Chapter 7: Twenty Archival Sound Recordings:
Subjective Aesthetic Comparison

The twenty recordings of the GMdm have therefore been seen to vary widely in terms of tempos. Ultimately, in the appreciation of any form of the re-creative art of music, there is the matter of subjective satisfaction, a tenuous quality beyond the mere empirical evidence of stated facts and printed notes. The present author has attempted in this chapter to apply fair considerations to these performances and to arrive at some aesthetic conclusions regarding them.

One general observation which may be made about performances of the GMdm is an obvious one: the better the performers, the better the performance is likely to be. Put another way, the more professional and competent are the orchestra and choral ensembles, the greater likelihood there is that the forces will have the capability of doing justice to Berlioz' demands. A virtuoso conductor here is certainly not a liability, but it may be considered that a conductor more in tune with the totality of Berlioz' oeuvre has the greater ability to bring out the force of Berlioz, personality in this highly personal and individual work.

To begin this survey in roughly chronological order, the Jean Fournet (FO) recording stands alone in many ways. It is more striking an achievement when one considers it in light of the prevailing political situation; the recording date, asserted by Henry-Jacques, of September 1943 places it squarely in the infamous Vichy regime, surely an unusual time to record a work of such a strongly French character.1 Fournet at that time was the musical director of the Opéra-Comique, and it may be thought that any Frenchman who held such an appointment during the Occupation was on politically thin ground (the Dutch conductor Willem Mengelberg lost his post with the Concertgebouworkest and, in fact, his right to conduct in his native land as a result of his wartime collaboration with the Nazis).

Nonetheless, Fournet's reputation appears to have remained in good order; perhaps the concept of recording such a national work in an occupied country was seen by the French as a sign of defiance too subtle for their German occupiers to comprehend (as Berlioz' music continued to be more popular in Germany perhaps than in France). Certainly Fournet was to remain associated with Berlioz' large scale works, as witness his later recordings of other works, and his performance of a staged version of the GMdm at Lyons in 1969; yet it is strange that, though his pioneering recording of the GMdm has been cited as one of the finest interpretations of the work, particularly by early critics, Fournet has never rerecorded the work.

FO was recorded at St. Eustache, the site of the first successful performance, in 1827, of Berlioz' juvenile Messe Solennelle, of the premiere in 1850 of the Te Deum, and of several performances by Berlioz himself of the GMdm. The recording was taken down onto 78 RPM discs, and as such represents an amazing concentration of effort, that this large work should be recorded under such relatively primitive mechanical means. (The album examined in the course of this study is of course a transfer to LP of the same recording.) Haggin refers to the "strong reverberation from the empty spaces of the Church of St. Eustache--except for the Quaerens me, which seems to have been recorded in a studio, and sounds the better for that."2 No confirmation, nor indication of what this alternate locale might have been, has been found. Haggin, citing a review by Jerome Bohm (not found by the present author), and Reed both indicate a cut in the Quaerens me; however, this appears to be these critic's misreading of the fact that Berlioz excised a passage to the words "culpa rubet vultus meus" for the Brandus choral parts of 1852 and the second edition (Ricordi) of the full score of 1853, a cut observed by every edition since that time.3,4 Pace these critics, this movement is given exactly as left in its final form by Berlioz, a practice observed in all other recordings and probably in most modern performances, though occasionally one may find the "culpa rubet' passage accidentally left intact in the printed matter accompanying a recording or live performance.

The one cut which is taken in FO is of the Sanctus; after the first four measures of orchestral introduction, the performance proceeds from measure 92, the effect of which is that the first statements of the "Sanctus' and the "Hosanna" fugue (there is no setting in this work of the "Benedictus") are omitted, along with a brief half-cadential passage leading to the repeat. The repeat itself is virtually the same as the first statements, except that a brief choral passage is now added to separate the "Sanctus' tenor solo and the "Hosanna" fugue; in addition, pianissimo possibile beats on the bass drum and striking of the cymbals (indicated here to be struck as pairs) are added, along with some additional quiet lines in the low strings; then the upper-strings' "halo" from the restated "Sanctus" is continued to the end of the "Hosanna," which this time has a somewhat more extended conclusion.

