TREETOP TECHNIQUES
You may recall that in my column in the January issue, I stated that digital recording is "the great leveler." I pointed out that among the generally used digital recorders, none has a significant performance advantage over the others. Since this is the case, the sound qualities of classical music recordings are therefore largely a matter of choice and application of microphones, choice of recording hall, and, of course, the skill of the recording engineer in integrating and controlling these factors to achieve a recording that mirrors the sound of the live performance.
In the old days, there were "sound labels," record companies that emphasized the high-fidelity sound of their recordings as well as their music values. We had Mercury, with Bob Fine's Olympian Series, Westminster, Vanguard, and my own Everest label.
Of course, some recordings with good-quality sound were issued by other labels, but the audiophiles of those days knew they could consistently get the best recorded sound from these specialized "sound labels." Gone now is the relentless striving for technical superiority in the recording medium itself. Digital recording, no matter what brand of recorder may be used, gives all engineers an equal ad vantage. In spite of this, we still have "sound labels" today. This certainly must include London/Decca, Telarc, Delos, Chandos, and some smaller specialty labels. EMI is also a frequent source of recordings with high-quality sound, and it must be emphasized that recordings with good sound are be coming more common, even from other major labels. Obviously, with the medium no longer a factor, the "new face" of recording technology resides in the skills of the recording engineer and, to a certain extent, in the philosophies and musical perceptions of the producer.
I think it would be instructive to examine some of the early techniques of classical music recording. Back when the great Arturo Toscanini was con ducting the NBC Symphony Orchestra, his performances were broadcast from the notorious Studio 8H in New York City's Rockefeller Center. RCA Victor also made recordings with Toscanini and the orchestra in the same studio, and 8H gained its notoriety because it was an acoustic "sponge." Granted, the recordings were monophonic, but the acoustics of 8H were dry as a bone and completely robbed the sound of any warmth or dimensionality, making it seem very compressed. Other recordings made during that era were pretty dry acoustically, but not to the extent of those from 8H.
Bob Fine changed all that in 1950, when he first used a Telefunken U-47 omnidirectional condenser micro phone in his famous "single-point" mono pickup of the Chicago Symphony. He reasoned quite correctly that when you are attending a concert--in Orchestra Hall in Chicago, for example--you hear not only the direct sound of the orchestra but all of the reflections and reverberation characteristics that constitute the hall's acoustic signature. Thus, the omnidirectional pattern of the U-47 was used, with the mike painstakingly adjusted to achieve a "focus" that would capture just the right ratio of direct-to-reflected sound.
Incidentally, Bob Fine also recorded pop music for Mercury, and to him goes the credit for first incorporating reverberant information into pop recordings. If I recall, he used it initially on vocalist Frankie Laine's recordings of "Rawhide" and "The Cry of the Wild Goose." The idea caught on quickly, but you must remember that there were no reverb plates then (like those made now by EMT) and certainly no digital delay or reverb. Consequently, all sorts of means were employed to achieve reverb, including Columbia's (CBS) using the stairwell of their building on Seventh Avenue in New York City! With the advent of the stereo disc in 1958, both Bob Fine and I began using the spaced-array stereo recording technique. We employed omnidirectional U-47 microphones at the left, center, and right, with the center mike discretely recorded on the center channel of the three-channel, half-inch Ampex. The spacing of the three mikes varied according to the size of the performing group and the acoustics of the hall. On the two-channel stereo disc, mixed from the three-channel tape, the center channel was heard as a phantom image, thus avoiding the hole in the middle which was a common fault of original X/Y two-channel stereo recordings.
Currently, the chief proponent of the three-mike spaced-array technique is Jack Renner, chief engineer of Telarc. Since this label was founded some 10 years ago, it has achieved a well-de served reputation for recordings with consistently high-quality sound. However, Renner is not dogmatic. He generally favors the use of Schoeps omni directional mikes in his spaced array, but he has used Sennheiser and Bruel & Kjaer mikes, again depending on the recording hall and the performers. In large-scale works such as oratorios, Renner is not above employing some discreetly used accent mikes. His main considerations are the reverberant characteristics and ambient qualities of the recording hall. Like most engineers, he is on a never-ending quest for good recording venues.
What is the rationale for using a spaced array of three omnidirectional mikes? After all, the British are always damning this technique as being "too phasey" and lacking in instrumental localization and stereo focus. There is a certain amount of truth in this, but there is also no denying that, from a subjective standpoint, many listeners like the feeling of stereo width, depth, and ambient spaciousness which this technique conveys. Another major plus is that the best omnidirectional mikes from the likes of Schoeps, Neumann, Sennheiser, and AKG-have the most extended low-frequency response.
Some B & K mikes respond as low as 2 Hz! Most reach below 20 Hz with ease.
In contrast, most cardioid mikes have a low-frequency response that rolls off rather steeply below 40 Hz. So to cap ture the essence of a big bass drum or 32-foot organ pipes, the omni mike is first choice.
Needless to say, one does not use a three-omni spaced array in a hall that has too short a reverberation period.
