The Fallacy of Accuracy

I was in a strange mood last January when I posted this on Facebook: "Do speaker designers strive for accuracy, or for a 'sound' they think potential buyers want?" I doubted that any designer with two working ears would even attempt to design speakers that merely measured well—there must be at least some subjectivity in their process. I also assumed that few designers would go on record about where they stand on the accuracy question, so I was thrilled when Elac Americas' speaker designer, Andrew Jones, responded:

Accuracy in terms of closest approach to the original performance is not practical nor even possible. There is no way to capture an original performance for replay over two channels that can represent the "truth" of that performance. All we can get is a facsimile that is the producer's attempt to capture what he wants to convey to you. As for a studio recording, it is a total construct and has no "original." Therefore, as a speaker designer, my goal is to try and keep the speaker neutral so it is agnostic to the type of music and, to a degree, the replay level. The difficulty is knowing from a design point of view what a neutral speaker is. From a technical-measurement point of view, we mostly think we want a flat on-axis response measured anechoically, with smooth off-axis response, although even this concept has been contested. But how do we know this is the desired technical performance? By listening, and tying listening tests to measured performance.

But listening to what? Facsimiles of a recorded musical performance? This is somewhat of a circular argument, compounded of course by the sound of the equipment we have hooked up to listen with. So in reality, I design to my idea of what neutrality sounds like, on equipment I find also represents my idea of the sound of neutrality, with music that I enjoy and want to hear in the way I imagine it should. The exception is that I also use music recorded by engineers whom I know who can give me some kind of comments on their impression of what I have designed.

Holy crap. That was a blast of fresh air. But I soon discovered that Jones is not alone. In an e-mail, PSB Speakers' Paul Barton contributed this:

When I design a new product I apply the experience I've had measuring and then interpreting the measurements in terms of natural reproduction. Once the new design meets or exceeds the performance expectations by measurements, then the "listening step" fine-tunes the final speaker sound. There is no substitute for the "listening step"—maybe someday a computer can do the listening, but I'm not counting on this in my lifetime. Besides, I love to listen to music; it's a huge perk of my job.

The response from Jeff Joseph, of Joseph Audio, was short and sweet: "I design for magic."

Exactly. It's all about the sound of music. Getting there is a journey, and while measurements can be essential tools, the best designers let their ears be the final judges—as do the people who buy their speakers. Apparently, for a lot of speaker designers, subjectivity trumps objectivity.

I put the query to Zu Audio's Sean Casey. "There is no perfection and there is no accuracy in loudspeaker design; it's all relative," he said. "Listeners hear things differently, there are always physiological differences, there are always biases based on individual experiences. Simply put, there is no absolute sound."

Nelson Pass, one of my all-time favorite amplifier designers, has also designed speakers, and he was no less forthcoming: "Most of the good designers who have control over the product work toward the sound that they personally want and, if the curve is reasonably flat, call that neutral." He continued: "As fragmented as the market is, there will always be a segment of the customer base that agrees with that assessment, and if the segment is sizable and the marketing department does its job, then the speaker will probably be successful. One percent market share will do it."

I didn't poll every designer, but if the majority of high-end designers all strove for accuracy, their speakers would sound more and more alike, and the most accurate would all eventually sound exactly the same. With the speakers I've heard that are listed under "Class A (Full Range)" of Stereophile's "Recommended Components," that doesn't seem to be happening: each sounds different from the others. To believe that every Aerial, Bowers & Wilkins, Dynaudio, Focal, Harbeth, KEF, Magico, MBL, MartinLogan, Revel, TAD, Vandersteen, Wilson, and YG Acoustics speaker on that list is accurate—or even in the neighborhood of accuracy—is to believe that accuracy can take many forms.

My take: Park your faith in accuracy and measurements on the back burner, and go for speakers that make a big chunk of your collection of recordings—the good, the bad, and the ugly sounding—shine. Last summer, when I visited a top New York City recording studio and listened to analog master tapes of classic rock albums, the sound wasn't a lot of fun—the studio's carefully dialed-in, super-calibrated monitor speakers revealed every last bit of distortion on the tapes. Sure, if all you ever listen to are exemplary recordings, a pair of studio monitors might be the way to go. But otherwise, accuracy in the design of audiophile speakers is a very rare commodity, probably because most of us don't want it.

Where you land on the accuracy spectrum will, over the years, likely be tempered by the evolution of your listening room, system, and tastes in music. Some audiophiles want to believe that accuracy is an absolute—but it might be only what sounds good to you at a the moment.—Steve Guttenberg

COMMENTS
jayg's picture

There is reference number one, for music accuracy and it's call live (un-amplified) music. Whether people want this or not is another question.

