from the true reference that our meager attempts are laughable at best? Or is it that we all hear differently? That what sounds right to me doesn’t sound right to you?

If this were true then a hundred people enjoying a concert would all hear something different. We’d have trouble agreeing that it sounded real, yet that doesn’t happen.

Or does it?

Could it be that because we’re actually at the concert no one would be silly enough to openly question whether it was actually live or not—yet none hear the same thing?

I don’t suppose there’s a way to easily answer this question but I suspect if we compared listening notes there would be as much divergence of sound quality per listener at a live concert as we do in a reproduction setting.

I would wager that the only real reference is what’s inside your head.

Subscribe to Paul's Posts

68 comments on “No real reference”

  1. Of course people hear differently. I would have thought the use of the term “reference” in audio should be limited to what each manufacturer considers their best product, against which their products can be measured. To consider their products to be “reference” for other manufacturers is presumptuous and conceited. The market will decide that by what they buy.

    Also bear in mind reference is a relative term, implying comparison to other (inferior) products. Many manufacturers, just make products and don’t compare them to other products, either their own or by other manufacturers. Harbeth is an example, as they do not have an upgrade path. Each product stands on its own merits. Life is easier when you can accept a product is for you rather than always having to compare it to other things.

  2. “I would wager that the only real reference is what’s inside your head.”

    Do you mean like “you’re only pretty as you feel?” It boils down to musical experience. After all, hi-fi playback systems should be all about the recreation of a live musical event.

    If you invest enough time with music, the impact of the human condition and emotional being on the ear -> auditory nerve -> brain connection, delivery and process system will be minimal.

    1. Comparing one system to the next really isn’t objective, with a music background one draws from their aural memory which may just have been the jist of Paul’s last sentence and comment.

      Does the system reveal the recording?
      Does it sound like real music?
      Can you sense the location of the microphones in the venue and can you get a sense of the space and air of the room the recording was made in.
      In many cases can you hear the mixing console, the actual space of the studio and the effects layered into the recording.

  3. I’m sure (not only) due to the different sound dispersial from an instrument vs. a speaker in a room, live can easily be distinguished from recorded. On the other hand, if a blind test live vs. close(er) mic recorded in a concert environment (artists and speakers further away from the audience) was possible, I guess many would be disappointed of the live sound, as it doesn’t sound as immediate in many ways, but better in other ways.

    "I suspect if we compared listening notes there would be as much divergence of sound quality per listener at a live concert as we do in a reproduction setting"

    I think this is not to the point. Put the same piano player, a singer or a cellist or a string quartet with their instruments in any of the rooms at Axpona…it wouldn't sound in any aspect as different as the hifi setups do and therefore also not as different to the listeners. Therefore I also think that listeners in the same concert wouldn’t have that kind of different opinions as they’d have from speaker playback.

    1. Hi to all, I probably do not attend as many live performances as some on this blog. I go to the Montreal Jazzfest every summer and to a few live concerts at other times. I’ve noticed that sometimes I prefer the sound in my room and usually though live will sound better. I suspect many factors are at play here, too many to get into the why.

    2. I’ve witnessed more than a few comparisons of live vs. recorded of the same musician playing the same music at audio shows and there is indeed very little difference when the system is reasonably good. Obviously, though, hearing live in a concert hall will sound different than the same recorded music played back in your room at home. The room imparts a significant signature on the music, which is why there is such a huge market for room treatments and tweaks that make your perceived music sound like you’re in a much larger space.

  4. Based on which criteria could you detect, Paul, that the same “reference” music sounded different? Did you have your own reference tracks being played or were there different recordings? Did you always listen from the same sweet spot. Obviously the room acoustics were not identical. If even capacitors or resistors showing identical values are said to “sound” different there is no way that two different stereo systems sound identical. Yes, people hear most differently based on their actual mood and hearing abilities- the latter varying during 24 h. And finally our senses are based on a complex pattern recognition process, patterns which are trained and learned most individually. And finally having different senses for perception a perception is a multi-modal process based on all senses. We rarely can sense pure visually or aurally and exclude the others senses. Thus again my question: where do you focus your attention when listening to your reference track?

  5. Hi @all.
    I think it is a dfiference to compare the music out of speaker drivers with a real concert. To many imponderables are there. What mic has made the record, in which environment (live or studio) and so on. I think there is a limes of sound against real what we want to have and yes, there are personaly likes like the light of a LED lamp (warm, or day light).
    Best regards

  6. Speakers and live music in the same room, at the same sound pressure should reveal whether a reproduction system is faithful to the original. I believe it was it was Peter Walker (Quad), who did a test with a musician and his loudspeakers + amplifiers behind a curtain.

    1. It was done by Peter Walker and Gibert Briggs, the founder of Wharfedale Loudspeakers, in 1954 and 1955. Several concerts in the newly built Festival Hall. It was written up at great length over several issues of Gramophone Magazine, which I read a while back (I’ve subscribed since the 1980s and they are all online).

      It was repeated at Carnegie Hall on 9 October 1955:

      OCTOBER 10, 1955
      The English sound engineer Gilbert A. Briggs, who sold out a London concert hall for his recent demonstration of recorded sound equipment, duplicated this feat at Carnegie Hall yesterday afteroon. He proved Hi-Fi enthusiasm to be at an equally feverish pitch on both sides of the Atlantic. A feature of the demonstration was the playing of records in conjunction with live performance of the same works by the same artists. One listener found the pipe-organ demonstration most impressive. Mr. Briggs’ battery of sound equipment reproduced i t aU, from the highest squeal of the” mixture .. stops to the boom of the pedal diapason. An observer looking away from the stage could not detect the moment at which the record stopped and E. Power Biggs, at the console, took over. The mellow tone of the Philadelphia woodwinds, too, was projected with almost startling fidelity to the sound of the players in person.”