Clearly the omission of the first statements in the Sanctus is regrettable from a musical point of view, as it softens the effect of the contrast between the two statements, and diminishes the majestic breadth of the movement, It is true however that given the capabilities of the recording medium, the difference in effect between the first-statement "Sanctus" and the second-statement "Sanctus" with quiet drum and cymbals is minimal in this case; also the restrictions of the wax disc in terms of recording time must be taken into account. FO was released on eleven discs, and it may have been that the inclusion of the complete movement would have necessitated the use of an additional side, with yet another left blank and wasted, a serious problem in material-poor wartime France; and the cost to the consumer of the extra disc ($1.25 in America in 1950) might also have argued for the cut, Still, Kolodin expresses the loss philosophically when he remarks that the cut "does not deprive us of any material by Berlioz, only omits a repetition of it."5

A subjective judgement of FO is of a carefully-prepared performance, though with a smallish and rough-sounding chorus, particularly the men (understandable considering the circumstances). The orchestra acquits itself well, though it is miked somewhat more distantly than the chorus. In general the conception itself is a fine one, and Fournet's tempos are as has been noted not too greatly at odds with Berlioz' score markings. If the orchestral playing is not blessed with the smoother qualities of other modern European and American orchestras, the balances reflect Berlioz' calculations of tone color. The chorus sings with fervor. The quality of the reproduction is naturally not as clear as most of the later recordings, but the contemporary critical appeal (at least when the album was issued in America in 1948) was favorable.

HO is tied to a pair of live performances near Easter Sunday, 18 April 1954, and was recorded shortly after the performances. This effort was the first GMdm conceived for issue on LP, and so of course offers the complete work without cuts. The orchestra and chorus--the latter known to be about 250 strong--were a semi-professional community ensemble, presumably specially augmented for the occasion, and Columbia Records secured adequate sound for the reproduction of a quite acceptable performance.

Hollenbach, quoted by his secretary in a personal letter to the author, has related the following:

The first performance that year filled the house. A second performance was given and also was a sell-out. The recording, done in Eastman Theatre, had the performers arranged in difficult positions. The orchestra and one "band" were on stage, another "band" was in the balcony and two more were one at each side of the hall. The chorus was behind me in the "orchestra" section of the seating. There were microphones everywhere.6

MI is a documentation of a live performance from the Salzburg Festival, taken down from a radio broadcast of 25 July 1956 from the Felsenreitschule. The performance was dedicated to the memory of the German conductor Wilhelm Furtwängler (who had died on 30 November 1954), and the recording as issued includes a spoken preface, presumably by Mitropoulos, acknowledging that dedication. The performance is warm and appropriately dramatic and devotional, though recorded in a very live acoustic with noticeable audience noises.

The advent of stereophonic recording and reproduction as a viable commercial medium in the late 1950s (experiments with stereo had been made as far back as the 1930s, with some transcriptions of performances by Stokowski and the Philadelphia Orchestra, and some spare studio sessions by Beecham and the London Philharmonic Orchestra) made the GMdm a fitting vehicle for demonstrating the new medium because of the ability to make clear the antiphonal brass writing. It is interesting that the first two recordings of the work in stereo were made during the same month, April 1958. It has not been possible to date the sessions precisely, but MA appears to have been tied to a performance on 16 April, and an unsigned Billboard article in the issue which went off-sale on 28 April refers to the sessions for SC having taken place the previous week, with no date specified.7 The two recordings thus cannot be separated chronologically, and since SC is reported to have taken five days of sessions, it is possible that they may have overlapped in time. The American release dates of the recording appear to have been identical--August 1958--and reviews in the principal American journals appeared simultaneously for the most part.

MH used the required orchestral complement, but with trumpets substituting for cornets and tubas for bass ophicleides in the outlying brass bands, which were arranged around the orchestra. The combined size of the two choruses is not known. The locale was Bushnell Auditorium in Hartford, Connecticut.

