Just How "Absolute" Is Recorded Sound?


John Atkinson at the 2012 Rocky Mountain Audio Fest

Since I gave this presentation at the Rocky Mountain Audio Fest in October 2012, based on one of the topics in my Richard C. Heyser Memorial Lecture and mentioned in my March 2013 "As We See It," I have repeated it at Music Matters evenings at Definitive Audio in Seattle and Listen-Up in Denver, and at audiophile society meetings in Minneapolis, California's Central Coast, and Connecticut. I will be repeating the presentation at a Music Matters event at Georgia retailer Audio Alternative, Wednesday April 24, at 6pm, at T.H.E. Show Newport Beach, at 12 noon, May 31, and at The Audiophile Society in Brooklyn, NY on June 22.

Share | |
COMMENTS
dalethorn's picture

If you had the singers and the various mics and gear standing up and ready to record, but you had a room with solid walls that you could nonetheless shrink and grow dynamically by pushing a button or lever, how much of the difference between the three recorded versions could you approximate with the variable room size, from any one given set of mics? I'm guessing if the mics are close enough to the performers, that closeness might mitigate most of the room reverberation etc.

remlab's picture

..Valin was a no show. Heh heh.

acuvox's picture

I was raised acoustically and have only heard a phantom center on a few occasions, with exceptional high end gear and tuned rooms - not the kind of acoustic treatments you can buy in a store and bring home, but expertly architected and adjusted rooms.  I have been working with conservatory trained musicians for the last year because they likewise do not hear the stereo illusion from pan pots and have never been confused between live and recorded nor between acoustic and amplified.  In fact, they view all audio as inferior including seven figure systems.

At a previous Heyser lecture by Manfred Schroeder I had an epiphany that the pinnae are directional phase encoders.  I asked Dr. Schroeder if this was valid and he replied "of course".  You can get a sense of this by use of a plug in one ear, a blindfold and an accomplice who walks around the room talking.  You can hear direction and depth and map room dimensions crudely WITH ONE EAR!  This information is completely ignored by the orthodox stereo which only accounts for inter-aural differences in recording.  

Natural spatial hearing determines the location of a source by triangulating all the specular reflections and comparing the waveforms through the directional phase encoding.  This also requires de-convolution of the source polar pattern with the acoustical signature of the room path from source to listener.  This perceptual facility has been ignored in audio because all audiological data since the 1930's has used subjects acclimated to the temporal and spatial distortions of loudspeakers which mangle this directional phase information.

Necessary criteria for "absolute sound" require not only that a separate speaker be located at the virtual source position of every instrument and voice in the recording, but also that the speaker match precisely the polar radiation pattern of the original instruments.  A compromise for fixed speaker locations in playback still requires the reproduction speakers have flat frequency and phase response at all angles both horizontal and vertical to emulate a portal into the original event.  I have been reading Stereophile sidebars for 15 years and not found ANY speakers that match this criterion - so I built some.

The ideal, as been stated in the literature, is to have a string quartet represented by a 'cello speaker, a viola speaker and two violin speakers.  My contribution is to construct a consort of speakers that mimic the acoustics of the violin family, replicating the temporality and spatiality as well as the timbre.  I have produced and recorded amplified concerts of solo violin, viola, cello and contrabass for audiences comprised of post-graduate string players with 99% reporting transparency.  They were completely unaware of the amplification.  It is unprecedented to fool people who play an acoustic instrument daily to this level with an audio artifice.  Therefore I am announcing the first occurrence of "Absolute Sound".

Athough my residency is offcially over, I am still producing occasional concerts in the Manahattan loft where I did my experiments including this Sunday at 3PM and 7:30PM, and next Wednesday 5/1 at 7PM: http://spectrumnyc.com/blog/spectrum-the-venue/

remlab's picture

..I owned Dunlavy Sc1's. They had astonishing imaging. To this day, the best I have ever heard. My children,10 and 7 years of age, with ears at tweeter level, could hear(Or, I should say perceive.) "absolutely" nothing in between the speakers, just what was coming out of the left and right. I could perceive absolutely nothing from the left or right, only in between with the same material. Still blows my mind.

acuvox's picture

The Grand Stereo Illusion is learned (hearing sound originating from a place where the speaker is NOT).  Likewise, seeing persons in photographs, movies and television is learned.  This technology is embedded in our culture so you have to find outsiders to avoid self-referential results.  Some children escape it for a while, particuilarly if the live in a quiet background environment and practice an acoustic instrument.

My reference "box" speakers are SC-1's and I also have a pair of "outside the box" Earthworks Sigma 6.2 which image somewhat better; but I only hear imaging on recordings made by near coincident pairs including baffled arrays.  I have heard good imaging in demonstrations by Michael Gerzon, James Johnston and Ralph Glasgal, but they were far beyond stereo into megabuck research prototype systems.

I have re-trained my ears to my acoustic childhood by listening to acoustic music more often than speakers for the last ten years.  This re-ordered the music budget priority as follows: (1) Concert tickets; (2) Musical instruments and lessons; (3) Recording gear including proprietary designs; (4) SACDs and select well recorded CDs;  (5) Education and research; (6) Playback gear; (7) Tools and materials for acoustic furnishings.

If you have children, PLEASE expose them to acoustic music and let them play off-grid.

dalethorn's picture

That seems like a great gain in accuracy, but how close is it to absolute? The listening room has to play a part in bouncing the sound around, maybe not as badly with the string quartet if you sit close, but how can you sit in some ideal spot for different types of recordings? I suppose you could program some futuristic music server to know where the speakers (and how many) go for each track, then the music server would send that info to mechanical devices that move the speakers around according to the info they get from the server? I'm getting dizzy already...

acuvox's picture

We hear acoustic sounds in rectangular rooms every day.  Placing a speaker in the room to reprepresent every player is the absolute "they are here" illusion.  Seat placement matters the same as if the performance was in the room, which is a key factor in the illusion.  A good seat for the performance is a good seat for the playback and the contra-positive is true as well - an inferior location to hear an acoustic performance sounds bad in precisely the same way on playback.

This construct is not practical for living rooms because you need to store a speaker for every type of instrument and as many players as can be represented.  It is possible for a chamber hall with limited ensemble size.  I am hoping that it encourages more people to go live acoustic music by preserving the intimacy of a private chamber in an economically feasible theater.

I vastly prefer the articulation and nuance of small ensembles.  During the Baroque era there was usually one player per part and historically accurate performance has less than twenty players in an orchestra.  I just attended the season of the resurrected New York City Opera where the orchestras number around 30 and the voices are more human scale, to my delight.

X
Enter your Stereophile.com username.
Enter the password that accompanies your username.
Loading