  7. Of course they are going to sound different as you walk around an audio show, every room is different, the speakers are set up different and interact with the rooms differently, and even though the music is the same it’s being played at different volumes. Every room could have the same exact sound systems and sound different if they are set up differently and played at different volumes.

    1. There’s probably many very high quality speakers that can pass as references. But if you had to choose one there’s no doubt you would get different opinions.

    2. Even if each room was exactly the same room, with the exact same setup, and sounded identical with no people in it, once people start entering the room, the sound will change, and will continue to change as positions, sized and shapes of people change. After all, don’t we all act as room treatments, at least to a certain extent?

      1. The best speakers, or the worst, retained their basic issues, good or bad, regardless of the qty of people in the rooms. Your ear can quickly detect a speakers deficiencies. Empty or full, the basic presentation remained.

  8. We don’t “hear” differently but we may pay attention differently. I have often said that a man with a bad left ear and a man with a bad right ear take those characteristics to the concert hall the same as to the listening room and will agree when reproduced sound mirrors the real thing.

    1. No, we definitely hear differently. This is because we all have different sizes and shapes of ears, different ear canals, and slightly different composition of every component affecting hearing. Also, does one person have a cold that might affect hearing, or something else that might temporarily affect his/her hearing? Also, our brains will process the signal it gets differently. So, we do all hear differently.

        1. But, that is an impossibility, as it is impossible to replicate everything related to the original. The key is to get close enough to trick our brain (and or memory) into believing the reproduction is accurate.

        2. There is a basic difference between real life listening and „artificial“ stereo listening: stereo mandatorily requires absolute symmetry even for both ears. If there is one “bad” ear and one good ear you fail to get absolute symmetry and the phantom effects of stereo are severely degraded! The result maybe a shifted center image and other localization deficiencies! The phantom images of stereo have totally different patterns compared to natural sound images!

          1. Not so sure about “severly degraded”. My ears are unbalanced by about 0.5 to 1.0 dB shift on the balance to the left, looking at the speaker.

            To place an image ACROSS a plane in front of you we need stereo. That is what we have to work with, and we are doing OK with it.

            Certainly we can understand a real instrument is automatically placed in the sound stage and thus far more stable and easier to listen to, even if it sound awful, and most live events I go to are awful sounding but are wonderful experiences.

  9. All this is compounded by the fact that even at live venues, the same person hears the performance differently from each position he might be in. A front row seat might sound very different than one four rows from the back. Does that make one more or less real than the other??

  10. As a long time performing musician, I can state with confidence that different people hear music – and everything else – differently. Getting on the same page with other musicians is an interesting experience.

    1. Different yes, but do you think people would argue about the sound, sound quality or sound characteristics as intensively with such widely varying opinions as they do when listening to a specific high end setup?

      1. The answer from my experience is yes. Different people I know fail even to identify which instrument is making which sound, while we’re listening to a composition at the same time through the same system. Timbre and voicing are perceived differently. Not to mention, some lucky people have perfect pitch – though there is some question as to whether that has to do with perception or analysis.
        Then there is the simple fact that everyone has their own frequency response characteristic. I can’t tell you the number of times that people were so sensitive to bass notes that they were satisfied only after I turned my bass amp off (true story). In my early twenties, my hearing went to 21KHz; I was tortured walking through the TV department at Sears because of the 18.5KHz carrier frequencies. (Unfortunately, I’m closer to 14KHz these days).
        So between hearing acuity, personal frequency response preferences and distinguishing of timbres, little question in my mind that different systems would appeal to different people – and even two “golden ears” could disagree on any one setup.

        1. I can follow your conclusions.

          All this is very interesting because I had some discussions with Paul and Ted somewhere in the forum if hearing loss plays a role when evaluating the absolute sound characteristic and tonality of components during voicing. Both, as far as I understood, are quite convinced that the brain compensates for hearing loss so that people suffering from it more or less perceive the same as people with full hearing capability. This would contradict the majority of your thoughts in my logic or your thoughts would contradict that hearing loss logic.

  11. My reference is a favorite system that I frequently listen to. It used to be a friends system, but now is my own system…but in a friend’s room. We actually did that experiment.

    Experience has shown me that the room is as important to a system being my “reference system” as the equipment itself.

    1. Both the space where the recording is done, and the space where reproduction is done, probably contribute more to the final experience than anything else in between.

  12. When I used to go to audio shows, I would make a deliberate attempt to not look at the speakers or equipment when I entered the rooms. It’s very difficult to separate one sense from another. Vision influences hearing, smell influences taste, and so forth. These days, I’m more concerned with whether the sound pleases me than I am with it’s comparison to an “absolute” standard, whatever that is.

  13. We are born unable to recognize any sound. It is all a jumble of neural firings. We learn to hear by millions of examples, and the sounds we recognize as spatial probes need to be closely related to specific acoustic sounds we hear daily like humans talking. Dick Olsher used “the Leslie test”. Because he heard his wife sing every day, her voice was a valid system reference – for him.

