“You Are There” Absolute Sound: Can We Get There From Here? Part IV

[tmallin]

The first three parts of this discussion were written long ago, but are available on WBF at the following threads:

“You Are There” Absolute Sound: Can We Get There From Here? Part One

Over on REG's Audio Forum, Robert E. Greene has recently been making the case that the problems blocking progress toward realism in home music reproduction come down to just two things: (1) how commercial recordings are made, and (2) how speakers radiate energy into the room and how the room...

www.whatsbestforum.com

“You Are There” Absolute Sound: Can We Get There From Here? Part Two

[Continued from Part One. . . .] There are two other stereo miking techniques I've heard which, to my ears, result in a fairly accurate spatial perspective, one where playback yields a fairly convincing replication of the perspective the listener would have heard from the recording mike...

www.whatsbestforum.com

“You Are There” Absolute Sound: Can We Get There From Here? Part Two

[Continued from Part One. . . .] There are two other stereo miking techniques I've heard which, to my ears, result in a fairly accurate spatial perspective, one where playback yields a fairly convincing replication of the perspective the listener would have heard from the recording mike...

www.whatsbestforum.com

And then there is also this sidebar discussion:

"You Are There" Absolute Sound: Do We Even Want to Go There?

Awhile back I posted a three-part series on "You Are There" Absolute Sound: Can We Get There From Here? here, here, and here. Since I wrote that series there has been further abandonment of the sound of live unamplified music from a good seat in the concert hall where the music was recorded as...

www.whatsbestforum.com

I've recently revisited this topic and here are my further thoughts.

I'm currently inclined to believe that, like it or not, when all is said and done, one is left with the Audio Circle of Confusion. http://seanolive.blogspot.com/2009/10/audios-circle-of-confusion.html That discussion concludes:

"the key in breaking the circle of confusion lies in the hands of the professional audio industry where the art is created. A meaningful standard that defined the quality and calibration of the loudspeaker and room would improve the quality and consistency of recordings. The same standard could then be applied to the playback of the recording in the consumer’s home or automobile. Finally, consumers would be able to hear the music as the artist intended."

Since there are no real standards at the recording end, there can be no standards at the reproduction end. Even sonic judgments using recordings made of one's own violin compared with that same person playing right between the pair of speakers as evaluated by neutral listeners are suspect since you cannot know whether the speakers are really reproducing what the microphones "heard." Sure, you can keep adjusting the recording technique until the match in such a test is as good as possible, but what can such a test tell you about the reproduction of any other recording made with the same mikes but where the mikes are not so skillfully placed, or with different mikes or miking arrangements, or with a different instrument or a group of instruments spread over a stage?

That changing the microphone can make significant, even gross, differences is readily apparent from JGH's reading of his "Why Hi-Fi Experts Disagree" essay on Stereophile Test CD1, as picked up through various professional-grade microphones.

The BBC may have had something approaching a recording/playback standard when they were able to directly compare the sound of various types of live acoustic music in the recording hall with the live mike feed in the control room. But those times are gone and that method is no longer practical for any loudspeaker manufacturer I can think of.

As far as experimental methods are concerned, I do not impugn the BBC's methods of measurement and listening comparisons of live in hall vs live or recorded in the monitoring studio. But I do wonder whether the BBC tested basic hearing acuity as Toole did once he figured out that normal hearing was essential to making consistent judgments as to speaker sound quality.

Then of course we have the fact that Peter Walker designed that other great classic British speaker, the Quads, outside the BBC umbrella and without much actual listening to the design at all. What does that tell us about the necessity of carefully comparing the sound of the speaker being designed with the sound of real acoustic music in order to design a great sounding speaker?

 

[tmallin]

If these days there is a microphone whose neutrality can be agreed upon--take the Coles ribbon, as a possible example--and a set-up which can be agreed upon to produce natural sounding stereo results--Blumlein, for example--then there is a procedure which interested speaker manufacturers and others could adopt for designing or picking neutral loudspeakers without having access to the type of live recording situation available to the BBC more than 50 years ago.

Here's how I imagine such a set-up could work. Start with an acoustically well treated room. Set up the speakers in stereo with the listening position chosen so as to provide a subtended angle of at least 90 degrees between the speakers. Bring in a musician and have the musician play from a centered position between the two speakers, with each listener taking turns listening from the predetermined listening position. Then record the musician playing the same material using the Coles ribbon Blumlein array set up at ear height at the listening position. Compare from the listening position the live playing of the musician with the recording and design or search for speakers which produce the best match between live and recorded.

No, this won't assure that the speakers accurately capture a soundstage of multiple instruments/singers arrayed on a stage. But it seems somewhat practical and not too expensive and definitely could use live acoustic music as the absolute reference. The test could be performed with various types of musicians, from singers to string players, woodwind, brass, and percussion. Obviously the cost and time taken could still be fairly high, but perhaps could be reasonably practical if, for example, university music students are used for the players.

The main problem with this experimental paradigm is that the recording will capture some amount of the room sound where the recording/listening are performed. "Well treated" or not, every normal room has a sound of its own. The recording will be at a disadvantage since the playback will replay the acoustics of that same room and thus the recording will have the listening room's acoustics overlaid on the sound of the instrument, not once, by twice as listeners hear the recording, whereas the live instrument sound only is subjected to the room's acoustics once.

One way around this problem would be to record the instrument either outdoors or in an anechoic chamber. Under such circumstances, the only room sound involved will be that of the playback room since the recording captured no room sound. That will add considerable cost and complexity to the experiment, of course.

 

[tmallin]

For those of us who have small listening rooms or who like to listen near field (say, six feet or less from the drivers) an important speaker selection criteria is the inter-driver coherence of the speaker system in question. You don't want to be able to hear out the individual drivers from your listening position. You want a coherent overall presentation where the sound from all the drivers blends together and the placement of individual drivers cannot be heard.

Harbeths, even the M40 series, are absolutely superb in this respect. Even with ears 20 inches away from the drivers, as long as the speakers are aimed at your ears and you are listening on the correct vertical axis (just below the tweeter) the individual drivers cannot be perceived.

Most large speakers are not nearly this fine at inter-driver coherence. I'm not saying that a large speaker, like, for example, the Revel Salon is poor at it. I'm just saying that anyone who listens in a small room or from close up who is considering any large speaker where multiple drivers are strung out in a line or arranged horizontally across a wide baffle (e.g., the Magnepan 3 and 20 series) should audition carefully to determine whether the drivers cohere at your required close listening position.

 

[tmallin]

Yes, listening near-field is my preference in any size room as well. The further you can get most speakers and your listening position from the room boundaries, the better the spatial presentation tends to be.

The cardioid dispersion of speakers like the Gradient Revolution and 1.4, as well as the D&D 8c, makes boundary proximity not so problematic, but padding the walls is still helpful in my small room.