Start with microphones and then we can add cables; some people will add a mixing console between the cables and the recoding device. The best we can do now is one microphone, one cable and a recorder. Playback over headphones.

Reference number two. Recording --> headphones. (Headphones eliminate the acoustics of the room.)

Then there is the listener; this is the weakest part of the chain because the listener's ear/brain system factors in variables such as their own listening experience to that "type" of music, their attention level, if they "like" the music, etc. Unfortunately for the engineer the greatest variable is the listener.

As the accuracy of transducers improves it does "narrow" the commercial possibilities. When you remove those elements that add to the "signal" that the transducers are attempting to reproduce, then differences between transducers will, diminish. They will tend to sound "alike" in that they add less and less distortion to the original recording. This is good for those who enjoy listening to recordings.

John
President, Planot, LLC

geickmei's picture

It's not about accuracy. We are not doing "accuracy," there is no fabled pair of microphones witnessing the live performance from the listener's viewpoint that is going to give us the live music if only all of the in between components were accurate enough.

In fact all of the components are already accurate enough to do anything we want with the recording and reproduction system. No, the problem is that we don't know what we want to do with the audio that has been recorded. Stereo theory has been wrong for the entire stereo era, to the extent that we don't know how to build loudspeakers or listening rooms or just what is important in attempting to do stereo.

Sounds like a wacky charge, right? Well, I've got proof. Hopefully you will be able to read all about it in a future issue. I have asked Mr. Atkinson if I could submit an article on this whole subject, expanding on Guttenberg's statements in the As We See It of this month.

Please stay tuned. This is possibly the most important topic in audio right now.

Gary Eickmeier

prerich45's picture

I'm very amped to hear what you have to say Gary!!!! One such anomaly that exist is our own hearing ranges....that are subject to change with age. That's one factor that can't be accurately calculated due to the fact that we hear differently. We can make estimates of what a person with reasonably health hearing should be able to hear - but we can't do it on an individual level.

I'm looking forward to that article Gary!!!!!!

geickmei's picture

Thanks prerich45 - I am too, and I hope Mr. Atkinson as well, but it is not about the human hearing system. In fact, we could do the whole recording and reproduction process without reference to the human hearing system at all. This lies at the root of this whole "accuracy" mislead. The system is not about streaming two beams of direct sound "accurately" to the two ears, it is about constructing sound fields in rooms. Unless and until we can shake off this concept of "accuracy" we will never arrive at the real goal of realism.

Gary Eickmeier

prerich45's picture

Even more excited!!!!

dalethorn's picture

We each hear live music the same, i.e. the natural sound as we hear it. If the system reproduces that, and we're good with our own hearing, then the same system should please us, despite our hearing differences.

prerich45's picture

There are certain types of music that are intended to be amplified. One of the number one areas of amplified music - that's often overlooked by audiophiles is African American Gospel Choir music - that has at its core - and amplified instrument - the Hammond B3 Organ. There is no acoustic equivalent to the Hammond B-3, that gritty sound from the tubes, the Leslie speaker (love to hear that switch flip).

I believe there's also a subjective "taste" paradigm that presents itself when looking a musical reproduction to include ones definition of music.

prerich45's picture

Your next to last paragraph represents the biggest paradigm to music reproduction. Great comment!

Stefan-air's picture

In my own view, i also try to leave room for the fact that the subjective nature of the interaction between machine and man is what counts, even though I might differ based on conclusive scientific data. But my scientific interest stretch wider than the machine elements of the value chain, ie recording, source, amp and transducer for example and include how we interact with this. A sounds system in someone's home serves more than one goal. It could be an escape from a stressful job or a sterile relationship with a life partner, it could be a status symbol for the well-heeled few.

I also have a problem with the notion of "accuracy in the reproduction chain" or "true to the source" from a perspective of logic and philosophy. What is the baseline; the performance, the primary recording after using a certain microphone, or after the recording engineer has done his bit, with a certain monitoring device, all with their own additive or subtractive elements , might even be a valve based mixer, who knows. And would I have preferred to sit in front in the concert hall to see the facial expression of the artists, or rather to get the full perspective from the back of the venue, assuming it was a life recording, just too many variables that are not specified to make any call on true to the original a very contentious concept. For me, the purpose of a music system is to bring the experience of music into our home environment, it is not the original but like the allegory of Plato's it is a reflection of the reality. From a personal point of view, if I accept it is a reflection of the original, I tend to relax and enjoy it for what it is and I don't try to fit a full orchestra in a 4*5 living room at concert levels. Stereophonics, imaging, depth of soundstage etc fuel my human imagination and if I am lucky the system disappears and I have access the intent of the performers or composers. If someone then prefers a leather and wood masterpiece from somewhere in Italy rather than an aluminium structure in their living room because it triggers an unconscious association with an archetype that the cradle of music is in Italy, I say so be it and that then also becomes forms part of my scope and scientific interest. For me, this hobby is a testbed or practise field to observe and think about many different disciplines of science and how they intersect.