Conversely, it is equally difficult to use this type of array in a hall with an excessively long reverberation period. In 1950, when Bob Fine first recorded in Chicago's Orchestra Hall, it had a reverberation period of about 1.9 seconds. It made for a lush, wonderful sound, with great dynamics, ultra-solid bass, and brass you could taste. When I was music director of RCA Victor classical recordings in 1965, I found out that a well-meaning patroness had given the Chicago Symphony Society some $3 million for a "general restoration and rehabilitation of Orchestra Hall." As you may recall, after Mercury ended its contract to record the Chicago Symphony, RCA took over and made all those fine recordings with Chicago conductor Fritz Reiner. In those recordings, they also used the three-mike spaced-array technique.
When I heard about the nature of some of the remodeling being done in Orchestra Hall, I had considerable misgivings about how this would affect its acoustics. Imagine my horror when we sent young Peter Serkin to record Bartok's Piano Concerto No. 1, with Seiji Ozawa conducting, and the engineer reported that the famous Orchestra Hall now had a reverb period of 0.4 second! That's when the Chicago Symphony Orchestra began to record in nearby Medinah Temple.
Since the 78-rpm days of ffrr recordings, has been justly famed for the generally high-quality sound of their releases. It is safe to say that no other major label produces recordings with such consistently high sound quality.
Decca uses a complicated technique that generally affords good sound, but it allows for flexibility to cope with a wide variety of recording environments. Like most of the "sound labels”, one of the cornerstones of the Decca technique is in the use of halls with outstandingly good acoustic qualities. At a typical Decca recording session, you will see the famous Decca "Christmas tree" mike array, mounted approximately 12 to 15 feet above and a couple of feet behind the conductor.
This tree is an equilateral triangle of 1 meter per side. Facing forward toward the orchestra, at the apex, is a Neumann M-50 mike. At the base of the tree are M-50 mikes angled to the left and right of the orchestra. These Neumann mikes are of a rather unusual design. At low frequencies, the M-50 acts like an omnidirectional mike, with the typically extended bass response.
As the frequency rises to about 2 kHz, the M-50 starts to gradually assume cardioid-like directional characteristics. At 3 kHz, there is also a rise in output level. Such a mike provides good low-frequency response, and the quasi-cardioid pattern affords more high-frequency definition. The directional effects give more localization to orchestral sections. Flanking the tree at various distances, depending on the acoustics of the hall, are a left and right omnidirectional microphone. Stereo pairs of cardioid mikes are used sparingly as accent mikes-again in accordance with the hall acoustics. This Decca tree mike setup works wonder fully well in most halls, but it is easy to perceive the sonic differences be tween halls. For example, Decca re corded the Shostakovich Fifth Sym phony in the great Concertgebouw Hall in Amsterdam. Just recently, they made a great recording of the same work in Walthamstow Assembly Hall in London, the scene of so many splendid Decca recordings. Spaciousness, ambience, depth, directionality, and instrumental localization are equally well delineated in both recordings.
However, the Concertgebouw has a somewhat longer reverb period, and bass response is fuller; the Walthamstow sound, on the other hand, is wonderfully detailed and more refined.
Some will prefer the massive sonorities of the Concertgebouw, while others will revel in the sparkling definition of Walthamstow.
Chandos is another "sound label" which has been steadily gaining respect for their well-engineered recordings, using a somewhat modified version of the Decca tree: Here is the same 1-meter, forward-facing equilateral triangle but with Neumann M-83 straightforward cardioid mikes at the apex and each base. Flanking the tree, left and right, are omnidirectional mikes, usually Bruel & Kjaer. Stereo pairs of accent mikes complete the setup.
Chandos makes some spectacular, vital, and very musical recordings.
They have a tendency to live dangerously with hall reverb. When they are successful, the sound is splendidly musical, although some recordings seem a bit overly reverberant and be come a bit diffuse.
Finally, we come to the Delos recordings engineered by friend and colleague John Eargle. As you would expect from the man who wrote The Microphone Handbook, Eargle's approach to high-quality recording is well reasoned and uniquely successful be cause of his engineering and musical savvy. His approach is different from most other techniques in that he uses a quasi-coincident pair of Sanken cardioid mikes in a modification of the French ORTF technique. This calls for a separation of 17 cm between the mikes, with an included angle of 110°.
Eargle uses flanking, omnidirectional mikes, left and right of the coincident pair. To round out the setup, he uses cardioid stereo pairs as very discreet accent mikes. In some halls, he uses a so-called house mike-a cardioid never more than 35 feet from the stage-in order to preserve the Haas effect by not exceeding the 33- to 35-mS limit of ear/brain fusion. Eargle's recordings are models of the art-wide dynamics, subterranean bass, and great orchestral definition, depth, and ambience.
There are, of course, a few other recording setups-including the overt, heavy-handed, multi-mike, multi-track situations-which, in general, do a disservice to the music. Next time you listen to some of the "sound label" recordings, try to visualize the mike set ups and the hard work the engineers have put into bringing you such great sound.
(adapted from Audio magazine, Apr. 1989; Bert Whyte)
= = = =