    Back in the 18th Century, before engines and motors, the elites of Europe and the newly professional musician/composer class heard violins, lutes, harpsichords and organs played daily. The richest had a wind section as well. By the end of the Renaissance, luthiery had advanced to unprecedented heights as had performance practice and ornate composition. Some pieces – masses, oratorios, passions and operas – lasted over three hours. This group of people had absolute references, Stradivarius and Guarneri string consorts and Ruckers harpsichords.

    I hear harpsichord and/or piano daily, and attend 70 orchestral, operatic and chamber music concerts a year (down from 150). Further, I have 800 concert recordings of chamber music that were made straight from ORTF pair (AKG C480B) to hard disk in DSF format. This comprises an absolute reference because I was in the front row or three for every recording. I also have many hundreds of CDs purchased at concerts so I heard the the same program by the same performers within 24 hours – and this caused me to give up on even boutique label 18th Century music recordings. By now I can’t stand to listen to Internet music, broadcast radio, DGG or Sony Classical.

    Maybe I am spoiled by subscription seats in front of the halls, but it almost always sounds like other people’s recordings have too much reverb. I think this is because I put a priority on articulation, hearing how the notes start, stop and transition, in essence the consonant phonemes of music. Commercial speaker designs, whether home stereo systems or studio monitors, have considerable time distortion so they do not reproduce the waveform of consonants accurately. Further, if they DID reproduce musical consonant waveforms consistently that would serve as a spatial probe signal to locate the speaker and the stereo image collapses! I have done this experiment hundreds of times on hundreds of people. The universal reaction is mono sounds better.

    “Imaging” and “Soundstaging” require the speakers to be locationally vague. If you can hear where the speakers are, then the stereo illusion is revealed for the fakery it is. The most common way to do this is narrow, sharp edged baffles with the drivers mounted symmetrically. This creates a lot of treble diffraction that scrambles the phase of off-axis waves that bounce around the listening room so you can’t locate the speakers aurally. This also causes mild comb filtering on-axis which is a trade-off for designers flatening the on-axis response.

    A way around this trade-off is high order crossovers which introduce phase wrap, mangling consonants until they are not recognizable to people who know what music sounds like. The speaker design is further abetted by studio engineers who have even more compulsion to twiddle and tweak than audiophiles. Every knob in a recording studio introduces temporal and/or spatial distortion of the signal, and recording engineers get paid by the hour to twist knobs. In the recording business, the engineer with the most knobs is perceived as king. There are hardly any records with low enough time distortion to serve as a reference, as the microphone pair (no mixing or mastering allowed!) needs to be close enough to the music that the direct sound is louder than the reverberations, i.e. inside the Schroeder Limit.

    SO yes, everybody designing and shopping speakers hears differently. Their hearing is trained to the temporal distortions of their recording collection, speakers, ancillary equipment and ROOM. Even professional musicians (with little overlap into the audiophile community) compartmentalize live music and reproduced music. The two are so different that their brain flips the switch in the decode algorithm within 100 milliseconds.

    I have bridged this gap, but it means throwing out the fixed channel paradigm. Since well trained hearing can detect where time-accurate speakers sit, you need a separate speaker to physically represent the location of every instrument in the recording. The upside is that music reproduced with temporal accuracy has greater depth of meaning so it enriches your musical life.

    For those in the New York area, there is an upcoming concert demonstrating the largest scale system to date – a double string quartet of two cello speakers, two viola* speakers and four violin speakers playing back an eight track close miked studio recording. This is with live performance of a premiere by the JACK Quartet, who are also the recording artists, so you can hear a direct live/recorded comparison of the prototype world reference sound system.

    8PM, 30 April 2019, Bronx Community College.

    *the viola speakers for this concert are actually two prototype voice speakers – let’s see if you can hear the difference

    1. Yes, a mono speaker designed for each sound isn’t a reach to understand. It is a reach to do it! I’m getting down to one system with two speakers let alone 22.

      Curious, I, and many others in my day, went to BAND, or CHOIR, some to both, all through junior high and high school. It seems the once you hear real music, instruments or voice, you immediately recognize it as your reference years later. I’m a sucker for an oboe!

      My memory of sound seems far more permanent than you suggest. True, more often can fine tune but that initial exposure seems to be a higher reference point than those that avoid the arts, or gad don’t have them, in school.

    2. “prototype world reference sound system”

      All it is doing is replacing instruments with speakers. The instruments may be close-miced, but the other instruments would only be a few feet away so sound would not be isolated to each player, and quartet players have to be close together to play properly. Doesn’t say anything about the quality of the speakers. Seems aimed at eliminate the need for stereo speakers to image a quartet, which mine do admirably, and I’ve heard plenty of live quartets (Takacs, Casals most recently) and am happy to listen to quartets in stereo at home. No guarantee that the dispersion of the speakers is the same as the instruments.

      Even if it is a “reference system”, it is for a quartet only, and I have no reason to believe it is, and it is not much good for anything else.

      All this illustrates is that loudspeaker designers have to compromise to meet the wide listening preferences of their customer base, and for the last 60 odd years 2-channel systems have been considered both practical and generally effective.

      More specifically, take a piece like Monteverdi’s Vespers of 1610, which I was listening to last night, it has a range of musical styles and musical groupings, most evident when you see it performed live. He was showing off, the piece was effectively a (successful) job application. During the performance it is normal for singers and instrumentalists to change location, whether playing solo, duet, in the choir etc. You can hear these changes in the recording. If each instrument and voice has a separate speaker, I assume you have to rearrange them during playback. you would spend a couple of hours listening to music and a couple of days rearranging speakers.