With the D&Ds it actually is an overall plus to locate the speakers close to the wall behind them for the better-in-all-ways bass this allows. That sort of placement just does not work well with most speakers, in my room, at least. Keeping the speakers close to the wall behind them means my listening position can be further from the wall behind me for the same distance to the drivers and subtended angle between the speakers, which is a plus in my small room.

 

[tmallin]

Another requirement for a speaker doing well in matching the original sound may well be time domain accuracy. Speakers of this nature are usually referred to a "phase coherent" or "time aligned."

The usual meaning of "phase coherent" is basically the same as time aligned, which means that the leading edge of transients at all frequencies arrive at the listening position at the same instant when measured on the design axis. You can determine whether this is true from examining the shape of impulse response or step response graphs of the speaker system. In recent decades Stereophile has regularly published step responses of all speakers it tests. Look for a right-triangle shaped step response such as that shown for the Quad 2805 in figure 7 at this page: https://www.stereophile.com/content/quad-reference-esl-2805-loudspeaker-measurements

The lack of a right-triangular shape of a step response indicates that a transient's sound from each driver is not arriving at the listening position at the same moment in time even when listening on the design axis. The step responses of many speakers indicate that the outputs of the various drivers are not even connected in the same absolute phase with each other, with some drivers moving in while others move out; this shows on the step response graphs as both downward and upward spikes in response. Even if there are not both upward and downward spikes in the step response, the step response of most passive multi-driver systems show a time dispersal of the output of the various drivers with separate upward spikes in response separated from each other along the time axis.

Passive speakers aimed at consumers, historically available in the American market, and which measure as reasonably phase coherent in such step response tests include the Quad electrostatic models from the 63 forward (the larger ones with extra bass panels a bit less so), Dunlavys, Vandersteens, Thiels, Meadowlarks, Spica TC-50, the various Sequerra Met 7 models (not measured by Stereophile, but clearly of a phase-coherent design), and perhaps a very few others I'm forgetting at the moment. Crossoverless passive designs have the potential for good step response, but this sometimes eludes the designer, as with the Audience ClairAudient models, see https://www.stereophile.com/content/audience-clairaudient-one-loudspeaker-measurements

For passive speakers with crossovers to be phase coherent (1) all crossovers must be first-order 6 dB per octave and (2) there must be some method of physically aligning the voice coils of the drivers in the vertical plane either via a sloped baffle (e.g., Thiel) or physical offset (e.g., Vandersteen). Without the first-order crossover, the physical offset will not do the trick. Thus, none of the Wilsons measure as phase coherent despite all the effort at physical alignment and adjustability of the drivers back and forth along mechanical tracks.

The Quads are a special case with the time alignment accomplished electrically through analog delay lines feeding the Quad's concentric circular segments so as to produce a time-coherent response at the listening position from all concentric segments.

 

[tmallin]

Folks would do well to actually spend the money to read the most recent version of Floyd Toole's book, Sound Reproduction. Some have belittled Toole's work on various grounds, such as not being real science in that it does not test for accuracy but mere listener preferences and ignoring how serious listeners actually listen to speakers in rooms. But Toole is good enough at what he did to be both widely cited in the academic literature on acoustical research and also to have made a decent living at it. Yes, he ended up working for a large manufacturer, Harmon. Some seem to think of that as somehow having sold his soul.

But just for starters, even a casual reading of Sound Reproduction will indicate that "preference testing" showed that people prefer what is demonstrably accurate in terms of frequency response and thus that preferences are an experimental stand-in for accuracy.

Second, those who participated in the tests came from many walks of life, including many who were involved in the sound reproduction industry and even some audio reviewers. Those who made their living from sound did no better, and sometimes worse, in terms of consistent judgment of sound quality than others. Most anyone, it was found, could, with a bit of training, become as discerning in terms of sound quality as anyone else.

Third, the only real outliers in terms of ability to make consistent sound quality judgments were those with demonstrable hearing losses and those were excluded from the tests, even Toole himself in later life.

Fourth, Toole found that listeners could judge preferences and therefore accuracy at least as well using recordings created totally in the studio which had no acoustic concert hall origin. Thus, the fact that some (most?) contemporary equipment reviewers use rock and other non-acoustic music in making their subjective evaluations of equipment quality should not be a factor in dismissing their judgments.

Fifth, listeners show a remarkable ability to separate the performance of the speaker from the effects of the room in which the speaker is playing. Thus, careful evaluation in a "neutral" room is not so necessary as might be believed.

Sixth, yes, the tests usually were done in mono, not stereo, listening to a single speaker. Many criticize this methodology since speakers obviously sound much different and, in most cases, much better, in stereo than in mono. The book explains the good reasons why testing was done with people listening in mono to single speakers: people pick out sonic differences more easily in mono. The book clearly states that in later years, at least, rigorous listening tests were made with one listener at a time to avoid the possibility of listener interaction coloring the judgment of any given listener and to avoid the problems of judgment from different parts of the room.

This mono testing is not at all unusual. Objective speaker measurements, most of them at least, are performed by most reviewers on individual speakers, not in stereo, and for good reasons. Stereo power handling and SPL capabilities are extrapolated from the performance of individual speakers. Frequency response, time domain behavior, and distortion are all measured for individual speakers whether anechoically or in room.

Are these measurements all invalid because they are not done in stereo? Of course not. And frequency response measurements from a centered listening position with mono signals are demonstrably invalid when taken in stereo, at least in the higher frequencies, because of interference/cancellation effects between the drivers of well-set-up stereo speakers where the drivers of both speakers are fairly exactly equidistant from the listening position.

Besides, no one argues that the irregular off-axis performance of favored BBC speakers is a design plus. All some say is that such is easy to fix by absorbing the sound bouncing off room surfaces with acoustic treatments, and by helping the situation by listening near field, angling the speakers toward the listener, and with the speakers as far as reasonably possible from the side walls. I think that off-axis radiation of speakers is less of a problem generally in larger rooms because the reflections are both later in time and less intense for any given reasonable listening position.

I've recently read three books about speaker design and performance: Toole, Newell & Holland, and Colloms. I don't recall any of them mentioning much about designing speakers with different directivity for stereo than for mono. If people have been talking about this for a long time, I'm sorry, but I must have missed it. I'll have to go back and look for such discussion again.

What has often been discussed in the literature is the problem with perceived frequency response of centered images in stereo and how to correct this. There is also the problem that for headphones and wide subtended angles with speakers, the highs seem more prominent from each phone/speaker since the sound is at less of an angle to the ear canal than when a sound emanates from straight ahead as it would with a centered mono speaker.

I'm not saying that Toole is an audio god. I'm just saying that people should read his book and draw their own conclusions. Surely he does not deserve to be dismissed as an authority out of hand.