I am happy to be in the corner of accepting that accuracy is a fallacy, and the same goes for objectivity. For me, once i admit to my subjective bias, I can deal with it, get to ways to minimize it, or at least be authentic in communicating it as my point of departure.

prerich45's picture

I'm glad you guys got a quote from him!!!! He told me the exact same thing at Axpona 2012 in Jacksonville!!! That same day, Bob Carver passed me in the hallway - I recognized him but I was too paralyzed to even speak to him (my first time at a show). He was sweating badly and look like he'd been working on a car....what had just happened? He was voicing his Amazing Line Source speakers for their public debut!!! This was done by ear.

I love this article and it validates something I've always believed to be true!!!

jayg's picture

* Different microphones have different sonic signatures.

* Good enough = Subjective differences

* Objectivity vrs Subjectivity

* Set a standard for a reference recording, by way of consensus of recording engineers, being, a setup with one mic, one cable, one digital recorder, and specifications for the “session setup and adjustment parameters.” Have an independent testing entity setup by a recording engineer professional organization. Manufacturers who have their speakers tested can proudly display a “statement” and logo to that effect on their products.

jayg's picture

Years ago I proposed, to one of the organizers of the Rocky Mountain Audio Fest, that "they" should produce a CD with standard tracks that attendees could use to compare speakers and associated equipment. It could be a binaural recording of a small acoustic group using the simplest setup and "recorded in one of the actual rooms being used for demos" at the Fest. The guy laughed and dismissed the idea because, "these guys would not like that; they would find some reason why this or that particular track or tracks were "unfair" to their product." (I am paraphrasing.) The consumer suffers on.

geickmei's picture

Jayg there are plenty of test recordings already available. A binaural recording would not be appropriate to play on stereo speakers, unless you are doing Ambiophonics with crosstalk elimination etc, but then that would not be stereo. Very different systems, different recording techniques and different playback techniques. Each manufacturer will have his own favorite demo recordings anyway.

Gary Eickmeier

jayg's picture

Gary, yes I know about binaural technology. I have had excellent playback of binaural recordings on a stereo pair of loudspeakers. Good results are very dependant on the playback transducers.

jayg's picture

Here is an illuminating article on microphones (one of several classes of acoustic transducer) intended to be used for voice. It was published in Physics Today and there are a number of other interesting articles on microphone technology and applications. After all every recording begins with a mic and ends with a speaker.

http://scitation.aip.org/content/aip/magazine/physicstoday/article/69/3/10.1063/PT.3.3116

TRCose's picture

As music enthusiasts, I believe that what we want out of our system is the ability to lose ourselves in the playback. This loss of self is an experience of connection with the artist's idea.

We seek this aesthetic experience, but experience is subjective. Because by nature we seek objective answers, we attempt to measure this experience and its source. Accuracy is a measure of how well a playback system expresses a recording.

In audio we often emphasize the accuracy of expression when what we actually seek is the aesthetic experience of the idea. An idea and its expression are not the same thing. The two often come hand-in-hand, of course, but we shouldn't mistake that for interchangeability.

ednazarko's picture

I love this line in Andrew Jones' response: "All we can get is a facsimile that is the producer's attempt to capture what he wants to convey to you." Because it's truth - you're not listening to the artist, or the performance, as much as you are the way the producer wants to portray that artist and performance.

This is true for almost all arts that capture one moment for playback in another. Even photography - many of the greats say that a photograph is as much about the photographer as the subject. Two different photographers using the same camera, lens, and film will produce very different images of the same subject. My guess is the same producer would come up with different portrayals on different days.

When subjectivity begins at audio capture (which microphones to use), the notion of measurable accuracy delivering the best listener experience is curious. Kind of like "acoustic concerts" in arenas, where the only thing you hear if not on stage comes from the PA system via the mixing board.