      Makes no sense at all to me. Doesn’t seem to have caught on either.

      p.s. How do you deal wit the fact that the choir is usually raised above the instrumentalists? Do you supply a carpenter with the recordings, or specify speaker heights? Do you need different speakers for each recording? Or should we just listen to quartets as they sit on the same level?

      1. These speakers are dedicated “sound boards” for acoustic instruments, in the same way that Leo Fender designed electric instruments and instrument amplifiers as a system. You would not use a guitar amp for keyboards, so the principle of single-purpose amplification has been established commercially for 60 years, longer in art music (Theremin, Ondes Martinot).

        The following advances incorporated into the system design:

        1) The speakers are arrayed in the same geometry as a pair of string quartets – same width, same depth, same height.

        2) The speakers mimic the acoustic size and shape of the target instruments – in particular, they are double sided quasi-dipoles with similar baffle area and effective tweeter apertures similar to the HF radiating area of the respective orchestral strings.

        3) The speakers have over 12dB more headroom full spectrum, as much as 25dB higher power+efficiency and air displacement in the tweeters than 1″ domes. This design criterion was verified by Meyer Sound, who documented typical 18dB crest factor in the top octave for live music. I have measured string quartets at 115dB peaks near field. The midranges are 6.5″, 2x 5.5″, 2x 8″ and 10″. They are bi-amped except the cello speakers which are tri-amped.

        4) The drivers have substantially lower inductance (3x -10x) than typical “high end” and pro drivers, meaning they follow transients and reproduce waveforms better. The woofers are in Bessel sealed boxes for lower group delay (time distortion).

        5) The speakers have substantially flatter off-axis frequency and phase response from the unique geometry so they fill the room more accurately.

        6) In dozens of tests with solo Violin, Viola, Cello and Contrabass, >95% of audiences of conservatory trained string players could not tell the difference when the concerts were amplified. In may cases the performers were not aware of the amplification. This is an unprecedented result.

        In some cases we were able to amplify a Steinway Model D transparently for audiences. The performer always knew, but we were able to fool veteran NY Times critics and a four-time Grammy winning Classical Engineer, with the acoustic grand piano in the room for comparison.

        This has many applications: fixed media concert pieces like the upcoming one, unbalanced orchestrations, micro-sounds of extended techniques, mixed electro-acoustic ensembles with acoustic-like sound from sampled, processed and synthesized sounds to blend seamlessly with acoustic instruments.

      2. So far I have undertaken only chamber music and Jazz. Per Gabriel Sakakeeny, it should be possible to use a stereo pair of speakers to represent sections in an orchestra. I have no solution yet for woodwinds which have a different acoustic shape for every note due to the tone holes which modulate pitch.

        All of the speakers have separate treble and bass sections so the treble speakers have adjustable heights, either on speaker stands or subwoofer poles. A way to deal with crowded stages is to use a lighting truss and hang the speakers, which could represent a raised choir.

        We only attend operas in concert version – the acoustics of singers strolling around are atrocious with theatrical wings and fly arch behind the proscenium, likewise for pit orchestras. Have you listened to live broadcasts of operas? Even with dozens of mics they miss things.

        You need to understand that convenience is ruining our perception. Audio degrades our hearing, imaging screens degrade our vision and all electronic media reduce our comprehension and retention of information. Further, cognition is the basis for all thought so if you only experience a virtual world your thoughts will not be so well informed, precise or utile.

  14. Keep it simple.

    There is no reference speaker sound, just as there is no reference Cabernet wine taste.

    Individual tastes vary as much as the choices available. Still aficionados can tell the difference between good and bad.

    All that really matters is what brings you pleasure. So pour a glass, select a piece, take a sip, and turn it up.

  15. Live acoustic sound is inherently coloured by the acoustics of the space in which it performed. So if your objective is to accurately recreate the experience at home, the challenge becomes one of creating the sound of one reverberant space within another, different, reverberant space. In theory, that is possible, but within the constraints of conventional two-channel stereo, it is not.

    1. Richard,

      I can recognize a real clarinet ANYWHERE and in any room. Yes, the space makes it different, but something else is going on as the brain adjusts for the room or space and yields, “is it real or is it Memorex” every time. And no, fake hardly ever is real.

      My gut is that the harmonics are severely missing or aren’t the right AMPLITUDE relation to the live instrument to sound real. That ratio of harmonic to the fundamental seems to be what the ear is really good at hearing.

      More proper harmonics and at the right amplitude equals a more real sound. I may be wrong, but I’d rather have transient response right over steady state tones in a speaker, as long as steady state is reasonable. Hit a steady state tone with more amplitude and how accurately does it track the input? Music is dynamic.

      Even steady state middle C throws off harmonics, the timber, that a speaker has to nail in AMPLITUDE and frequency to the fundamental, so there we STILL have dynamic response and the ability to respond to ALL the harmonics.

      1. The difference between a violin and an oboe is mostly how the notes start, stop and transition in sequences, the musical consonants. Think about notes as syllables – each instrument has its own language, but the vowels are more common to the various resonances and method of oscillation. The transitions have a different waveform than the vowels, just as in speech, and vary far more with the details of the physical resonating interactions we call music.