SC was recorded at La Chapelle de l'Hôtel des Invalides, site of the 1837 premiere, and while Smith claims that conditions approximated those of that performance "as closely as possible," this is doubtful, because the presence of the tomb of Napoleon (translated to the Invalides in 1840), and the partitioning of the main dome off from the adjoining Church of St. Louis, undoubtedly affects the reverberation space of the interior. Smith and Billboard concur that the chorus comprised only 130, and the latter source relates that "Scherchen used first desk men from every major group in the country, plus the entire Paris Opera ork [sic]"; Smith cites an instrumental complement of 170, where Billboard lists the brass bands separately.8,9

As might be imagined, MH and SC present a most contrasting pair, particularly by the nature of their tempos (see Chapter 5). Performance competence is greater in the more professional SC, though the choral forces are more clearly audible in MH. Unfortunately, the rapid tempos of MH tend to vitiate the power of the music, and the extraordinarily slow tempos of SC remind one of Rossini's remark about Wagner's quarter-hours. Weinstock, in his review expressing disappointment for both recordings, expressed hopes that a far more satisfactory recording might be made, specifying that the conductor "must either be Beecham or French."10

Curiously, the two GMdm recordings made the following year are by a French (or more precisely, Alsatian) conductor, and by Beecham. Munch's first of two recordings (he is the only conductor to have recorded the GMdm more than once) is the first to utilize a fully-professional standing orchestra, and in fact one of world rank. Furthermore, Munch had already communicated his long familiarity with the score to the Boston Symphony Orchestra on a number of occasions, as enumerated in Chapter 3. The orchestra plays with particular panache in the recording, and since Munch's approach is very sensible, the recording is quite successful. The choir, though youthful, has been well-trained and adds to the luster of the performance.

Lawrence protests that Munch uses trumpets instead of cornets, the critic preferring "the small-bore French brasses, with their bright, clear sonorities" to "bands [which are] heavy in tone."11 The precise disposition of the brass choirs is not known, but the separation works well in the recording, especially in the most recent issue, which has been remastered using half-speed cutting. The locale was Symphony Hall in Boston, with the audience seats removed.

While BE emanates from a performance at the Royal Albert Hall on 13 December 1959, it was not issued until about 1972, and then only privately, under the auspices of the Sir Thomas Beecham Society. As Beecham was a particular champion of Berlioz' music (he performed most of the major works, recording several of them and leaving performance tapes of several more), the performance is of considerable interest and merit, pointing up in particular Beecham's conception of the work as a progression of moods, rather than a loose set of stylized tableaux, The sound is somewhat constricted, coming from acetate discs transcribed from a radio broadcast, and in monaural sound. Orchestral playing is of first-rate quality but the listener must allow for the fact that this was a live performance, not a commercial recording, and retakes were not possible.

OR enjoys once again the luxury of a world-rank orchestra, and the vigor of a student choir (here comprising about 250 singers). The recording was made using the requisite cornets, but as usual substituting tubas for ophicleides. (It may be noted that Lawrence, who so commonly complained in his reviews of the lack of cornets in some of the earlier versions, was called upon to annotate this release.) Ormandy's lack of grasp of the Berlioz medium hurts this version, though the playing of the individual choirs of the orchestra is extraordinarily sumptuous. (The Hostias in particular is a fine example of the unusual nature of Berlioz' sonorities here.) The recording was made, following live concerts, on 1-2 April 1964, in the open spaces of the Athletic Club in Philadelphia.

MZ has, as noted, the one duplication (with MX) in conductors in these performances. Munch achieved somewhat weightier and more modern sound, as well as a more spacious acoustic, at the Benedictine Abbey at Ottobeuren, but the Bavarian Radio Symphony Orchestra does not have the elegance of the Boston ensemble. Chorus is effective, and this July 1967 taping preserves essentially the same performance as the earlier one, though with tempo differences as noted in Chapter 5.

1969 was the centenary year of Berlioz' death, and as such was a fitting occasion for performances of his works, especially the GMdm with its memorial associations. Four of the performances in this year (see Chapter 3) led to the issuance of recordings tied to those performances, though two of them were tapings of the performances themselves, and the other two were separate recording sessions tied to the performances.

The first of these, from 5 March 1969, is recording CH, and presents a special problem because of its nature, requiring some background information. It originated in the Soviet Union, where performance practice considerations are sometimes dependent on political situations. Conductors and other performers are generally considered to be able to achieve their complete training entirely within the Soviet Union (and its allied nations) and few are given the extensive experience which world touring, more readily available to Western artists, can allow. This leads to a de-cosmopolitanization of styles and performing "schools"; witness the oft-noted "Slavonic wobble" of female singers, the lighter tone of horn players, and the percussiveness of so many Soviet pianists; the incestuousness of their playing and teaching has emphasized these aspects of their specific performing profiles. Conductors, too, chiefly learn from one another, and while the present author does not claim that there has resulted any form of homogeneity in styles (witness, for example, the contrasting styles of Yevgeny Mravinsky and Gennady Rozhdestvensky), still a recognizable profile can be observed for the Soviet-trained conductor.