Measurement accuracy is kind of a hygiene factor - not decisive, unless it's really wrong. I do read the measurements section of speaker reviews here, but I can't remember very many that didn't conclude that the speakers measured really well - a bobble here or there, and EVERY speaker has a bobble here or there. Then I go listen, if I can. I can appreciate Vandersteens and B&W Diamonds, for example, but they leave me cold. Other speakers that MEASURE almost identically speak to my heart and not my brain. I've never been able to see why in the measurements. But it takes me about four bars of music I love to know that it's my kind of speaker.

rom661's picture

As a designer wannabe who's knows a lot of designers, part of the reaction to any speaker is how well you tolerate specific inaccuracies. The flip side of things you like are things you don't like. "Crisp" top end is highly desirable to some people but makes me think of listener fatigue. I despise what I term smearing in the high frequencies. Two pieces of brass (a high hat cymbal) rattling together should should like that, not a haze of high frequencies. If a speaker does everything you like but is not very dynamic is that inaccurate? In my view, at any kind of reasonable listening level, it is. I consider compression a distortion. All home speakers I'm aware of compress at some level. Down the rabbit hole.... To Jeff Joseph's comment, and I like Jeff's speakers, I love something that makes me look up, startled at how realistic, how right, how magical it sounded. As long as I'm not paying too high of a price elsewhere. Happy listening.

avanti1960's picture

If speakers were created to be accurate they would have a dead flat frequency response curve (or the curve of the sum of the components in the signal chain which is relatively flat).
Playback through "flat" speakers sounds awful to most people. Everyone has a "target" response curve that they prefer whether they realize it or not- example "house curve". The speaker designer tries to sell you his "target curve" based on his preferences or sales goals.
All else being equal, there does exist a frequency response curve that sounds best to most people.
I'll take mine with a gentle slope from left to right, the smoother the better, and no distortion please.

HammerSandwich's picture

Some of the author's conclusions confuse me.

Quote:

...if the majority of high-end designers all strove for accuracy, their speakers would sound more and more alike, and the most accurate would all eventually sound exactly the same. With the speakers I've heard that are listed under "Class A (Full Range)" of Stereophile's "Recommended Components," that doesn't seem to be happening: each sounds different from the others.

How long is "eventually," Steve? Is it possible that we're heading toward universal speaker performance, but that the journey's not nearly over?

Along those lines, how similar are today's Class-A speakers as compared to those on the list in 1996? 1976?

However, those questions skirt a more fundamental issue here: Recommended Components comes from Stereophile's reviewers, not speaker designers. The subjective reviewers try to answer something like "What's most musically satisfying?" Wouldn't the results be different if JA asked "What are the most accurate speakers, regardless of how much you enjoy using them for music?" And different again if the panel was made up of full-time speaker designers? Deriving these designers' priorities this way cannot work.

Quote:

Last summer, when I visited a top New York City recording studio and listened to analog master tapes of classic rock albums, the sound wasn't a lot of fun—the studio's carefully dialed-in, super-calibrated monitor speakers revealed every last bit of distortion on the tapes. Sure, if all you ever listen to are exemplary recordings, a pair of studio monitors might be the way to go. But otherwise, accuracy in the design of audiophile speakers is a very rare commodity, probably because most of us don't want it.

So we DO know what speaker accuracy is and also how to obtain it! Briefly ignore the fact that you didn't enjoy the audition, so I can emphasize that the calibrated monitors struck you as more accurate & revealing than typical audiophile speakers. If I've misread your point, please clarify. The conclusion seems to rely on that detail.

And I find that conclusion a stretch. Because the studio's sound differed from audiophile systems', audiophiles must not like accuracy. (Otherwise, they'd buy it, and more designers would aim for it.) However, most consumers haven't experienced calibrated monitors, and expecting them to buy something they're unaware of strikes me as a little unreasonable.

Andrey's picture

I believe that audio accuracy is absolute. Because it's measurable. That's why I listen to studio monitors, though before I had "audiophile" speakers from Dynaudio. I hate those audiophiles who prefer beautiful lie to the truth. Because of them market is overloaded with "audiophile" loudspeakers, which have designed sound (in simple words - deliberately distorted). And it's more and more difficult to find the speakers, which reproduce original recording as close as possible.
The lack of an unambiguous criterion for sound quality helps marketers extricate huge amounts of money from the audio market, which they would never have received if such a criterion were widely known. Many people turn out to be deceived, thinking that they get a sound as close to the original as possible for a lot of money, but in fact they get “sugary shit” wrapped in natural wood veneer and with gold-plated terminals, and so on.

geickmei's picture

Stop! Everyone pull over and show me your license and registration! We have already established that it is not about "accuracy." I realize how difficult this idea is to shake off, but shake we must. Stereo reproduction is an acoustic process - a reconstruction of the sound fields that are in the recording within your listening room. The factors are the speakers' radiation pattern, their positions in the room, and the acoustic qualities of the room. Those factors are not directly measurable, and my story is that we are not doing most of that right, which is because stereo theory itself has been screwed up since the beginning. Again, it is NOT about streaming the sound from two speakers to your ears (the "accuracy" notion), it is about how the speakers put the sound into the room. Live sound has a small amount of direct sound, followed by a huge early reflected field, and finally the full reverberant field. Hi Fi changes that pattern to two streams of direct sound from just those two points in space, changing that pattern drastically, all because of this longstanding "accuracy" mislead. My article is about where this all went wrong and how to do it right.