        It is a standard example in composition studies to play a recording of alternating violin and oboe with the notes faded in and out instead of traditional technique. If done carefully, even conservatory trained ears have difficulty telling them apart.

        Speaker design has been focused on getting the frequency response flat, using sine waves as test signals or more recently the frequency component of a Fast Fourier Transform (FFT). Those are measured in pseudo-anechoic conditions (thanks to Richard Heyser, RIP) at one point in space, nominally 1 meter distant on the tweeter axis. This indicates the accuracy of vowel spectrum reproduction, given sufficient absorbtion and diffusion in the listening room.

        To reproduce music, you need to accurately replicate the consonant waveforms too. As in speech, consonants are less than 5% of the time and energy but over 50% of the meaning. Because they are so short, even non-repeating waveforms, you have to get the phase right and that means eliminating frequency dependent delays and phase shifts. We need to test with impulses, square waves, triangle waves, sawtooths or at least rectangular gated sines and optimize for waveform accuracy.

        As if that were not difficult enough, you also have to get the frequency and phase response flat off axis. There is a social reason for this, most systems serve a single listener into a sweet spot; but even if you like to spend a lot of time by yourself locked away from the world, a lot of those off-axis waves bounce around the room and arrive at your ears. Any phase scrambling off-axis will distort the spatial information in the echoing transients, and weaken the transcribed identity of the instruments.

        1. I played a clarinet, and not very well. The tone is very complex. The bell, the body, the finger holes, the reed, the mouthpiece and your face all change the tone. Worse, for recordings, every one of those has various vector amplitudes and directions onto the air. The tone is a composite entity. I say vectors, but that is a gross understatement to how we hear sound. Hearing real sound from a recording is all but impossible.

          Look at speakers. They are F=MA devices. Sure, they can play a test tone, or static superimposed tones to produce differing waveforms, but they ALL have a severe deficiency, they can not match the amplitude changes in music accurately enough, and over a broad enough range.

          Two cars go 60 MPH, that’s the current ampIitude. Ask them to go to a new amplitude in a specific amount of time and back again, and what happens? We see severe differences based on the properties of the engine and mass of the car.

          My experiences say that THAT is what we hear as good or bad, and not so much the steady state flatness of a speaker. Sure, we have to be reasonably flat but very few speakers are flat dynamically.

          This does not mean that the eletromagnetic speaker motor assembly can’t get close in frequency ranges, and this accounts for speakers sounding “real” playing specific things.

          But again, for some reason we sell steady state frequency response for speakers and virtually ignore accurate dynamic response…which is what music mostly is. Sure, a waterfall plot shows excessive RESONANCE properties, but not the instantaneous ability to track an amplitude shift.

          Speakers I enjoy are better at DYNAMIC accuracy. But frequency linearity is important, and is why I do not like horns.

          1. I am with you, rower30. Winds are one of the most the most elusive sounds despite being the oldest surviving instruments (Swan Bone Flute). The notes change timbre and shape with volume, and every note has a different shape. They have a phenomenal dynamic range and with good technique can range from pure legato, even perfect glissando, to biting staccato.

            The lips and tongue have the most finely controlled muscle movement in the human body, which together with breathing technique impart tremendous range of expression – and yet, the sound itself, however complex, has an internal consistency such that we can identify not only which family, but also which register, and as you say which mouthpiece and other fine details of construction as well as the player.

            Altogether, it is too much information for a microphone let alone a speaker. I despair of ever hearing a sound like that come to life with our Frankenstein audio inventions.

    2. Richard Murison if the microphone picks up those cues inside the live venue whether they are colored or reverberated then a highly revealing stereo system will be able to reveal that in your room at home through good speakers or even a high quality pair of headphones. Your room at home doesn’t need to have the same room characteristics of the live venue for you to hear the live venue in your home. That’s what high fidelity is all about and that’s capturing the sound of instruments in a room with a microphone and duplicating it at home which is why we get closer to that goal with better recording technology. In the case of high quality headphones there is zero room effect and yet we can hear the atmosphere of the hall it was recorded in.

      1. More often then not instruments recorded in a studio might be more accurately reproduced then in a live environment due to a studio built to prevent room coloration’s. But once again we come to the question of accuracy versus what we prefer to hear. We might actually like the reverberations and coloration’s that we hear in a live hall versus the studio.

        It’s not the job of the stereo system to create reverberations that we like but rather let us hear when it’s in the recording. If it’s not present in the recording then maybe Paul is right with his theory that the micro-phonics of tubes and even turntables create this reverberation effect that we like even if it’s not accurate or true to the recording. Turntables and analog have other virtues that I like besides the theory of micro-phonics. So to me that doesn’t entirely explain why so many prefer analog over digital and I’m sure the same could be said for the sound of tubes that people prefer over solid state. Micro-phonics is only one aspect of it and it’s only a theory.

        Myself I prefer sound reproduction to be true to the recording. If you want to add reverberation that’s not in a recording you can do that through electronic stereo reverberation. They used to sell those for stereo systems and I’m sure you can still pick those up on eBay.

      2. Right, but the comment was made in the context of an objective of wanting to hear music that sounds like it is being played in a live acoustic space. So the question is, is our objective for it to sound like the original acoustic space in which it was recorded, in which case, if the listener wasn’t at the original event, how is he to know? Or alternatively, is the objective for it to sound like a plausible facsimile of real music being played in some acoustic space … where we don’t care much what that space is, so long as we are convinced of its plausibility?