 

[tmallin]

More on the monophonic testing paradigm Toole adopted:

The fact remains that speaker performance measurements are mostly all done in mono. That is the almost universal practice. It is thus entirely reasonable that subjective performance evaluation of speakers, especially subjective evaluation of frequency response smoothness/flatness, distortion, and dispersion, also should be done in mono. The burden of proof should not be on Toole to prove the reasonableness of his procedure. The burden definitely is on those who would dispute it. Toole adjusted his testing protocol since he noticed that mono listening produced a higher and more consistent ability to distinguish among speakers' sound quality. That is a truly scientific method--redesigning a test for higher discrimination on the variable under test. Likewise Toole eliminated the variable of hearing acuity because he noticed that those who did not test normal on standard hearing tests were not as discriminating or as consistent in their judgements of frequency response, distortion, etc. as other listeners.

I think Harry Pearson used to enforce hearing tests on TAS reviewers, but I haven't seen that discussed in TAS in a long time. There is no reason to believe the BBC tested for hearing acuity unless that is in print somewhere. It might well have been viewed as ungentlemanly or unprofessional to imply that an audio engineer might have a hearing problem. Toole discusses the occupational hazard that audio engineers, even if they know they have a hearing problem, claim to be able to compensate for it due to their years of experience. He notes that while this might be plausible in a familiar system, once an unknown variable, like a new speaker under test, is interjected, the hearing problem becomes an actual problem for making sonic value judgments and adjustments.

The stereo performance of a pair of speakers is basically a judgement about their imaging and staging. Stereo imaging/staging is primarily an illusion produced by the human ear/brain analysis of the objective sound field in the room. But knowing how speakers should image or stage in your listening room usually cannot be known without reference to detailed notes by the recording engineer who may have adjusted things "just so" to achieve the desired illusion to his ears in that room. Again the audio circle of confusion raises its head.

It is also known that the spatial aspects of music are strongly influenced by the speaker set up and the measured frequency response of the individual speakers, not only on axis but also off axis. Diffraction effects from cabinet shapes and driver placement on the baffle also plays a role in this. One literally usually does not know how a particular sound should image unless one is intimately familiar with the recording microphones, their placement at the recording session, EQ applied in recording/processing, etc. Toole points out these issues in his discussion of why he used mono tests. The imaging/staging performance of a pair of speakers is interesting/engaging/enjoyable/and even fascinating, but who knows what is "right"? Robert Greene has often referred to at least the finer points of imaging and staging as ephemeral, epiphenomena not really in the recording but in large measure manufactured by the ear/brain in response to the objectively measurable characteristics of particular speakers and room set ups.

I certainly agree that I only or most always listen to stereo speakers in stereo. I even find mono programming heard through two speakers at a centered point a bit annoying, at least for the first few seconds, until I settle into the presentation.

Stereo listening is obviously vastly more enjoyable than mono and I certainly agree much more realistic in terms of comparison with the real unamplified acoustic music. I can't say that I've tried hard enough to get surround systems optimized in my listening rooms and it's been a decade or more since I even tried. But my memory is that for music, as opposed to video, surround sound as I heard it from commercial recordings in my systems, was always a bit contrived sounding, but, again, that's not meant to be any sort of pronouncement.

But in my experience in my rooms, getting the stereo imaging and staging right is less a part of speaker quality than it is of speaker and listener positioning and acoustical treatment than it is of the speakers themselves. Adding deep bass adds to general spaciousness and maybe a bit to depth, but all the speakers I've owned over the past couple of decades or more "do stereo" to a T when set up properly in a well-treated room. This is especially so for recordings where one expects solid stereo, such as Blumlein or other minimally miked recording methods. No, the stereo presentation is not identical from one set of speakers to another, but "doing stereo" is easy once the room, speakers, and listener are set up appropriately.

What usually bugs me eventually about speakers and triggers a new purchase are non-stereo aspects, such as spurious mechanical noises, distortion at higher volumes, apparent strain at higher volumes, perceived frequency response problems not apparently easily amenable to EQ or room treatment, or lack of inter-driver coherence. I do object to low positioning of the stage/images or lack of image height (as in looking at the sound from a balcony), but that's often a problem caused by relative listening height vs speaker height or trying to listen too close to the speakers in my small room. I also object to sound which is very dull sounding outside the stereo sweet spot, but I realize that's an entirely personal thing which others may not find objectionable at all and may even regard as a plus since it indicates a lack of unwanted room reflections.

Pink noise is actually also useful in stereo since it is a good indication of stereo imaging focus. In phase should be tightly focused to a point in the center and out of phase should maximally have that "diffuse and directionless quality."

 

[tmallin]

I totally agree with those who argue that flatness or smoothness of frequency response is not the only criteria by which we judge whether a particular speaker sounds like real music. The fact that Maggies are one of the biggest sellers in the "high-end" market (considering how many are sold, relatively few appear on the used market) just goes to show that flatness/perceived accuracy of frequency response is not everything when it comes to hearing the "gestalt" of live music from a speaker system.

I hear the appeal of the huge, wide-open window on the sound as well. To me the disadvantages (e.g., lack of bass impact, lack of SPL capability in the mids and highs before the fuses blow, vertical venetian blind effect of the side-by-side drivers from a nearish field listening position, and the stick-out-like-a-sore-thumb quality of the ribbon tweeter) of many past models have been off-putting to me. Recent models have considerably reduced at least some of these problems, however.

 

[tmallin]

I also agree that while many speakers can indeed be made to sound closer to live sound through the judicious use of electronic equalization, proper use of some of the most capable software programs for this are probably beyond the technical abilities of most audiophiles. This is probably true even as to REW. That program t is simpler to use with the D&D 8c's because of the D&D's integration of its firmware with REW, but it still is not super simple and in no way automatic.

The DSPeaker equalizers are quite easy to use and sound good in terms of results and are quite flexible in terms of user adjustments.

I also agree that it is impossible to sufficiently flatten low frequency response with passive cylinders, pillows, etc. Note, though, that the antinodes (low frequency dips of a sharp nature) are not all that obnoxious or even audible. Getting rid of the low frequency peaks is the main subjective battle. Then you just increase the overall low frequency gain to make up for the sense of thin bass. This is the approach adopted by the DSPeaker products and available via REW and it works very well indeed subjectively even if a sharp low frequency dip or two remains. However, it is very important to have adequate fullness in the 100 to 300 Hz region, the area where "the usual floor dip" occurs.

 

[tmallin]

Other objections to Toole's work are:

1. It was commercially motivated. But much of applied science is or should be. If you can't learn from your investigations so as to apply what you learn to create products people actually like enough to want to spend considerable sums to buy, you are spinning your wheels. You should be doing basic research rather than engaged in applying science to technology.