Gary Eickmeier

John Atkinson's picture
geickmei wrote:
Hi Fi changes that pattern to two streams of direct sound from just those two points in space, changing that pattern drastically, all because of this longstanding "accuracy" mislead.

Longstanding yes, as the British Patent Application describing the theory behind stereo reproduction, written by the English engineer Alan Dower Blumlein, was filed in 1931. But misleading? I don't believe so. This is what Blumlein wrote:

"The fundamental object of the invention is to provide a sound recording, reproducing and/or transmission system whereby there is conveyed to the listener a realistic impression that the intelligence is being communicated to him over two acoustic paths in the same manner as he experiences in listening to everyday acoustic intercourse and this object embraces also the idea of conveying to the listener a true directional impression. . . . An observer in the room is listening with two ears, so that echoes reach him with the directional significance which he associates with the music performed in such a room. . . . When the music is reproduced through a single channel the echoes arrive from the same direction as the direct sound so that confusion results. It is a subsidiary object of this invention so to give directional significance to the sounds that when reproduced the echoes are perceived as such."

In other words, if you can record not only a sound but the direction in space it comes from, and can do so for every sound wave making up the soundstage, including all the reflected sound waves (the reverberation or "echoes"), then you will be able to reproduce a facsimile of the original soundstage, accurate in every detail.

See the relevant section from my Richard C. Heyser Memorial Lecture, which I presented at the 131st Audio Engineering Society Convention in 2011: Case Study 1: Recording.

John Atkinson
Editor, Stereophile

jayg's picture

You may have reached that conclusion but many have a different perspective and have reached a different conclusion. Yes, there are room acoustics. Yes, there is direct and reflected sound. So. These are givens when listening to speakers. Let's move along and solve the problems we can rather than saying nothing can be done. Sorry, I have work to do; we have work to do. It's not over untill the...

michaelavorgna's picture

...listeners, we'd have perfect hi-fi.

avanti1960's picture

whether we realize it or not. loudspeaker accuracy meaning freedom from distortion, freedom from enclosure vibration, accuracy in resolution, phase response and coherency.
what we don't want is accuracy in frequency response because it simply does not sound pleasing.

geickmei's picture

Sorry if I have hijacked this thread, but I am doing it to arouse interest in this most important subject. Jayg, I did not say that nothing can be done, I said that the answer lies in some surprising directions and is about the most important topic in audio right now. I said that I have written an article for Stereophile and asked Mr. Atkinson if I may submit it. It explains how the stereo theory contained in the Blumlein patent is at least incomplete, that many discoveries have been made in the 80 years since. An entire Audio Engineering Society exists to study the matter of the reproduction of auditory perspective and related issues. The Anthology of Stereophonic Techniques is one example of the number of ideas that have developed in the field.

JA, if my statement that we have established that accuracy is not the goal is all wrong, then you are denying the "As We See it" in your own magazine. In your lecture on the Heyser Memorial you say

"I tell this tale because it illustrates one of my points: that thinking you are right about something in audio doesn't mean you are right. No matter how much you think you know, there will always be new things that upset your world view."

In the Blumlein patent he says

"The fundamental object of the invention is to provide a sound recording, reproducing and/or transmission system whereby there is conveyed to the listener a realistic impression that the intelligence is being communicated to him over two acoustic paths in the same manner as he experiences in listening to everyday acoustic intercourse and this object embraces also the idea of conveying to the listener a true directional impression. . . . An observer in the room is listening with two ears, so that echoes reach him with the directional significance which he associates with the music performed in such a room."

This is obviously wrong. The information in live sound is not communicated to the listener over two paths. It comes from a multiplicity of incident angles, as described by me above. The spatial nature of live sound is well understood by architectural acousticians but not by speaker designers or room treaters. If Blumlein thought that the live sound field could be communicated to the two ears by two channels, and that this would encode all arriving sounds to the two signals, then he was confusing stereophonic with binaural. This binaural confusion lies at the heart of the accuracy confusion, but once the ears are free to hear all sources presented in front of them, with no crosstalk elimination techniques, then the system is not operating binaurally, and the theory that depends on just those two channels of direct sound from the speakers does not work - all recorded sounds, including the ambient reflected sounds that are so important to the enjoyment of a good hall, now will be compressed to arrive from just those two points in space or somewhere along a line between them. Siegfried Linkwitz describes this as a two dimensional articles strung on a clothesline kind of stereo.