Performance practice, too, as it applies to editions and performing versions is subject to external influences as well in the Soviet Union. It might be supposed that the anti-religious attitude of the Soviet government would lead to a wholesale suppression of liturgical and other religious music, but this is not quite so; however, performances are far more likely to concentrate upon the spectacular and musical values of the compositions than on their more worshipful aspects. Antiphons by Arensky, Tchaikovsky, and even Rachmaninoff are common in a cappella performance in the Soviet Union, although one is most likely to find their works given "off the track," that is, outside of the musical centers of Moscow and Leningrad, especially in the outlying "independent state S' of ethnic groups outside of the R. S. F. S. R. which make up the non-Russian portions of the Soviet Union.

In this instance the "outsider's' nature of the performance clearly alienates it from the standard Berlioz practice. Chekijian had studied with Fournet, but this study does not seem to have transmitted any clear conception of the Berlioz style, rather as Fritz Mahler's tutelage under Weingartner did not seem to be of much help in the Hartford performance. CH is beyond doubt exciting, but the inaccuracies of the performance (in particular the way the tenors get separated from the orchestra at the outset of the Lacrimosa, and the harsh, wobbly sounds of the brass instruments) interfere with the enjoyment, and the extremely rapid tempos once again rob the music of its heft. The Leningrad Philharmonic Hall appears to have been quite spacious enough for the closely-miked performance, but there is little to enjoy here.

AB was recorded in sessions following closely on performances in April 1969, and is also notable as one of three recordings of the GMdm which have been made available in four-channel format, allowing for the separation of the brass bands in the Tuba mirum. (Indeed, it may be that this recording, and the technology which compassed it, are among the reasons performances of the GMdm are often given with the brass bands in the corners of the hall rather than around the orchestra. Conductors and public have become accustomed to the intriguing effect of the greater antiphonal separation.) The conductor's approach is volatile rather than elemental, and the Mormon Tabernacle in Salt Lake City is a most effective space for the performance. Orchestral and choral performance is not up to the level of some of the other "world-class" ensembles. The four-channel separation, auditioned through a suitable decoder, is somewhat effective.

November 1969 brought two recordings of the GMdm which cannot be precisely dated. DA, though recorded at Westminster Cathedral, appears to have tied-in with the performance on 6 June of that same year with the same conductor, orchestra, boy's choir, and tenor, at St. Paul's Cathedral (see Chapter 3). Of this recording little need be said save that Davis' grasp of the Berlioz idiom is complete, and as with BE moods are not played out for mere effect (as they frequently are in OR) but drawn into a unified scheme. The four brass bands appear to have been placed in positions surrounding the performing ensemble, which took up much of the available floor space of the Cathedral. The chorus was large and well-trained. Benvenga's cited source on the "built-in" reverberations of Berlioz' work, as well as Davis' remark quoted in that source, are equally applicable here (see Chapter 3).12

The other November 1969 recording, BU, need not detain the serious listener. It is a provincial run-through with instrumentalists and choruses which are semi-professional at best. The Lacrimosa, the central movement to the work, catches fire, but this is totally defused by the conductor's mishandling of the hemiola at measures 182-83, which blunts its entire climax, The performance is spirited and the recording adequate, The locale was the Oetker-Halle in Bielefeld.

WE represents another non-professional effort, in fact a cooperative one between the University of California at Los Angeles and numerous supporting brass and choral ensembles from surrounding area colleges. The stage of Schoenberg Hall on the UCLA campus was filled to capacity with the 90-piece orchestra (in which composer Paul Chihara played back-desk viola), and 400-voice choir, while the full complement of brass choirs (trumpets instead of cornets and tubas rather than ophicleides) were arranged at the four balcony corners of this relatively small auditorium. The present author, who participated in the performance of 8 March 1972 as a chorister, recalls that the full chorus sang together only in the Dies irae, Rex tremendae, and Lacrimosa; the rest of the movements were carried by a 150-voice choir, with a somewhat smaller ensemble taking the Quaerens me and Hostias. Further, though the timpani were skimped in number, all required timpani notes were sounded by means of having each player able to play rolls on two instruments at once (Berlioz' indications imply that he believed it possible for one player to play a roll on only one drum at a time); it is suggested that this application of modern timpani technique may be standard to modern performances of the GMdm as a space- and money-saving approach which does not short the composer's musical aims.