2. Toole is accused of lowering his standards for what is audible in terms of frequency response deviations when judging speakers. While plus or minus 0.1 dB is demonstrably audible wideband, Toole lowered his standards to plus or minus 1.0 dB when specifying whether a loudspeaker should be judged as accurate. But practicality tempers what is observable at the limit. It is commercially impractical to mass produce loudspeakers with a tolerance from sample to sample of 0.1 dB broadband or even over narrow bands. Compared to the plus or minus 5 dB frequency deviations which passed for high fidelity in the 1960s and before, plus or minus 1 dB is ruler flat. The Quads loved by many certainly are not that flat.

3. Toole is often represented as not being willing to apply equalization to level the playing field when judging speakers. But in numerous spots in the book Toole recognizes the benefit of and need for equalization of programs to make things sound more accurate. He even approves of old-fashioned tone controls. This stems from the audio circle of confusion: one needs to be able to correct for recordings made via systems with vastly different response than what one is using at home.

4. I don't think Toole says Harmon's Revels are better in terms of response than other Harmon brands, but many graphs of various Harmon brands are displayed so you can judge for yourself. The spinorama catalog at https://pierreaubert.github.io/spinorama/ certainly acknowledges that some of the less expensive Harmon efforts score even better by that measure than the best Revels. But perhaps the Revels have other desirable characteristics, such as lower distortion, better looks, deeper bass, etc. Quad, Harbeth, Spendor, and Graham all have models which may not be as literally accurate as their smaller speakers (Quad 2812 or LS3/6-type speaker), but which sell for more and may be desired above the smaller speaker for various reasons.

5. Toole's book does acknowledge his decision to use flat untreated walls as the standard for his tests. Many audiophiles cannot decide whether they like absorbing or dispersive treatment better on walls and if so, how best to arrange it. Many hate the sound of absorbing foam. Many are not in the position to use much of either because of decor considerations.

I know that even with absorbing foam, at least in my small room, limiting dispersion to narrower patterns than BBC-type speakers does seems helpful to the overall realism of the presentation. I also have found out that the smoothness of the off-axis response indeed makes such reflections far less obnoxious. For those listeners who want a "wide open" presentation with little fuss paid to room treatment, smooth off axis response such as the D&D 8c provides is a true boon with relatively little price to pay in terms of imaging, staging, and any overly bright or gritty sound.

6. Granted, the ability to tune the response of any given speaker to maximally mimic live sound produced between the speakers with skillful use of EQ is not addressed by Toole. But relatively few users, even audiophiles, are skillful enough to do that by ear. If they want to do it, programs like REW will do as well or better than most people can manage by ear. Robert E. Greene's writings and investigations are very valuable in alerting users to this possibility. But to single out Toole as ignoring this is unfair. Other than Greene, no other reviewers I'm aware of talk about withholding judgment on speaker quality until you try equalizng it with an aftermarket device or program.

7. As to Quads: I've liked Quad speakers since I first heard them many decades ago, even though I've never owned a pair for fear of damaging them by overdriving them in the bass. Many astute listeners agree that Quads are special as well and have used them as references for many years. Even the authors of my favorite pro-audio book, Newell & Holland, while extensively extolling the numerous technical and sonic advantages of active dynamic speakers, admit that for their own private musical enjoyment listening they rely on Quads.

 

[Duke LeJeune]

Yes, the tests usually were done in mono, not stereo, listening to a single speaker. The book explains the good reasons why: people pick out sonic differences more easily in mono.

Imo Harman's study of the relationship between loudspeaker preference and objective measurements is superb for what it does, but there are things which it does not do, such as:

- The single speakers auditions were conducted with the speaker located along the centerline of a 22-foot wide room, such that sidewall reflections arrived after an unrealistically long time delay, and thus would also have been lower in level than if the reflection paths had been shorter. So the subjective effects of room interactions under normal conditions were not adequately represented and evaluated.

- The test conditions were unfavorable to some types of loudspeakers, such as the hybrid electrostatic Martin Logans which scored so poorly (I can elaborate if you'd like).

- While single speaker testing is more revealing than stereo testing regarding sound quality differences, it fails to adequately evaluate the spatial quality of loudspeakers. In a paper referred to by Toole in his book, Wolfgang Klippel finds that the contribution of spatial quality to listener preference is comparable to, if not greater than, the contribution of sound quality.

Don't get me wrong - I think Toole's work and Harman's studies have a huge positive impact on our understanding of what matters in loudspeaker design. But I do not think they have answered all the questions which might be worth asking.

 

[tmallin]

For those who want to use after-market equalization to alter/correct the sound of your speakers to better approach live sound, here is some information about OmniMic V2, a system I use in addition to REW. Unlike REW, OmniMic V2 is basically just a manual measuring system and must be used with a hardware or software equalizer to apply your desired corrections. However, it is much simpler to use than REW, I think.

OmniMic v2 was developed by the same who developed Liberty Instruments' Praxis. I had Liberty's Praxis before this, but for modern computers with USB inputs, you can use a USB mike like the one that comes with OmniMic. OmniMic's USB microphone makes for a much simpler measuring setup than Praxis since you don't need to mess with the computer's internal sound card or find a better external sound card. Omnimic also dispenses with the Audpod box used for Praxis.

REW seems to yield at least equally helpful results but is quite a bit more complicated to learn, I think, and OmniMic has the advantage of offering a continuous real-time display of measured response so you can easily tinker with filters while watching their effect on response in real time on your screen. (If anyone else has the D&D 8c speakers, however, you are far better off with REW since that program integrates superbly with the speakers' firmware.)

Note that OmniMic does not work with Mac OS. Use a Windows computer to take measurements. REW works on both Windows and Mac.

For ultimate convenience, convert the OmniMic test CD to a music file and access it through Roon. Put the chosen test tone into repeat via the Roon controls. If you just use Roon as your source you can use the Roon DSP filters to equalize your speaker response to taste.

A few other hints about OmniMic:

• the sine sweep test tone is said by the manufacturer to give more accurate results at high frequencies than the RTA pink noise
• use SPLs of at least 70 dB for best results; I prefer 80 dB
• the USB microphone will drive extension USB cables of at least an extra 15 feet longer than what comes with the microphone so so you can get your measuring computer pretty far from the audio system. In my case, the cabling will reach far enough so that the computer display is out of the room entirely and at my normal computer work station
• for measurements from the listening position, move the listening chair totally out of the way so the chair surfaces are not near the microphone; aim the microphone forward toward the speakers, not up toward the ceiling
• you should exit the room before making measurements and if you do serious listening with your listening room door shut, make sure it is shut for the measurements as well (same for windows, of course)
• use a boom stand to support the microphone and arrange things so that the boom is angled down and away from the microphone; for impulse response measurements, pad the mike stand boom and vertical support to avoid spurious early reflections from the stand
• for most home audio listening heights a short kick drum set stand with boom will allow the best arrangement of the microphone on the boom
• I find that the OnStage Stands Pro Heavy Duty Kick Drum Mic Stand provides excellent value for money for this application in terms of sturdiness, ease of adjustment, ease of locking into place, and stability.

