Audio Vanity

In the quest for absolute fidelity we find ourselves running into an impenetrable barrier. Between recording techniques and playback there is little possibility of perfectly recreating an acoustic event. Think about it. Trying to reproduce the sound of a full orchestra playing in a 2000 seat concert hall, or a rock concert in a 20,000 seat stadium, in a domestic room of perhaps 200 square feet, is hubristic arrogance. But, how close can we get to creating the illusion of fooling our ears into believing? What are the means of producing that illusion?

There are many theories about how to accomplish this. Some are more complex than others. The problem itself is extremely complex—to the point of invoking chaos theory. Every proposed method is to a degree a simplification of the task, and necessarily so. How far are we willing to go to solve the problem, and at what point do these efforts start slipping into a obsessive pursuit of diminishing returns, or more commonly, zero return?

It's obvious that stereo is a poor approximation. It needs to cover a lot of territory, the direct sound, hall reverberation, and the location/distance cues of sound coming from every direction. No matter how good the recording or playback systems are, squeezing all of that into two channels and putting all of it in front of the listener will never truly fool the ear. Instead of trying to do it from the view point of the source, what if it's done from the view point of the listener? Binaural recording captures sound from the subject's reference point. Theoretically this is the most exact way to recreate a live event because it records what a surrogate head hears at the point of the eardrum.

Assuming binaural is the best method, immediate questions arise. Few recordings are made using the binaural technique. It has not made any headway. It requires earphone or headphone playback. Typical in-room speaker playback still sounds good, but it can't deliver the binaural effect. Of course, begging the question : why do we have to have playback with speakers? Earphones don't take any floor space, can reproduce the full audible spectrum with full dynamic range and extremely low distortion far, far more easily than full size speakers can. They don't need huge amounts of power. They can do it all with a single small driver. Multiple headsets can accommodate any number of people with individually controlled volume and no compromise in seating position channel balance or room interaction concerns. There is no disturbing others in adjacent rooms or the neighbors. Sounds like a winner. If we want background music, which isn't a critical listening situation, then inexpensive speakers could do the job. Think about it. For home cinema or serious music listening, use earphones; for casual listening use speakers. It solves or simplifies a multitude of problems, but it requires binaural recordings. Ordinary recordings can't capture the "I am there" reality. And there is the biggest problem. The industry is entrenched in standard recording techniques for stereo and multichannel cinema. A major shift in direction would be required.

However, there's one more consideration. As accurate as binaural recordings are from the ear's point of view, there's something missing. A musical event isn't merely heard, it's also felt. Lower frequencies conduct through solid material. Even airborne low frequencies can be felt. Those vibrations go through our entire body. Loudspeakers can provide the airborne and structurally conducted vibrations. Although the ear's experience with binaural recording & playback may be near perfect, the rest of the body's experience is nil. So close, you can taste it; too bad the texture is missing.

If we insist on trying to recreate an acoustic event in a room, we'll have to put more effort into it. A 7–channel arrangement seems to be a likely candidate, 3 channels front, 2 side, and 2 rear. But this would be the absolute minimum necessary to even begin to simulate a concert setting. It would require 7 full range channels of discrete information, 7 full range speakers and all the necessary amplification, peripherals and wires to go along with them. We're getting into a serious amount of equipment, cost, and complication. Are we really that serious or is this pursuit purely vanity?

More on Binaural Recording & Binaural FAQs.

back to the Observations main page.

— Sidetrack —

Good Recommendations

On good recommendation, I recently picked up the box set of Bernstein's complete cycle of the Mahler symphonies. I already have the Bertini recordings— excellent recording quality, outstandingly well interpreted and performed. So, why did I buy another box set of Mahler symphonies? The review I read of the Bernstein set claimed that these are superlative interpretations of Mahler. It states, ". . . after Bernstein other conductors' Mahler sounds bland. . ." Being the skeptic, I didn't take the reviewer's word for it. But, being familiar with Bernstein's treatment of the Beethoven symphonies, which are way over the top and far too romantic for Beethoven, would on the other hand be perfectly appropriate for Mahler. Indeed it is. Bernstein's somewhat exaggerated tempi and generous use of rubato accentuates the idiosyncratic nature of his interpretation and heightens Mahler's über romanticism. Clearly, Bernstein had a vision for these symphonies and an insight into Mahler. His insight is passionately presented and his vision explicitly expressed. I can't say it's more exuberant than Bertini's, but that could be because there's something getting lost in translation.

On the downside, I am disappointed with the recordings. In direct opposition to the reviewer who said, ". . . almost on the level of best current practice," and "Fidelity is no longer the issue. . ." Well, I must take exception. They aren't really bad recordings, but I find myself distracted by their glaring weaknesses. Once again, I am reminded that analog recordings cannot, do not, and never could have the dynamic range, clarity, or overall fidelity of digital recordings. From the first bars of the first movement of the 1st symphony I was aware of gain riding. Normally I use a standard volume setting for orchestral music. The beginning of the 1st movement, which starts out very soft, sounds artificially loud on this recording at my standard setting. So, I turned it down by 9 dB to get it soft enough and to avoid getting blasted later on in the movement. When time came for the blast, I found the forte-fortissimo sections lacking in power. So, I turned the volume up. And down again. I ended up spending the entire 52 minutes of symphony #1 trying to find the "correct" volume setting. The next day I pulled out the Bertini recording of the 1st symphony. The opening passage averaged 10 dB softer. The bombastic introduction of the 4th movement showed peaks going more than 5 dB higher. This adds up to at least a 15 dB greater dynamic range in the digital EMI/Bertini recording. It would be highly doubtful if Bernstein didn't draw out every decibel of dynamic range possible from the New York Philharmonic. His brilliant vision has been compromised. Compression (gain riding) is getting in the way and suppressing the performance. I wish it had been an evil plot perpetrated by the recording engineer, but it wasn't. It was done because the limitations of analog tape recording in the 1960s, when these recordings were made, required it. Could it have been de-compressed in the new remastering? Maybe, but it wasn't. That is not the only distraction. These recordings are excessively bright. Minus 4 dB on the treble control is needed to bring them back in line to an acceptable balance. And that's not all. Except for the loudest passages, tape hiss is a continuous accompaniment. And there's more than an average amount of incidental noises in these recordings too. I'm no longer used to all the tape hiss, nor the limited dynamics. I hope on subsequent listenings I can ignore the flaws and concentrate more on these historic performances. At least the treble can be attenuated to normalize the equalization and simultaneously reduce some of the tape hiss. (It might be excusable if it were a budget label, but it's not. It's Sony Classics. I have experienced too many poorly made, subpar recordings from them.) Normally when I have a choice, I avoid analog recordings, especially on classical music which requires full use of all the signal to noise ratio and dynamic range available. The most demanding recording companies, adhering to the highest standards have been using digital technology since its inception. Not one has returned to analog. Thank you.

— 20.04.10 —

(|| ) Audio Home
© 2008-2016 Parallel Audio all rights reserved