My article is a path leading us out of this dilemma. It is some very different thinking about the whole concept of the reproduction of auditory perspective on loudspeakers in rooms. It explains the mysterious differences in sound among loudspeakers that measure the same on axis but sound radically different. It shows the speakers and room that I have designed for my own home to prove my theories from my 1989 paper "An Image Model Theory for Stereophonic Sound" that I could not demonstrate until now, and shows the design details of the speakers that have resulted from the theory.

I know that the article will probably spark a raging controversy in the magazine and has a good chance of upsetting your world view of audio, but if you are interested in these ideas, please permit me to send it in.

Gary

John Atkinson's picture
geickmei wrote:
Once the ears are free to hear all sources presented in front of them, with no crosstalk elimination techniques, then the system is not operating binaurally, and the theory that depends on just those two channels of direct sound from the speakers does not work - all recorded sounds, including the ambient reflected sounds that are so important to the enjoyment of a good hall, now will be compressed to arrive from just those two points in space or somewhere along a line between them. Siegfried Linkwitz describes this as a two dimensional articles strung on a clothesline kind of stereo.

Yet there is no doubt that listeners to a conventional stereo system are perfectly capable of perceiving depth in the soundstage, just as Alan Blumlein wrote. You appear to be confusing stereo playback over loudspeakers with the enveloping experience you get with a true binaural-to-loudspeaker auralization transform, such as that developed by Edgar Choueiri as Bacch-SP. See www.stereophile.com/content/bacch-sp-3d-sound-experience.

John Atkinson
Editor, Stereophile

Bart A's picture

This is a great article and good feedback. There is one hugely important aspect not mentioned. The room in which the speakers are measured is critical. An anechoic chamber is no substitute for a normal room, but what is the average room? When you go to concert, you go to hear the artists. Imagine you’ve paid a good price for great seats and then a few seats away some yo-yo is singing, off key and off time…

That is exactly what the average rooms does. Lingering bass, unbalanced time decays, modal peaks & troughs are all part of the in home listening experience without an intelligently designed room acoustic system. How in the world can one design a speaker that will sound same once it leaves the testing room and is placed in home? The speaker designer must make some allowances for this, be it with speaker placement recommendations and bottom end roll off, and radiation patterns to mention just a few considerations.

geickmei's picture

John - thanks for that Bacch article on loudspeaker binaural. I hadn't seen that one. But surely you understand that it is a head-related, binaural system and stereo is a field-type system. About every ten years or so some genius rediscovers binaural and proclaims it the answer to all of our problems with stereo. But it is not stereo, it is binaural.

I have no beef with binaural. I think loudspeaker binaural can be a lot of fun and very realistic, but it is a totally separate system from stereo, and my paper is about stereo. You saw in some of the comments on the Bacch system the incompatibility of the two systems, as in you should not play stereo recordings on a binaural system and vice versa. If you do that, the soundstage can wrap around you nearly 180 degrees, very unnatural and sometimes unnerving. But at least I hope that this adventure emphasizes the difference for you between the two systems.

In my humble opinion stereo can be the superior system, simply because it does not depend on your hearing system, the separation between your ears, the HRTF, pinnae shape, or anything else to do with a head-related system.

Simply stated, stereo is about making sound in rooms, not in your head. My theory says that the correct approach is to make music in the room with speakers the same way we make it with live instruments. There should be no difference in how the sound is distributed into the room simply because it is being made with speakers instead of people and instruments. With stereo, you can present to large audiences, you can move your head around, walk around the room, interact socially, and enjoy the music the same way you do with live music, if done right. No measuring of your ears or infrared beams tracking you around, no special recording techniques for binaural.

Bart A (I know who you are!) yes, the room is part of the answer, but the direction I will recommend in my article will be quite surprising to you and I hope to convince you with a demonstration some fine day. I want to have a mini-meet soon, before our remodeling does something drastic to my acoustic qualities.

John just let me send the damn article - please? Then you can judge and decide whether it is worth printing and sharing with all of these good folks.

Gary

David Harper's picture

If you had live performance of acoustic music in a room the size of your listening room at home, and two microphones were placed ten feet away from the performers, out in front of them, ten feet apart, and then in your listening room at home,your left speaker plays back what the left mike picked up and your right speaker plays back what the right mike picked up,would this be binaural?
Has anyone here actually done this?
What does it sound like?

spacehound's picture

See below

spacehound's picture

We can't do that and we probably never heard it anyway. And all this 'microphone' etc. stuff is nonsense. It is entirely out of our control. Try to 'compensate' for it (not that you have heard the live performance anyway), messes up all your other recordings.

The closest approach to the SOURCE. Equals CD, WAV file, whatever.

So I check the specs. Frequency response on ALL equipment (not including a PC, obviously - music does not come out of a PC). If it is flat over a wide enough range, then rise/decay time etc. (PRAT) is covered.