 

[tmallin]

For such recording/playback experiments the Blumlein stereo pickup should be positioned at ear height at the listening position, with the recording made either outdoors or in an anechoic chamber. For playback of the recording, the speakers should be positioned in room to the right and left of where the soloist or small ensemble is so as to subtend an angle of at least 90 degrees at the in-room listening position. Listeners should be about as far from the instrument as the Blumlein stereo pickup was during recording.

If you are present at the original recording--the test situation being described here--and actively working to get as good a match as you can between the sound coming from the speakers and the sound from the live instrument, it seems to me that you are short-circuiting to the maximum extent possible the effects of the Audio Circle of Confusion. This is potentially the best circumstance you can hope for since, for individual instruments or small ensembles, the live and recorded sound could be occurring in the same room with the ability to compare back and forth with little time gap. This is even better than the BBC paradigm since there the recording was usually being made in a room separate from the playback.

You still need to eliminate the double venue effect by making the recording outdoors or in an anechoic chamber. And you still have the possibility that the microphone response or placement is misleading you about the speaker accuracy. But, with enough comparison work, you have the best ability to separate errors being caused by the mikes and their placement (or errors caused by room acoustics) from those caused by the speakers.

So, yes, speakers that do better than others in producing such a match can reasonably be said to be better than others to which they are compared in such tests. Many speaker manufacturers and others can perform such tests with soloists or small groups if they so choose. The real question is whether most manufacturers and other audio professionals would care enough to do such tests, and that may depend on what they think the consumer market wants in terms of speaker sound: do people prefer literal accuracy or something "enhanced" in some way.

Note that I am not arguing that recordings made in small rooms, outdoors, or in anechoic chambers should ever be commercially released. We usually want to hear the more commodious acoustics of a larger hall. I can tell when our local radio station does live or recorded broadcasts from its small studio recording venue, versus when they do them from larger venues. The larger venues are "wetter" sounding and more like what you would hear at a concert. The acoustics captured on a commercial recording should be those of a good concert hall with the small-room acoustics of your listening room suppressed as best you can via room treatment and speaker and listener positioning.

My point is that as a practical matter, contemporary speaker manufacturers or other audio pros might be able to use this recording/playback technique to pick or design speakers to better match real acoustic instrumental sound without the impracticalities of access and expense involved in recording in a larger venue and having day to day access to such a venue. For soloists and small ensembles, the small room is probably good enough to make such judgments. Of course, if the manufacturer or other audio company has the financial wherewithal of a Harmon and the desire, they might aim to use the BBC method.

 

[stehno]

Hi, Tmallin. I've read a number of your posts and I appreciate your being so fastidious in all you post. That said, I did not read your posts above but only skimmed for several reasons. 1) My attention span ain't what it used to be and 2 it's really a bit of the same ol' same ol'.

No offense I hope but let me explain.

First I would like to say everything matters when it comes to the quality of sound / performance of a playback presentation. Recording studios can always do better but so can we - far better.

That said, I say same ol' same ol' because I'm confident that high-end audio is a bit inudated with folklore with few if any ever willing to think outside the box regarding any of our playback system's potential shortcomings. Even your main theme above seems to have a primary focus on the recording side of the house.

I find that rather interesting because the more I focus on what I consider to by far our biggest shortcoming (the playback system's noise floor), the more I'm amazed at how well-enough engineered the vast majority of recordings I own really are. And I'm talking recordings from the 50's until the present day.

IMO and in sharp contrast with some of your statements, I'm quite certain the most serious problem was, is, and always will be within our playback systems and how fastidious we're willing to be. And not on other sectors - though there's always room for improvement in every sector.

For example. By far our worst enemy has been a playback system's much raised noise floor induced by electrical energy - both coming from the wall / street but also all along the signal path from source to speaker drivers. It's a sonical catastrophy and when insufficiently addressed instills a universal performance-limiting governor such that our components may only operate far closer to their base performance levels rather than their optimal. And by optimal I'm implying performance levels far exceeding what even their designers imagined.

A distant second example is the speaker placement / subwoofer tuning within a given room such that there becomes a superior acoustic coupling between a speaker / subwoofer and its associated room. There are many degrees to this acoustic coupling which always starts with the token or half-assed approach. But I personally find this to be an often times daunting and neverending task but the benefits are potentially massive. Though the benefits are still far outweighed by those when we're able to drastically lower our playback system's much raised noise floors.

A note about noise floors. IME/IMO, a noise floor is the accumulation of a given playback system's mostly inaudible distortions that generates an imaginary horizontal threshold on an imaginary vertical bar. 100% percent of the music info embedded in the recording is read and processed but by the time the music info reaches the speaker drivers, we're hearing far less than 100% of music info than what was read from the recording. It's a percentage thing and it has almost solely to do with the severity of the system's noise floor. The more drastically we're able to lower the system's noise floor the music info we realize was embedded in the recording and the more we realize that the recording studio mic's were capturing far more music info from the live performance than previously thought.

That's the playback system's noise floor. But the speakers' (and subwoofers') level of acoustically coupling to the room also induces a noise floor of their own. Move a speaker / sub here or there and bass notes will come and go along definition, depth, warmth, balance, etc. In fact, the more fine-tuned this acoustically coupling the more the playback presentation becomes more balanced with richness, depth, warmth, etc rather than shrill, overly bright, thin, etc.

There are other less severe but still important issues on the playback side but these are the two worst that will absolutely cripple every last playback system as they are both universal performance-limiting governors. But again, when sufficiently addressed it becomes more clear that the recording side of the house actually has much less work to do than we think.

You mentioned Floyd Toole in your posts above. He and a few others are also quite fastidious in their work. But I really pay little attention to him for the simple reason that he and everybody else to one great degree or another all face the same issues crippling our playback systems' performance. Namely the two items I mentioned above.

With regard to Floyd, et al I mean no disrespect but as fastidious as he and others were it really doesn''t matter much because in all of their tests and evaluations they too were dealing with the same/similar sonically catastrophic shortcomings that severely cripple every last playback system, measuring instruments, sound generators, etc. If that is true, then that implies that all of their findings would be equally limited, crippled, and ultimately catastrophic. If true (and it is), there can be little or no value to their findings. Perhaps not much different than an exotic sports car researcher / reviewer testing and evaluating and publishing their findings when the only gasoline available to everybody is 79 octane and all the problems associated with that.

Actually I should take a step back here because THE most catastrophic matter facing this industry IMO is that there seems to be so few focused on training themselves to develop some ability to discern / interpret what they hear. Sorry but that had to be said and it's rather important because we have plenty of enthusiasts out there claiming every playback system they hear sounds just like live music or that there's a direct correlation between cost and performance or if some silly measurement is met, it automatically is more musical. That's why our listening skills as a collective whole is the most important issue. This is quite a dilemma because until more of us take our listening skills more seriously, not much else really matters, does it?

