MCS Review On-Line Reprints

Still More About Recording Perspective

by Peter Fellgett

Reprinted from MCS Review, Vol.6, No.4, Spring 1985.

The BBC has a regular record review feature 'Building a Library' in which each week a distinguished critic selects his recommended version of some substantial composition.  Recently the chosen work was Bartók's Music for Strings, Percussion and Celeste.  After reviewing all the modern CD and other issues, the critic Stephen Walsh selected as his recommendation the twenty-year-old Solti recording.  Conclusions of this kind are far from unprecedented.  Older recordings are frequently recommended for the listener's mythical library, in some cases even, if you will forgive a four-letter word, mono.  Not long ago the choice even fell on an LP transfer from pre-war 78s.

Another recent record-review feature was the challenge of the Sunday Times to Norman Lebrecht to establish a basic collection of recorded classical music for less than 100 (about $129 at the time of writing, but anyone's guess by the time it is published with rates of exchange emulating the yo-yo and other proverbial exemplars).  His solution was to choose from among the 90-minute cassette re-issues intended for the fashionable 'personal stereo' portable players.  For reasons of cheapness these re-issues are mainly of older music where the recording costs have already been amortized, but with the added bonus that many of the recordings are classics in their own right, attracting such descriptions as 'definitive interpretation', 'unsurpassed performance' and 'immaculately engineered'.  Most date from 1955-'75, which Mr. Lebrecht describes as 'the heyday of analog stereo.'

The common theme linking these two items is that the older, simpler recordings were found superior in their ability to give the only things that ultimately matter, listening enjoyment and musical satisfaction.  What price then the vaunted advantages claimed for digital audio?  What about the claims for sophisticated creative use of dozens of recording tracks and scores of microphones?

Well, I was brought up in a fashion industry and early in life learned to speak of 'creativity' in the throw-away manner in which Goldfinger is supposed to have discussed the price of gold; to those within the trade, creativity is a bit of froth one throws in the face of the public while the real hard professional work goes on underneath.  Later in life I acquired a smattering of education and learned that 'sophisticated' early acquired a pejorative connotation, as for example when one speaks of a wine having been sophisticated (i.e., faked); as so often, it goes better in French where the description of an ideal meal may be rendered 'les viandes bien cuisinees main pas les vins.'

In his superb article ("More About Recording Perspective") in the Winter 1985 issue, William Sommerwerck quotes an excellent example of truly creative use of technology in the service of the composer; John Culshaw's treatment of the recording of Wagner's Complete Ring Cycle by Decca in the late 1950s.  He contrasts this with the all too common attitude that because something is technically possible it ought therefore to be done, especially if it can be represented as trendy, up-to-date, and if possible having some connection however tenuous with microprocessors.  He also goes to the heart of the matter by his reference to jamming a microphone down the throat of each section, instrument or voice, positioning the sounds with panpots, and covering the results with 'a glaze of synthetic ambience.'

Ambience Labeling

This is indeed an apt description, because a fundamental difference between natural and synthetic ambience is that the usual kind of 'reverb' only blends, whereas a natural acoustic not only blends but also labels each source of sound in a way that is characteristic of the place it occupies in the acoustic space.  This ambience labeling is an important aid to discriminating and to the discernment of inner lines in music.  When properly reproduced it also helps to maintain the stability of stereo images as the listener moves, whereas pan-potted images tend to 'disappear into the loudspeaker' unless the listener is centrally placed.

It is ambience labeling rather than shoot-the-pianist directivity which, for example, enables the listener to pick out individual voices among a large choir in a recording made with a single Soundfield Microphone in the Royal Albert Hall in London.  Another illustration of ambience labeling is our ability to localize sounds even when the classical directional cues have been removed or falsified by blocking one ear and covering the pinna of the other; in lecture demonstrations I have found that 70 to 80% of the audience, with eyes shut and their ears masked this way, can point correctly to a sudden sound made in the lecture room.

It may indeed not be beyond the capabilities of modern technology to simulate ambience labeling, although the equipment might cost as much as building a concert hall with a good acoustic.  Indeed Mr. Robert Parker, whose historic jazz reconstructions have attracted favorable attention in MCS Review, recognized this by stating his objective to be the reproduction of the sound quality of a live performance 'in which the band plays in a three dimensional space and the walls, floor and roof of the venue contribute to our enjoyment of the sound of the instruments.'  This would be a true 'glaze' in the same sense as a painter can use glaze to help convey his intentions, but the usual kind of 'echo return' does not do this and is more like the notorious dark 'gallery varnish' which Victorians used to daub over paintings to give them a subdued and 'antique' appearance.  Those who have had the privilege of experiencing the pristine freshness of paintings cleaned and restored by modern meticulous methods, at least as far back as van Eyck (d. 1441), will appreciate how misguided the Victorians were; perhaps this is a sound analogy.