        1. The objective of recorded live music is to capture the sound of the hall and then hearing it at home on our stereo’s as if we were sitting there in the hall. On good systems set up properly you can actually hear the atmosphere of the hall and the distance from the stage that the microphones are placed so if you close your eyes it sounds as if you are in this gigantic hall with the crowd and the performers.

          1. Absolutely that’s the theoretical objective. My point is that the practical objective is “The sound of **A** hall”, or even “The sound of **A SIMILAR** hall”.

  16. One of my favorite venues is Boston Symphony Hall. It has wonderful acoustics. One time I had tickets for the 5th row. I could hear the individual instruments quite distinctly. The volume was a tad too much for my tastes. Another time I was halfway back. In this location the sound was more blended. I could not hear the individual instruments as distinctly as before. But overall I preferred that smooth, rich blend. From my experience I cannot help but conclude that you can have several thousand people hear the same performance; but there will be a variety of ways in which that same performance is heard.

    I realize that very seldom can recorded music provide the equivalent of a live experience. A few months ago I heard a performance of Bernstein’s Chichester Psalms, something I had heard only from recordings previously. That performance was so dynamic, it actually gave me goosebumps. No recording ever came close to the live experience.

    When I listen to recorded music, I’m not looking for a reproduction of the live event. I’m just looking for something that has been recorded and reproduced well enough to make the listening experience enjoyable. That’s one of the things I appreciate about PS Audio; that you are putting so much time and energy into giving us the best possible experience of recorded music that’s possible.

    1. And standing on the conductor’s place would give you another totally different sound perception. Thus the question is what does the conductor want the audience to hear! 😉

      1. >> And standing on the conductor’s place would give you another totally different sound perception. <<

        Not to mention, partly deaf… 😉

    2. The musicians are trying to tell us something, and it is in a language that is extremely detailed, more like dolphins.

      Dolphins have four sound generators which can independently produce chirps, warbles, clicks and other syllables at a very high speed. Wide bandwidth recordings reveal grammar-like patterns which one researcher claims extend seven layers deep in structure. This means they can communicate orders of magnitude faster than humans, which goes along with their enormous brain. This is like those occasional moments in opera when four principals are singing different lines (my favorites are Mozart and Bernstein), except it is one individual singing four lines simultaneously, and over ten times faster.

      Every musical instrument design speaks a different language with different phonemes. Orchestration is the art of blending those into a coherent message – but so much of the content of music is in the articulation. I find deeper meaning at close distances, specifically inside the Schroeder Limit where the direct sound has more energy than the reverberant sound fields. This is the sound the musicians hear when they are making music in an ensemble, it is the language they speak with their fingers and lips. They are talking to each other, we are overhearing the conversation and I would rather hear the words than than the sound of a crowd murmuring at a distance. Musical instruments are also faster than human voice and have more range, again more dolphin-like.

      With very clever acoustics one can maintain musical articulation over 600 seats (Zankel Hall), but that is the most extreme example. Even the 400 seat Weill Auditorium turns to mush by the 20th row. Keeping halls to under 500 seats is tribal size, and you can still recognize faces and large scale facial expressions of the people on stage.

      This is why I find it so distressing that recording engineers back the mics off to outside the Shroeder limit for Classical recordings, or add dripping ladles of artificial reverb over their mixed multi-tracks. I wonder if they were brought up in a cave. Probably their few acoustic experiences as a child were 20th row or further back.

  17. What is the final objective of a HiFi system? To me and many others it is to recreate real music in one’s home. We know of course what a violin or a saxophone sounds. Why Paul argues that there is no real music reference for his speaker to achieve sounds like he is not interested in presenting us some high fidelity reproductive HiFi item. Or is he?

  18. Your hearing would cancel out if the equation, yes? You reference to live only as you hear it, so that’s the reference, you have no way to define another one. Certainly we don’t all hear instrumented flat response the same.

    At AXPONA I went to about every room this year, whew, over three days. Of all the speakers and systems I heard, maybe 15 were what I call very good. And, that was agnostic of price. An inexpensive 3K set of planer magnetic Maggie type speakers were excellent, for instance. True they are not bass slam, but upper bass to treble they were superb. The small sub gave them plenty of vocal expanse. A great entry speaker. The small ELAC were very good, so the driver tech was also spread out to even include T+E CWT 1000-40 design, which were VERY good.

    Do I hear so poorly that 130 of the 150 speakers were really the good ones, and the 20 I liked were not? That’s certainly possible. But to my live references I hear every day, 20 or so speakers got close to that, the rest didn’t.

  19. “I would wager that the only real reference is what’s inside your head.”

    Absolutely! Experiences and exposures to live acoustical music, or for that matter-good reproduction playback sessions, are in-our-head and become “personal references” that we carry around with us till there is no breath left. Each one of us has our own unique understanding and appreciation of what LIVE is!!

  20. Paul… Take ten happily married men. (if you are so lucky to know that many).

    Not one husband would want to trade his wife for someone else. Each wife is totally different than the other wives.

    In other words? There is no reference! Its what uniquely pleases you. And, pleases others like you in a shared appreciation for certain traits.

    But, every system is uniquely our own reference. As long as the person is happy with what he has? ….. He has his reference!!!