So with that said and if there's any truth to what I've said, I would say, yes there is an audio circle of confusion big time but there need not be the case. But so long as we refuse to think differently or think outside the traditional status quo box, we will continue to propagate this circle of confusion as well as perhaps a dozen or so folklore that go along with it.

Anyway, I appreciate your efforts and your bringing up this important topic but I probably disagree with most everything in it for the reasons I stated above. And with that, I'll close by providing a couple of oldies in-room recordings that hopefully will substantiate my position a bit. Particularly that recording studios for the most part have been doing a sufficient-enough or better job since at least the 1950's. Even though there's always room for improvement there too.

BTW, speaking of folklore. Robert Harley of TAS said in the 2009 Mar/Apr issue, "I believe that something catastrophic occurs at the recording mic's diaphragm such that much of the music never makes it to the recording." Paraphrased. Harley then went on to describe an odd experiment that Ed Meitner performed using a guitar and amp. So the implication may be that Harley was parroting Meitner's conclusions. Regardless, it was folklore and I hope the two old in-room recordings below will substantiate my folklore claim because I'm unable to detect anything that might fall into perhaps anybody's "catastrophic" category.

I mention this because this is some of the crap we have to sift thru to get to any real truth in this industry. Moreover, the playback system below retails for roughly $25k - a mere fraction of some of today's so-called SOTA-level systems. And there are no bass traps or room acoustic treatments other than just a mid-sized carpeted room with minimal furishings and a couple of book cases in the back of the room. IOW, just more folklore that we need to sift through to get to the truth of what really matters. But IMO, it's all well within our scope as enthusiasts.

 

[tmallin]

"By far our worst enemy has been a playback system's much raised noise floor induced by electrical energy - both coming from the wall / street but also all along the signal path from source to speaker drivers. It's a sonical catastrophy and when insufficiently addressed instills a universal performance-limiting governor such that our components may only operate far closer to their base performance levels rather than their optimal. And by optimal I'm implying performance levels far exceeding what even their designers imagined."

So says stehno. Sorry, but I just can't agree with this quote from stehno. The fact that he then gives as examples two recordings from the 1950s or 60s made on analog tape should lay to rest the idea that a raised noise floor in the home playback system is the problem. The signal to noise ratio of analog tape in that era was at most 60 dB. If you think these recordings sound good today it is not because you have lowered your playback system's residual noise in some significant ways. The S/N of the recording will always be the limiting noise factor and with such recordings the S/N ratio of the actual recording is at least 30 dB less than what is theoretically possible even with Redbook CD 44.1/16 standard digital.

Then I compared stehno's YouTube recording of his system playing the Take Five (outtakes) recording by Dave Brubeck's group with the streamed Qobuz "Hi Res" version, both played at the same volume through my M-1 iMac (2021) computer's headphone jack as heard through my NAD Viso HP-50 headphones. This playback system has no significant electrical or mechanical noise reduction measures included. The differences between the two versions are clear and significant. I won't enumerate them other than to say that there is no aspect in which the recording of stehno's system sounds superior; it is clearly tonally colored and lacking in detail (tonal, transient, and "small sound" venue-induced) compared to the streamed version. Granted, this is comparing a recording to a recording of that recording as played back. This is not an apples-to-apples comparison, but stehno offered it as proof of his thesis, so I'm just telling what I hear and stating that I hear nothing to demonstrate that stehno's system has in any way produced sonic benefits through lowering mechanial or electrical noise at the playback end.

And then there is the real-life experience of people who go to actual concerts. The acoustical background noise as heard from any realistic audience position is higher than it is in a reasonably quiet room at home. Yet that clearly audible difference in noise does not cause the concert hall experience to sound any less "live." That background noise and the way it is presented to the ears in an enveloping fashion is in fact part of what makes it sound "live."

"Folklore" includes, in my opinion, the current fascination among some audiophiles with "noise" or "noise floor" which stehno defines as "the accumulation of a given playback system's mostly inaudible distortions." Well, if the system's components' measurable distortion is very low--such as with Benchmark electronics, for instance--and the electrical noise produced through the speakers is so low that I cannot hear any hiss/hum/whatever when I put my ear within an inch of the tweeter or any other speaker driver when no music is playing at any setting of the system volume control--then what "noise," exactly, are you talking about and how is it getting through to your ears when music plays?

This type of fast and loose talk about "noise" has become the reason behind vast expenditures by audiophiles who apparently swallow the talk from manufacturers who claim their products reduce all sorts of subtle electrical and mechanical "noise," lifting veils and bringing the audiophile closer to sonic nirvana. Most of this talk is pure marketing hype.

But I'll give stehno some help on his fixation on "noise": actually, speaker drivers have "self noise" such that the actual distortion of most dynamic speaker drivers is no better than in the 50 to 60 dB range, even above the bass frequencies. But the system stehno is demonstrating appears to have dynamic and ribbon drivers and certainly nothing he is doing in terms of system mechanical or electrical tweaks is altering the self-noise of his speaker drivers. If you want speaker drivers with truly low self-noise and distortion above the low bass, you need to look at electrostatic speakers such as the Quad 63 which are specified to produce only about 0.1% distortion at all SPLs up to fairly loud above the low bass, which is far better than the 0.5% or 1.0% most other types of drivers are limited to.

 

[tmallin]

As to Duke's comment in his post #11 above, I refer readers to my post #7 above. I certainly agree that most speakers sound better, more lifelike, in stereo than in mono. I, like most other listeners, greatly appreciate the contribution of spatial quality which stereo brings to the table and I don't dispute Klippel's finding that most listeners find the spatial aspects quite enjoyable and important to their overall quality rating.

Perhaps surprisingly, Toole found that bass level and smoothness was also quite important to listeners' judgment of quality; I believe he put a 30% figure on bass quality as a part of his listening panel's overall assessment of a speaker's quality. That may be unexpected, given the long-time fascination of many audiophiles with "mini monitor" sound with its limited bass. I just attribute that to the unwillingness of many audiophiles to do anything significant to smooth the sound of low bass in their rooms, such as using electronic equalization or adding enough disparately placed subwoofers to the room to acoustically smooth out the response, as is possible with Duke's four-sub Swarm system. For many audiophiles, rolling off the low bass is sonically more desirable than lumpy low bass.

Back to stereo: As I discuss in post #7, many speakers produce great stereo effects (imaging, staging, etc.) when the listener and speakers are properly positioned, and the room is acoustically treated in an effective manner. Sure, there are still subjective differences in the quality of the stereo produced by different speakers, but determining whether one given pair of speakers is "better" at this than another is frequently a difficult determination.

One may regard this as a product of the Audio Circle of Confusion. If we listen at home to a system using different speakers, in a differently sized, shaped and treated room, with speakers and listening positions not in accord with how the sound engineer had things arranged, we may not be hearing anything like what the sound engineer heard or intended when the recording was made. Hearing the intended "space" as heard by the engineer during the recording process could be one definition of "accurate" in terms of one's home stereo presentation.

