SIGNALS & NOISE (Letters to Editor) (Apr. 1990)

Are You Shure?

Dear Editor:

In the September 1989 "Spectrum," it was said that the original Shure V15 was the first elliptical diamond stylus made. Not so, if you'll forgive me for picking nits. There was an Edison phonograph with an elliptical diamond stylus long before magnetic, or even ceramic, cartridges were around. I read that somewhere-probably in Audio!

-Steve Graham; Studio Engineer; University of Michigan Public Radio Stations’ Ann Arbor, Mich.

No Mere Coincidence

Dear Editor:

Having read the correspondence on stereo miking technique in the July 1989 "Signals & Noise" column, may I throw in my penny's worth from this side of the Atlantic? The first most important fact we must accept is that all forms of mechanically reproduced sound are the result of a physiological and psychological confidence trick.

No way will we ever be able to simulate with 100% success the original source sound; all we can ever hope to do is get close to what we think is correct. I use the word "think" advisedly, because another factor enters into it. By definition, recording and reproducing sound is as much an art as a science.

So, anything goes-provided it produces results in conformity with our expectations of what we believe it ought to sound like. If anyone needs an illustration of this, it is an established fact that genuine high-quality sound reproduction and commercial hi-fi now have very little in common, and the gulf is getting wider. But that is another argument altogether.

It is what we expect subjectively that determines the recording technique used. Within these expectations, a disparity exists between the two sides of the Atlantic. I suggest that, in the case of the U.K., they are linked strongly to Blumlein's work and the broadcast practices of the BBC. Similarly, in the U.S., it is the early work at Bell and broadcast practice. The average balance engineer of any nationality is notoriously reactionary, and understandably so. With an accountant looking over one shoulder and the musicians' union rep holding a stopwatch and peering over the other, the engineer is far less likely to deviate from what he was brought up with, understands, and has hitherto used successfully.

It is, I suggest, inappropriate to talk of "purist" techniques because nowhere, to my knowledge, are recordings produced that use so-called pure coincident miking technique. I have one or two early (circa 1955) EMI tapes that probably were made this way. In the real world of commercial recording, it just does not work in many cases, and cannot be expected to. It also follows that, in the real world, to simply use coincident figure-of-eight mikes without signal processing is not appropriate either. I have tried it myself many times; the rear pickup produces confused images in a reverberant environment. If one insists upon trying coincident miking, the use of M-S gives better results.

As it so happens, I have been studying with fascination the fine print of the original Blumlein patent of 1933, which should be mandatory reading for anyone interested in the subject. The first discovery I made was that Blumlein devotes as much detail to a spaced omni technique as he does to the use of coincident ribbons. (We must remember that this type of directional microphone had only just been invented.) In fact, his original work established that above roughly 700 Hz, intensity differences were the determining factor in re-creating the stereo image; below this, phase. He also recognized that when reproducing signals over spaced loudspeakers in a real domestic situation, preservation of the phase information was impossible.

So, below 700 Hz, he converted phase to amplitude with his so-called "Shuffler," essentially a first-order low-pass filter in a difference channel. He retained this technique when using coincident ribbons with a cosine pickup pattern.

"How effective was it?" I hear you ask. Well, as it so happens, I also have copies of the test recordings Blumlein and his team made in 1933. Similarly, I have those made by Bell some three years later. These are quite remarkable, when one remembers they were made well over half a century ago. My (British) subjective expectations were fully met. The Bell recordings reveal all the inherent defects of spaced miking-a vague, amorphous image lacking in detail, and a massive hole in the middle. Those by Blumlein are perfectly natural-a sharp sense of direction and natural soundstage image, one that's impressively so when heard via headphones. The image is also much sharper and more natural when reproduced on speakers with a bi-directional characteristic (Quad ESL 63s). I haven't the slightest doubt that Steve Graham, author of one of the aforementioned letters in the July 1989 issue, might take the opposite view.

I know which I prefer, and if my fellow readers think this is an example of British fence-sitting, they are absolutely right.

-Reg Williamson; University of Keele Staffs, England

Stanley Lipshitz Replies:

I would like to comment on just one point made by Reg Williamson-namely, that in his original experiments (as described in his 1933 patent), Alan Blumlein used a "spaced omni technique." This does not accord with my reading of his patent (reprinted in the Journal of the Audio Engineering Society, Vol. 6, April 1958, pages 91 to 98). My understanding is that, at the time of the original experiments, he had only omnidirectional microphones available; in his patent, he describes how, by matrixing and "shuffling" the outputs of a closely spaced pair of omnis, he was able to create an approximation of the desired coincident pair of directional patterns he needed for intensity stereo recording. His capsules were spaced by about a human head's width, and his shuffler converted the phase differences created by the angle of incidence into intensity differences dependent upon this angle-that is, into the desired sideways facing figure-eight polar pattern, below about 700 Hz. Further matrixing with the sum signal produced an approximation to an angled pair of coincident directional microphones below 700 Hz. In his later patents (see references [5] to [7] of my paper, "Stereo Microphone Techniques--Are the Purists Wrong?" in the JAES, Vol. 34, September 1986), Blumlein specifically referred to the use of crossed figure-eight microphones (not available at the time of the original experiments) as the basic X-Y concept, and showed how conversion between X-Y and M-S was possible. The RCA ribbon microphone had, you see, become available in the meantime. His only reason for using spaced pressure microphones in the first place was to synthesize coincident directional microphone pairs! His original shuffled pair of omnis effectively behaved like an M-S pair.

- Stanley P. Lipshitz, Audio Research Group, University of Waterloo, Waterloo, Ont., Canada.

(Audio magazine, Apr. 1990)

= = = =