    Malcontents want to measure everything. 😉

  21. So true, Paul. Been lobbying for “IMS” (in my system) to become a standard acronym on the Forums like IMO, and I similarly think “IMH” (in my head) would go a long way to solving the Audiophile Problems of the World ; )

  22. What happened to “the absolute sound” Harry Pearson was looking for and then Paul and Arnie joined an army of others on a snipe hunt?

    WARNING, If you don’t like reading me repeating myself or saying some of the same things in different ways then don’t read the rest of this.

    Because I’ve spent so much time studying the crux of this problem, 45 years to be precise, and because of the wonderful education that trained me to solve all kinds of problems, I’m in a position to give some insight into all of this. Consider it an informed opinion.

    The first of my conclusion’s I’m going to say is that to exactly duplicate sounds experienced one place from a recording in another will likely only ever be possible in a special laboratory. It will have to use anechoic recordings, measurements of the specific spot where the sound was first experienced and in my scheme reconstructed. This is not a suitable idea for an entertainment device, its only value is scientific research into sound, acoustics, hearing, and psychoacoustics.

    However the sound of live music has qualities that make it pleasing to may people who enjoy music that recorded music lacks. There are both audible and inaudible qualities and since pursuing the inaudible qualities is pointless I’m only going to focus on the audible ones.

    Is there such a thing as “the absolute sound” in a recording? Sadly no. Does it exist at all anywhere? As I pointed out recently again if you move yourself from one place to another during a concert you will hear something different. If the musicians move from one place to another you will also hear something different. Which one is the absolute sound? All of them? None of them? However, while chasing the elusive snipe it may be possible to create something with the same qualities although not scientifically identical equally pleasing and convincing rather than accurate.

    So then the problem is to study how musical instruments project sound into space, how that sound is affected by room acoustics, what arrives at your ears, how does your brain interpret it, and how do the physical attributes of sound correlate to the subjective qualities of sound. Simple? Well not zakly. So the final goal is to start with a recording you like and wind up with a sound field reaching your ears that has the same physical attributes that produce the desirable qualities most people like.

    So what are the major flaws in recordings and the sound they produce with current technology that distinguish them in your room at home from live sound at a concert venue. To put it in a nutshell, the recorded sound when reproduced is sharply truncated in time and space. To make the difference more interesting, as the sounds you hear in live performance sounds are experienced in time, they change altering the perceived tone. Until the digital compact disc they were truncated in loudness too but that problem at least has been solved. Necessary in some cases but not nearly sufficient by itself.

    So what do the sounds you hear live or recorded consist of. There is a first arriving sound that gives you a cue to the direction of the source and then there are a vast barrage of reflections of the sound you hear over and over and over again countless hundreds, possibly thousands of times. These are reflections and are most of what you hear live so studying them is important. This is an important major difference and the devil is in the details. This is why moving from one seat to another or moving the musicians at a live concert causes you to hear something different, the details of these echoes change. At home the echoes are entirely different. And if your speakers are like most the tone quality of the first arriving sound will change drastically too since most home speakers beam high frequencies over a very narrow angle that decreases with increasing frequency while musical instruments don’t. Not good IMO.

    Why do these reflections decrease and change from the live performance to the recording? Several reasons. With microphones close to the musicians the relative loudness of the first sound is proportionately much louder compared to the reflections you hear in the audience. They are also different in character. Your two channel sound system is not capable of capturing these reflections separately and then segregating them from the first sound at all. So whatever recorded reflections you hear will come from the same direction as the speakers reproducing the first arriving sound compressing what is normally heard from every direction in rapid fire succession to coming from two points. This is a truncation in space and time I was talking about.

    But what about multichannel systems, they reproduce different sounds arriving at you at home from different directions? Yes but for one thing they don’t capture the sounds that reach you from all those other directions and second they come from additional point sources while in the live music they are so diffuse you can’t hear any of them individually from any one direction. Each one comes from a different direction, is brief in time, and there are so many coming from so many directions that once you point a loudspeaker playing them at yourself the illusion is gone. BTW, this is where the quality of what acousticians call listener envelopment or what audiophiles call involvement comes from. So what is spread smoothly out directionally including vertically and front to rear in real music comes from a handful of point sources whose direction is easily identified. The system is multidirectional but directionally lumpy, intense from some directions where the speakers are, missing from most others.

    What about binaural recordings made in the audience? They hear what you’d hear. Well that is not exactly right and we’ve known why for a long time, at least about 60 years. When you turn your head the sound turns with it. The directional properties of both the reflections and first sound is gone. The sound is truncated in space to two points and appears to come from inside your head. Edgar Chouieri has figured out how to duplicate this for the first arriving source in one dimension over a limited change in angle. Not nearly good enough. No cigar. He does it by delaying both stereo signals and when you turn your head, a camera follows it increasing the delay to one earphone and decreasing it to the other earphone. If the delays are done one way, the source will appear in front of you. Reverse them the opposite way and it will appear behind you. If you rock your head front to back there will be no change. If you rock your head sideways from one shoulder to the other there will be no change. Only when you rotate your head and then over a limited angle will the effect work. He can’t do both behind you and in front of you at the same time and is only able to produce the illusion through headphones in one dimension over a limited range of angles.

    So how does my solution work? The measurement technique is still proprietary and not necessary to know how to build a sound reproduction machine that meets these criteria from commercial recordings anyway. In fact it has never been built or tested. You start with your recording and you want to arrive at a sound field that reaches your ears comparable to what you’d hear live. But how? It is based on one simple assumption, that there is a fixed mathematical relationship between the sound you have in the recording and the sound you want to arrive at your ears. In essence each recording is different for the way it is recorded and the most appropriate sound field you want to hear. Choosing parameters at the current state of the art has no one right or wrong answer. So there is no absolute sound and a snipe hunt is a prank.