Another definition of "accurate" space may attempt to correlate the sound of the space one hears at home with what one might have heard at the live event where the recording was made. This is probably unknowable for most recordings of acoustic music and is meaningless for recordings which are the product of studio mixer manipulation of recorded tracks.

Based on my decades of listening in stereo to different kinds of speakers, I think that what one hears at home will also depend to some extent on the dispersion characteristics of the home playback speakers, regardless of how you set up your speakers or treat your listening room. Tall "line source" speakers will tend to produce taller images than concentric drivers, for example. Dipoles and omnidirectional speakers will tend to sound more "open" than most box speakers. Which is the "correct" presentation is difficult to say. Listeners may just have subjective preferences and those preferences may not be consistent across many recordings.

One thing to watch out for, I think, are speakers which in stereo tend to overlay all program material with a consistent spatial aspect. The vertical size enhancement of line sources is one possible example. The consistent reverberant/spacious quality of dipoles and omnis is another possible example. I keep coming back to differentiating between space which is actually on the recording and space generated at the playback end via your small listening room's acoustics. A test track like the Clap Track on the Sheffield/XLO or Reference Recordings/XLO test disk is telling, I think. Ideally, if you set up is not adding "space" via reflections from your listening room's acoustics, this track should sound identical through your speakers and through closed headphones--a single clap transient with no reverb tail. Most set ups will not even come close to passing this test. Most speaker set ups as heard from the listening seat will add clear and significant reverb tails, even clear echo. That reverb is added to everything you hear via your speakers. You may find it pleasant, but I don't think the reproduction can be called an "accurate" representation of the recording.

 

[stehno]

Clearly the audio circle of confusion lives another day.

I described and pointed you to the most crippling (think catastrophic) element that universally instlls a severe performance-limiting governor on every last playback system. And what do you do? You go out to Youtube and play similar verisons back on your iMac? Of course you completely by-passed the most crippling element and claim what you hear on your iMac is clearly better than my own in-room recordings? You're right, it's not comparing apples-to-apples. It's comparing apples-to-watermelons.

What you did was listen to a counterfeit of a countereit of the original. What I provided was a counterfeit of a counterfeit of a counterfeit of a counterfeit and perhaps of yet another counterfeit of the original. And in so doing you deliberately by-passed the very subject matter I tried sharing with you i.e. the playback system and it's much raised noise floor. What could possibly have been your motive for this apples-to-watermelons comparison?

Presumably, if one wanted to be fastidious as I presumed you were, you would at least take similar recordings and played them on your playback system for your evaluation. And if one wanted to be a bit more fastidious and objective, you'd publish a couple of in-room recordings while those or similar recordings were played back on your system so there's a bit more apples-to-apples comparison and then share those in-room recordings here so we'd have a very close apples-to-apples comparison. And this is where you might get a grasp of what I'm talking about. Otherwise, are you not just whistling dixie?

Since you've not shared the links of what you supposedly heard from your iMac, I'll take an educated guess that what you heard there on your iMac was nowhere as distinct a difference as you made it out to be. But that's an educated guess since you're obviously operating entirely in the closet here with nothing to substantiate your findings - which we already know to be seriously skewed.

Seriously though, reagarding what you just did here. Wouldn't it roughly the equivalent of you giving me what you consider your favorite omlette recipe and I make the omlette following your recipe but I intentionally leave out the eggs, and report back to you and of course others that your omlette recipe sucks because I can't taste the egg?

Fast and loose? All I can muster is, AMAZING. I guess it's my bad as I just assumed you were somewhat serious about this stuff and I blame myself for thinking you were fastidious and your efforts might include at least some due diligence rather than this obviously token or half-assed effort.

Like I said, the audio circle of confusion lives another day. At least until some of us are able to take matters seriously and starting with some due diligence.

BTW, you also seemed pretty clueless regarding a playback system's noise floor and other things but obviously no reason to pursue those matters any further either.

Nothing to see here. Move along.

 

[tmallin]

I'm sorry if I misinterpreted the reason why stehno included the YouTube video recordings of the sound of his system playing a couple of music selections. I thought he meant them to demonstrate his assertion that his system has conquered or at least significantly ameliorated the noise floor problem in home systems he asserts is the major problem preventing audiophiles from achieving life-like sound at home. (I hope I am not wrong in assuming that these recordings are of stehno's own home system playing these recordings; he seems to say as much in his second message.)

His second message suggests that perhaps those recordings were only there to show that recording engineering problems (and thus the Audio Circle of Confusion) had been "solved" as early as the 1950s since he apparently thinks these two recordings are examples of fine engineering and lifelike sound.

I agree that proper stereo recordings were possible back in the 1950s. Simply miked recordings using either quasi-coincident microphone arrays like Blumlein or two or three mike widely spaced omnidirectional microphone recordings were done back then. The engineers back then (or in more contemporary reissues) merely had to correct the frequency response of the sound captured by the microphones to make the recordings sound quite realistic. Old microphones had known high frequency response peaks and those can easily be corrected with modern or even quite old equalizers. The simplicity of the recording techniques in stereo's early days fostered recordings of exceptional potential. The real problems came later with "more sophisticated" mixers and the forests of microphones pan-potted into a "stereo" mix and all sorts of post processing possibilities.

Still, to hear what the early engineers desired you to hear, you would need similar speakers, similar room, and similar room treatment, and similar stereo set up geometry. Much of that is unknown and unknowable. Thus the Audio Circle of Confusion was there even for the 1950s stereo recordings. For example, the monitoring speakers used back in the 1950s may well have hidden to a large extent the peaky high frequencies of the microphones used. Vintage Acoustic Research speakers often used for monitoring in the USA back then rolled off the top octave considerably compared to what modern speakers usually do.

I will let others judge whether the two videos stehno included are examples of fine audio engineering producing lifelike results in early recordings.

I will just comment a bit further on the Dave Brubeck Take Five Outtakes recording. This recording is available on both Tidal and Qobuz. In terms of stereo presentation, the Take Five composition is what I would call a "bloated mono" recording in the Outtakes version. That is to say, there is very limited left-to-right stereo separation. The classic Take Five album release has the opposite problem of "ping-pong" stereo separation. Which one prefers is a matter of taste, but I could see how one might prefer the Outtakes version both for recording technique and for how the tune is played at a faster tempo with a much more inventive and entirely unaccompanied drum solo part.

 

[tmallin]

There apparently are audiophiles for whom the phrase "Audio Circle of Confusion" are fighting words. Here I will attempt to explain why I think this is so.

I think that these audiophiles mostly include those who are primarily interested in classical music recordings, those made in a good concert hall or other good interior venue of live unamplified acoustic music. For this type of music, those audiophiles who at least occasionally attend live concerts of such music have a mental reference or sonic picture of what such music should sound like.