In addition to the above, it is recalled that an intermission after the Lacrimosa enabled the additional choristers to exit the stage, and further allowed some of the brass band players to take up positions on the stage for the remaining four movements, thus increasing performance accuracy in the trombone-and-flute chords of the Hostias, for example. Lastly, it is specifically recalled that the cymbals at the moment of Judgment in the Dies irae were suspended and hit with sticks, as specified by the composer.

In all, the slight shortage of the instrumental complement (four bassoons rather than eight, and nine horns rather than twelve) little disadvantaged the effect of the musical composition. The effect of the entrance of the four brass bands in the Dies irae was splendid, if a bit inexact, on the date of performance and recording, but had been precise and overwhelming at the dress rehearsal, on which occasion the individual band directors of the four colleges supplying bands had been permitted to act as subconductors, a nicety of insurance regrettably not carried over onto the performance night. It was in fact this particular dress rehearsal and performance which was directly responsible for the present author's interest in the GMdm, and consequently the production of the present research.

Yet another institutional performance, recorded live, was the basis for AV, a most problematic performance. Here again it is possible to enumerate the performers who took part in this presentation, at Portland (Oregon) Civic Auditorium, on 18 May 1974, and while the chorus included no fewer than 240, the instrumental complement is starkly inadequate, a total count of 108 players (excluding the brass bands) hiding the fact that 82 of those instrumentalists were string players. The consequent weakness of the wind sound weakens Berlioz' orchestration considerably, particularly as youthful string players are in general less able to manage proper intonation than their wind-playing peers. The smallness of the brass bands--roughly half of the complement Berlioz calls for--additionally detracts from the monumentality of the work. The performance is further sabotaged by cuts: the identical cut in the Sanctus as noted with FO, though here there were no recording or cost considerations which might have justified it; and the first thirty-nine measures of the Agnus Dei are entirely cut, presumably on the supposition that they are immediately repeated, and therefore unnecessary (not true, because the chordal progression at the beginning which leads from A Major to a D Major durational half cadence to the choral passage modulating from G Major to B Flat Minor is replaced the second time by a progression from A Major to a B Flat seven, as durational cadence to a choral passage beginning in E Flat). This, along with the exceptionally rapid tempi already noted in Chapter 5, make for a quick, sleek GMdm that is over and done with in only 66 minutes (not counting the missing data correction required for the statistical analysis). It is truly a Requiem for the 1974 recession.

1976 was a seminar year in the marketing of recordings and playing equipment for the quadraphonic medium. It is true that AB had been recorded with the four-channel technique in mind, and indeed was released in a discrete four-channel tape format; but the introduction of the SQ modulation quadraphonic disc format by CBS-Sony in the early seventies allowed for the storage, mass reproduction, and retrieval of four-channel information on a standard and compatible stereo disc. AB was released in an SQ modulated format, though it appears not to have received any reviews in American journals. (The Tuba mirum alone was also included on a demonstration disc of classical and popular repertoire by the Vanguard label.) So it was that the more aggressive recording and marketing concerns of Columbia Masterworks (now CBS Masterworks) and EMI Records (along with its holdings Capitol Records and Angel Records in the United States) turned their attention to producing quadraphonic recordings of the GMdm.

The first of these, FR, was recorded 19-22 April 1975. The Great Hall of Birmingham University was the locale, achieving a sense of spaciousness which Hamilton has compared with that in SC (the Invalides): "the space [in FR] is evidently of comparable dimensions if somewhat less overbearing resonance." He further suggests the conception of the GMdm as primarily a choral work, and indicates his feeling that FR fails in the Lacrimosa, an opinion not shared by the present author.13 The chorus is well-trained and the direction and instrumental execution quite secure; overall, it is a very agreeable performance. The brass choirs separate well enough to be heard as discrete entities, and the recording in general is quite clear (the basses' "quidquid latet" in the Dies irae being audible about the timpani for once). One's main regret is that the conductor does not have as firm a grasp on the ultimate Berlioz style as, say, Beecham or Davis.