A further consideration is implied by Mr. Sommerwerck's recognition that jazz, chamber and orchestral music belong to one category, and by implication pop and rock to a different one.  The difference is that in the first category the musicians listen and respond to each other.  By contrast, in most rock and pop productions they hardly do so, indeed may not even be present in the studio at the same time, or if they are all there together are prevented from hearing each other by screens intended to produce 'good separation' (whatever that means).  In what we may call listening-music, however, the performers respond not only to each other but also to the acoustic in which they are playing.

An elementary illustration of this is that if recording levels are set during rehearsal, it is usually found that they remain the same within a decibel or so during the actual performance, despite the greater absorption caused by the presence of the audience.  What matters however are the musical aspects, including not only adjustment of tempo and dynamic, but also many subtleties of phrasing and expression.  Thus if the acoustic experienced by the eventual listener is different from that heard by the players, the performance is to that extent falsified.

Of course the technical limitations of the earliest days of recording, particularly during the acoustic era, made it necessary (or at least so it was thought) to place the performers in unnatural acoustic conditions anyway, and well-designed artificial ambience may give a net gain; it is tinting the picture rather than color photography, but the same may be said of Ackermann prints.  For a modern recording, however, every effort should be made to reproduce the true and natural ambience of the real event if maximum musical authenticity is to be achieved.

Natural Balance

To take an extreme case, if everybody plays mf and is pushed up and down on the faders of the 'mixing desk', the result resembles instruments playing with natural dynamic about as much as the Metro Goldwin Mayer lion resembles Calvin Coolidge.  The very act of calling the recordist a 'balance engineer' tends to perpetuate this kind of approach, known to punsters in the trade as 'carry on mix-mastering'.  The reactions of the players themselves are interesting; some have become so used to such methods that they may plaintively ask, 'Where is my microphone?' when presented with naturalistic recording methods, but others are absolutely delighted to have the opportunity of being responsible for their own balance and dynamic.  After all, players and conductors had been engineering good musical balance long before recording came along, and it is they who are the true balance engineers.

Even those who have the taste to use a single stereo pair for 'main balance' are often tempted to add one or two 'spot mikes' if there is a soloist.  Done very discreetly, as it is by the best people, the result may be acceptable, but it has a number of dangers.  All too often an otherwise excellent recording is spoiled by the impression that the soloist is in a different room from the orchestra.  It can also produce a curious effect of double-imaging for off-center listeners, since the image from the main stereo pair is better stabilized by ambience labeling than is that from the spot.  Taken to extremes, quite grotesque effects can occur, such as the recording of a guitar concerto in which the entire orchestra appears to be performing inside the soloist's giant instrunent.

Another temptation is to use a high microphone position to obtain 'balance'.  Indeed this does help the players to sound equally loud, but this may not be what is musically intended; and, if it is, good players can achieve the effect anyway.  High microphone positions are at the expense of almost everything else.  They distort the perspective, and one knows of many otherwise good recordings which sound somehow wrong and disturbing in a way which is mitigated if you imagine yourself looking down on the performers.  Moreover, if one climbs a ladder and listens to an orchestra from above, it will be realized that the sound can be quite unpleasant, particularly that of strings.  For concert recordings, it may of course be necessary to compromise and raise the microphones above sight-lines, but in most other cases it is desirable to adopt a credible microphone position, i.e., one which the head of the listener might occupy either sitting or standing.

In conclusion, all reproduced sound is more simple than the original, however hard we try, as try we should, to make it not so.  The true aim of surround reproduction is, in my view, to reduce this loss of information.  It can indeed put the listener in the best seat in the house, certainly surrounded by the ambience of a fine concert hall, but in the middle of the performers only in music which calls for this, e.g., that of the School of St. Mark's or the music of either the ancient or modern Taverner.

Either approach, whether direct or only indirect sounds come from the rear, is true surround listening if it fulfills the just intentions of the composer.  Although the wide availability of broadcast and recorded music has greatly aided the development of public taste, it does have the danger that our ears may become unaccustomed to the true complexities and rewards of live sound.  In popular music in particular, what may be called the 'echo spring generation' has become accustomed only to simplistic, even if 'sophisticated', sounds with a corresponding blunting of musical perception.  Surround reproduction can help give us sounds which are a little less simplistic, but in the end there is no substitute for refreshing our ears with live unamplified concerts.

Reprinted from MCS Review, Vol.6, No.4, Spring 1985.
Copyright © 1985, 2000 by Laurence A. Clifton

Last updated: August 6, 2000

MCS Review On-Line Reprints