Then I go and listen to the FLAT boxes, no others. If one sounds greatly different from all the others, it's broken, whether I 'like' it on the recordings I hear is irrelevant, it is broken.

I pay for ACCURACY, not 'nice'. 'Nice' isn't HiFi at all. If I don't like the sound of a particular recording I just buy a different one. HiFi equals quality. Quality is accuracy. It's not rocket science, and myth and magic are not needed.

dalethorn's picture

I have not worked with equalizing in-room loudspeakers as yet, but I've done 120 headphones over a 3-year period, and I've discovered some things: Among those are, starting with a headphone that has a reasonable balance and isn't too choppy, by smoothing out the worst peaks and dips in the sound, the sound becomes more natural, and the soundstage greatly improves. The process isn't simple - it's iterative, but it works. I think it would work really well with loudspeakers too. Once you get there, or close enough, a wider variety of recordings should be pleasant to listen to.

tonykaz's picture

Eq. doesn't quite work with "Reflected" Axis ( room acoustics ). It works wonders for "Direct" Axis as in headphones.
Loudspeakers in Rooms present both Axis of radiation.

We can treat those Wall surfaces with non-reflective but we can't null those low frequencies easily. High power combined with Bass Wavelengths setting up double pressure zones and Bass null zones are the larger problem. Small listening rooms solve the worst of the Bass effects.

Rolling off the Bass, like a small Monitor Speaker with bass decreasing at 50 to 60 hz. is a desirable solution.

My miniature Office in Europe has a shoebox sized pair of Pro-Ac loudspeakers, it's the best sounding system I've owned in decades.

However, one friend has a pair of Large Cerwin Vega speakers in his home Bar Area. His booze helps with the headaches his sound system creates.

Tony in Michigan

dalethorn's picture

You missed the point. Equalizing as I've described it here and elsewhere is about the mids, and the low-mid treble mainly, although the upper treble may need treatment if it's off far enough. The point is not to use a 30-band equalizer etc. - a 10-band parametric equalizer is sufficient. You get the frequencies that are way off of even balance back into a relative balance, then you experiment with different recordings that you're extremely familiar with, to tweak that initial balance into the most natural sound. It can't fail, even if it isn't perfect, and that's because nearly any speaker-room combination has several opportunities through the mids and treble for some improvement.

AJ's picture

...purely on the imagination of the listener ;-).
*Perceptual* "accuracy" is a bit harder to define than some arbitrary measure.
JA ought to know this one:
http://www.aes.org/e-lib/browse.cfm?elib=9136
In the usual stereo audio presentation, a partial sound stage consisting primarily of the front elements of the sound stage is created by two channels, either sampled from several microphones set in the original sound field or more often by a mixdown of many microphones placed both in proximity to the performers and out in the hall to capture the ambience. The information presented by the two channels, in either case, is a small fraction of the information in the original sound field. Additionally, this fraction is presented to the front of the listener. The presentation does not create an envelopment experience, where one is immersed in the original sound field, as the information is not present. While some processors mimic the effect, such effects are not based on the actual venue but rather on some hypothetical model of a venue. : In holographic or auralized two-channel presentation, a presumed human head-related transfer function (HRTF) is used to create an impression of sound arising from other than the front of the listener. This works well in headphones or with interaural cancellation for one listener facing directly ahead and on the central axis between the loudspeakers. This method can, with some difficulty, produce an immersive effect for one point in the sound field, assuming that the subject maintains the proper head position, and the subject's head has an HRTF like that of the presumed functions. The ultimate form of this is, of course, binaural recording, where an actual head model is used to capture the information for one head location. : Beyond two-channel presentation, one can think of analytically capturing an original sound field to some degree of accuracy. This would require the use of many channels, perhaps placed in a sphere about the listener's head in the simplest form, requiring very high data rates (1000 to : 10 000 channels, perhaps) and creating a very high probability of influencing the sound field in the space with the microphones and the supporting mechanisms. As a result this technique is currently infeasible, and is likely to remain infeasible, for basic physical reasons as well as data-rate reasons, and actual analytic capture of the spatial aspects of a sound field in this fashion is unlikely.

AJ Soundfield Audio LLC

jayg's picture

In my opinion, binaural and two microphone stereo recordings are "more than good enough" for playback. The only need for a reproduction of the "whole 360 degree" (read that as unnatural) recording of the sound field would be when enhancing virtual reality and not for normal playback without visual information. I feel two channel is more than adequate, when done properly, for movies as well.

After all we are trying to closely simulate the two channel system we are born with.