Though BS was planned for release in quadraphonic sound, some technical problems apparently prevented this from occurring, though copies with a "QAL"-prefixed stamper number were among those distributed at the set's American release. Regrettably one of these was not available for audition. The recording was made once again at the site of the work's premiere, La Chapelle de l'Hôtel des Invalides, being with SC the only instance in which the same locale has been used more than once for a recording. The conductor is one known (in his art as conductor and composer) for a dramatic, even theatrical approach, even in (or perhaps especially in) religious subjects. The reading is well-played with the occasional roughness often found in French ensembles, and the forceful choristers have the "wonderful, characteristically Gallic timbre" which Hall finds admirable.14 The recording sessions occurred on 28-30 September 1975.

Excellent orchestral playing but a generally phlegmatic approach spoil MA, recorded in August 1978 at the Masonic Auditorium in Cleveland. (The chorus master is the same as for OR, though here he has a more professional and responsive group of singers.) The thicknesses of the score are brought out, with little of its grace. Similarly BA makes a tremendous noise but ultimately gives little impression of the conductor's familiarity with the score, surprising from one who is among the principal champions of Berlioz' music active today. The French orchestra and chorus (with the same chorus master as that for DA) perform more than adequately, in sessions taped at La Maison de la Mutualité on June-July 1979.

It has already been noted that Berlioz' music has been of extreme interest to recording concerns at each juncture of major development of the recording technique; the pioneering electric Weingartner Fantastique (see Chapter 2) is one example, followed by the pioneering GMdm recordings: FO (the first recording), HO (the first long-playing recording, MH and SC (stereophonics), and AB, FR and possibly BS (quadraphonics)- Hence it is fitting that the twentieth and final GMdm recording to be surveyed here is the harbinger of a new technique, that of digital sound recording. The crux of digital recording is that it represents not a copying of wave-form patterns onto a receptive medium (grooves in wax, or magnetic analogues of those waveform patterns on recording tape), but the storing and retrieval at extraordinarily high speeds of numerical acoustical level readings which when reconstructed create anew the original wave-forms without the attrition, distortion, or other drawbacks of the previous "analogue" methods. It is clear that digital recording, whether it be by the hybrid medium of digital tape to analogue record, or the fully digital Compact Disc introduced worldwide in the Spring of 1983, is the sound storage medium of the future, which may seem a trite statement but nonetheless bears the evidence of truth in a growing technological society.

The pioneering digital recording of the GMdm is PR which to date --Spring 1983--has been released commercially only in the hybrid digital tape to analogue disc format, in which form it was auditioned. Recorded in Walthamstow Town Hall, London, in October 1980, it almost certainly represents a tie-in with the performance by the same executants at Festival Hall the preceding 20 April (see Chapter 3). Not much is known about the technical aspects of this recording, but some slight indication of the logistics of the sessions is given in the somewhat confusing technical note on the placement of the brass bands by the recording producer, Suvi Raj Grubb, reproduced here:

At the first appearance of the brass bands Berlioz asks that band one should be at the North, two to the East, three to the West, and four to the South. The entry of the bands is in the order No. 3, No. 1, No. 2, No. 4--in other words cyclic order from West to South. The illusion of this movement is obtained in the recording by the sound of the bands moving from front left (No. 3), through back left (No. 1), back right (No. 2) and then front right (No. 4). The bands could not have been accommodated round the full orchestra without uncomfortable crowding, so they were placed behind the conductor. Since the stereo microphone facing the orchestra and chorus which picked them up would reverse them laterally, the order of the bands in the picture is clockwise from the right--orchestras 3, 1, 2 and 4.15

In all, PR is certainly a splendid achievement in terms of the orchestral and choral performance, conductorial conception and recording technique, certainly a more than adequate introduction of the work to the inquiring listener. If it misses but the ultimate element of panache of Beecham, or the stoic grace of Davis, or yet the sense of occasion and Gallic rightness of Fournet, as an achievement of the recording art it is a fitting conclusion--for now--to the recording history of this composition.

Copyright © 1983, 1995 by Matthew B. Tepper

Table of Contents

Chapter 6

Chapter 8


Back to The Hector Berlioz Page

Back to Matthew B. Tepper's Home Page