For these folks, the basic job of a two-channel home music system is to play stereo recordings in a way which maximally mimics at home the gestalt of such music heard live in the hall. For their recorded references, they tend to identify a body of recordings which they think are well made. These primarily include those made with very few microphones and little processing. Such recordings include those made with fine ribbon microphones deployed in a Blumlein configuration, M-S configuration, or some other quasi-coincident configuration. They might also include two or three microphone recordings using widely spaced omnidirectional recordings, such as those made by such companies as Everest, Mercury, Telarc, and early RCA Living Stereo.

Using such recordings as their references, they then pick speakers, room set ups, and room treatments so as to produce as realistic a sound from such speakers as possible, both in terms of tonality and spatial reproduction. They swap out components and arrange things in the room until they approach this goal as closely as they can manage within their budget, room considerations, patience, and ingenuity.

This group of audiophiles then assume that once these "properly made" recordings are reproduced with maximal realism, all other recordings will also be reproduced as well as possible. This assumption is based on the further assumption that with other recordings there really is know way to know what they are supposed to sound like, so why alter their system to chase realism from them? Either that, or they just discount the importance of reproducing other recordings more realistically because then they may well sacrifice the realism of their best, "properly made" recordings.

These audiophiles stress the importance of only subjectively evaluating speakers and other equipment in stereo since stereo is how they listen to their best recordings as welll as most other recordings. How speakers sound in mono (such as the test methods employed by Toole) is irrelevant to them. The Audio Circle of Confusion does not exist since they know the result they are aiming for--maximal stereophonic realism (i.e., maximal match to concert hall sound of unamplified classical music in a good concert venue) at home from what they regard as well-made stereo recordings. They just adjust things so as to approach that goal ever more closely.

Those who follow this method believe that this method involves real science as it strives to compare the sound of the real musical thing with the sound achieved at home, albeit in a primarily subjective sensory manner rather than primarily through measurements. What really matters, they say, is the resulting sound quality and that can be approached through experimental comparative listening tests at home versus in the concert hall.

I see nothing wrong with this paradigm. It is, in essence, that followed by the BBC researchers back in the 1960s and 70s where the sound of the real thing in the hall could be contemporaneously compared to the playback from the live microphone feed or the recording in the well-designed control room stereo set-up. This paradigm resulted in speakers and descendants of those speakers which have stood the test of time among audiophiles interested primarily in classical music reproduction. It is a relatively simple process which can be pursued by most audiophiles regardless of the amount of technical knowledge they have.

Many people who acquire and use such speakers live happily with them for many years. Another such speaker which often creates long-term critical audiophile satisfaction among this group of folks are the Quad electrostatic speakers even though Peter Walker designed those speakers based on a theory of how speakers should work rather than by comparative listening tests.

 

[tmallin]

Against this approach you have those who believe that the job of the home audio system should be to match, as closely as possible, the sound those who engineered the recordings intended you to hear. It is when this goal is adopted that the Audio Circle of Confusion concept rears its ugly head.

Everyone acknowledges that most recordings even of unamplified classical music are not made in the way described in my post above, with minimal miking. And as for other types of recorded music, I think it's safe to say that a vanishingly low percentage of recordings are made this way. On the whole, I think it's safe to say that less than 1% of all recordings made since from the 1970s forward have been made with minimal miking. Most modern recordings are not really "stereo" at all: they are pan potted multichannel creations of engineers in a studio, made with a forest of microphones placed relatively closer to musicians than any audience listener ever would be. And then there are the huge percentage of non-classical recordings which originate largely from direct box feeds, electronically generated sounds, and gobs of post-processing. Such recordings are truly creations of the recording/mixing/mastering studios for which there was no live acoustic event which could serve as an acoustic reference.

Back in 1992, I wrote a letter to The Absolute Sound which was published in Issue 77, the one with the rock canyon picture on the cover. While at that time i had not heard of the phrase "Audio Circle of Confusion," I was quite aware of the concept and described how audio Rosetta Stone recordings could be made to help short circuit the circle of confusion and create recordings which would allow listeners at home to more closely approximate what the engineers intended.

Now, the best way to eliminate the Audio Circle of Confusion with any recording is to yourself personally record/mix/master the recording with a particular set of equipment, room design, system set up and room treatment and then play back those recordings at home in a very similar room, with very similar equipment, system set up, and room treatment. While this is possible, it is entirely impractical for the vast body or recordings made by others.

Once the home listener and those in charge of recording/mixing/mastering a recording are different people, the Audio Circle of Confusion arises. I think one has to admit that even among audio professionals there seems to be a variety of subjective tastes and opinions as to how recordings should sound. If it were not so, then most commercially released recordings would have similar tonal balance and perspective on the music and most recording studios would use similar equipment, studio design, system set up and room treatment. That this is not so is evident from pictures of such studios, discussion among audio professionals on forums aimed at them, and from books such as Toole's. In fact, the tastes of audio professionals spans a wide gamut, probably as wide as that of home audio enthusiasts.

Then of course there is the equipment, room design, system set up and room treatment. Unless the home audio listener matches what was used in the studio where the recording was made/mixed/mastered, the home listener is unlikely to hear that recording the way the audio professionals who made it intended it to sound.

It is against this background that Toole and others have argued for standardization of the recording paradigm. The idea is that if that can be standardized, then a home listener could arrange the home system so as to mimic the studio arrangement as closely as possible given the constraints of the room and pocketbook.

This is a lofty goal and perhaps one much more difficult to achieve than the BBC-type method described in my prior post. For one thing, it is difficult for most home listeners to mimic the room design, acoustic treatment, and equipment of a professional audio studio from a cost standpoint.

In addition, adherents to the BBC-type approach argue that the research done so far in an attempt to discover the type of speakers, room set up, and room treatments which are maximally beneficial is highly flawed for a variety of reasons. Some of those reasons include monophonic (as opposed to stereo) judgement of speaker quality, a failure to use solely acoustic classical music recorded with minimal miking in making speaker quality judgements, and a failure to determine the effect of room treatments and equalization on speaker sound. They claim that speakers designed according to Toole's/Harmon's design principles which have resulted from this research clearly are not as realistic sounding as the BBC-heritage speakers on music recorded with proper minimal miking techniques.

On the other hand, those in favor of standardizing the recording end of things argue that since the vast majority of recordings are not made with minimal miking, the best way for most recordings to be heard as intended is by mimicking the studio set up as closely as possible. It matters not whether the speakers used in the studio are not the ideal for reproducing minimally miked concert hall recordings of unamplified classical music. The audio professionals making the recordings are staking their professional reputations and livelihood on the quality of product they produce and home listeners should trust that they adjusted things so that it sounded maximally realistic, or at least maximally "good"--the way they wanted it to sound--when heard in the studio and it can't get better than that at home for the vast majority of recordings.

Your choice . . . .