The approach falls short, way short, in practice, because of a myriad of variables both in the production chain as well as the playback system.

dalethorn's picture

If we're talking about recording,
Some recordings sound great because the recordist and mastering engineers have either gotten really lucky, or they have better-than-average skills in making recordings. There are reputable physicists who believe that the 3D universe may be a hologram-like projection from a 2D surface somewhere. Guys like Richard Heyser got into these spatial things in the 1970s at AES. Whether the "really good" recordings are still poor facsimilies of a good concert experience or not, at least they're a good anchor point for further exploration.

If we're talking about playback,
Probably few audiophiles have the funds to get consultants into their house to spend whatever time is needed to install and tweak their system to have the best reproduction for the range of music they like. But the average audiophile, from reading Stereophile and other sources, can use common sense and a little specialized knowledge to set up the right system in the right room, and then tweak it from there.

In the end though, great sound is only going to come from great recordings, and we already know pretty much what those great recordings are.

jayg's picture

A general purpose microphone for stereo recording. It would have a polar pattern similar to human ears.
(A microphone's directionality or polar pattern indicates how sensitive it is to sounds arriving at different angles about its central axis. Wikipedia)

It would have an A/D converter built in. It would use wireless technology to send digital signals to a digital recording device. No "cables" just internal wiring. Analog to digital conversion right at the microphone capsule. Any processing, to correct for shortcomings of the microphone element, is done in the digital domain in the recording device.

How could this scenario be improved upon while maintaining only two microphones?

John

Russell Dawkins's picture

Are you suggesting, when you say 'polar pattern similar to human ears' the sound arriving at the ear drum, as modified by the influence of the pinnae and outer ear canals?

If so, this model could be improved upon by the elimination of the pinnae and canals, since the sound thus captured will be modified by those during the listening process, hence the Schoeps KFM6 Sphere:
http://www.bhphotovideo.com/c/product/743597-REG/Schoeps_KFM_6_KFM6_Stereo_Sphere_Microphone.html

tonykaz's picture

I once traveled with a successful Country Music Vocalist who explained that he explored a wide range of Microphones and Tube Mic amps to "get his sound". Since I own some of his recorded music I was surprised to hear his regular voice in everyday conversations. You'd never guess he was the same man and voice as in his wonderful recordings. We even sang songs together while on a plane for Europe ( I don't have a great voice, we were just having a bit of fun ).

Today, some of the Headphone Community buy little Garage1217 single Twin-Triode Tube amps and "Roll" various tubes to interesting results which they share on their Head-fi site. It makes interesting reading.

So, we music lovers and high-end Audiophiles enjoy building great "Sounding" music transducer systems. I doubt that most of us have any Sound Measuring Instrumentation equipment. Nor should we, that stuff is very expensive and requires careful interpretations to be effective.

We simply buy stuff that "sounds" good to us.

Besides, the Rooms we live in have a nearly "fixed" acoustic affect on everything we hear, our first investment should've been to build a "Proper" music listening room. ( mega-bucks )

I think Magnaplaners can sound wonderful in a Large Room, Pro-Ac Tablettes and the old Quads ( 1957 original version ) sound superb in a Tiny Room. In the mid 1980s I sold Theil CS3s for typical Room sizes.

Today, in a Great Room, using Powerful Mono Amps, a person can create a out-of-control listening environment that Sound Engineers would struggle to harness. I read Harmon reporting about careful Sub-Woffer placement issues and standing waves.

Well, if we're trying to re-create the Detroit Symphony Orchestra in our Family room, there's gonna be problems getting things to sound right!

I think that I'm hearing about Paul McGowan of PS Audio having trouble setting up his Infinity Reference System ( at his office, not his home ). Geez, if he's having trouble what hope is there for us mere mortals?

I'm not sure I care how "Accurate" a Loudspeaker is, if it doesn't sound wonderful in my space it's rubbish to me.

Tony in Michigan

ps. Sennheiser headphones always sound wonderful! ( for under $1,000 )

Andrey's picture

Accuracy is simply how accurate the speaker reproduces input signal. So, does somebody know what is the most accurate loudspeaker relative to the input signal? If I know that I have such a speaker, I know at last, that all the inaccuracy occurred outside of my "responsibility", i.e. before signal was recorded. With such speaker I have the most accurate sound (I don't take into account room acoustics and so on, just absolute lab measurements, in an anechotic chamber for instance). I'm not professional audio engineer, but I'm.. audiophile :-) So I'm quite deeply involved into nature of sound, but still, cannot find any materials where real speakers (or at last drivers) performance compares to the input signal. For instance, if we take file with 0.2 sec of pink noise, play it on a speaker recording it's output, and then, compare waveform of the original recording with the recorded sound. The speaker that produces the most alike waveform is the best speaker in the world, isn't that simple?

X