    From Wikipedia;
    “A snipe hunt is a type of practical joke or fool’s errand, in existence in North America as early as the 1840s, in which an unsuspecting newcomer is duped into trying to catch a non-existent animal called a snipe. While snipe are actual birds, a snipe hunt is a quest for an imaginary creature whose description varies.

    The target of the prank is led to an outdoor spot and given instructions for catching the snipe; these often include waiting in the dark and holding an empty bag or making noises to attract the creature. The others involved in the prank then leave the newcomer alone in the woods to discover the joke. As an American rite of passage, snipe hunting is often associated with summer camps and groups such as the Boy Scouts.”

    So a combination of signal processing circuits form what is in the current prototype a hybrid analog and real time digitial computer that generates what’s missing from what you have, channels it through 4 separate amplifiers that drive 16 small Minimus speakers arranged around the perimeter of the room that aim their sound at the walls and ceiling (never at any listener) to increase diffusion and when they arrive at your ears have those same relationship and qualities as the part of the sound that is missing by reconstructing the mathematical relationship between those sounds and the first arriving sound as you would hear live. Here’s the kicker. I’ve only built two prototypes and through careful shopping including buying used equipment neither of them cost over $3000. This second one is better than the first one in every way, is easier to operate, and remains true to the concept. There is no commercial interest in it and the likelihood that there ever will be is slim to none.

  23. I’ve also left live concerts with a very different impression of the sound experience with friends because we were seated in different places (close or further away, central or on the side, etc.) I therefore don’t even think there is a single reference for even live music.

  24. You asked the questions and you gave the answers too.The fact that we all hear differently is the explanation which seems the most logical. And why not ? All human beings are similar yet no two are the same so why not the ear-brain mechanisms ? In an audio show one gets to hear many systems in a short period of time so do many others and they all feel that each individual system sounds the same to all just like live performances do. How else would one explain the existence of tone deaf, tin ears and golden ears ? Different hearing mechanisms. Things get further complicated by having less than perfect reproducing systems. Regards.

  25. This is a puzzling discussion. Paul ends by saying that reference is defined by each individual. In one sense that’s just a way of saying that everyone hears differently, or in practical terms, to each their own. As kids used to say, Duh! Except to those who believe their ears are the only correct ones. From a philosophical viewpoint, Paul’s conclusion could be interpreted as pure subjectivism, egotism run amok. Developers do have big egos, trying to convince others of their preference in sound (and typically design). That said, I’d like to think that PS audio’s perspective is broader than that, trying to find some real world criteria that’s based on something beyond one individual’s taste.

    What I would have expected – or hoped for – as a lead in is a discussion of “reference” is what is meant by the term in the hi-fi business, or whether or not the concept is even valid. After all, Paul and colleagues aspire to develop equipment that is true to something.

  26. There are different views as to the actual meaning of the word reference. Some people feel the top-of-the-line DCS digital stack is the reference standard to which all other digital products should be compared and judged.

    Soon, Paul will have a new reference, the unnamed recording studio he is assembling with Gus Skinas, at which point he will no longer feel a need to design speakers by comparing one against the other. The attended performances and DSD recordings will become his own personal reference, or guiding light, or inspiration or whatever you wish to call it.

    If one wasn’t present at the venue recording the performance, you really have no clue how it should sound on a playback system other than what pleases the ear and once again we’ve crossed over the line between objectivity and subjectivity.

  27. Addressing the main premise: the reference is inside your head.

    Absolutely – or there is no reference. The reference is the music you heard as a child. From birth, when you take your first breath and your brain comes out of an oxygen starved coma, your brain tries to make sense out of sound. Repeated stimuli cause neurons to grow, wire and program to recognize those sounds. Infants raised with only the sound of other infants vocalizing incoherently never learn to hear aural objects or images. Their brains don’t grow and they have permanently stunted perception.

    As you grow in a more normal sonic environment, your hearing develops hierarchies that enable us to recognize voices, speech, sounds of impending danger and pleasure, etc. BUT, the underlying hierarchical layers of cognition eventually stop growing. Your discrimination of phonemes stabilizes during your teenage years as you grow vocabulary and elicit the subtleties of semantics. This makes it very difficult for humans to learn a second language as an adult. Only rare adepts and/or extraordinary effort can achieve native fluency starting after age 13.

    In the same way, professional musicians almost never start practicing after puberty. They can’t even make it into conservatory because they don’t have the proper neural development to hear music as a native language. If you grew up listening to bad audio in urban and suburban noise pollution, you will not be able to learn to hear REAL music, the absolute reference.

    This came out of an informal survey when I was on the basslist. I was in a vocal minority with consistently different preferences in speaker design, mostly related to temporal distortion and extreme treble and bass reproduction. Because it was a mail list, the email addresses were listed and I noticed that my cohort had a lot of under-developed country codes. I emailed them privately and determined that all of them grew up in a rural environment without audio, or had music practiced acoustically in their house as a child.

    To achieve full competency in music listening it takes 10,000 hours listening to acoustic music, and you have to start young. Otherwise, you never grow the brain cells to decode it fully.

Leave a Reply

© 2023 PS Audio, Inc.

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram