Only a few conductors have dared to risk ‘Bolero’ at the slow speed Ravel requested. And, not surprisingly, even fewer of those performances are successful. At seventeen minutes, the inexorable build of the piece becomes disturbingly obsessive, and can easily tip over into heavy-handed bludgeoning. But when ‘Bolero’ is bopped off at thirteen minutes, it sounds like a pops concert in hell. Many prominent recordings can be eliminated from the connoisseur’s attention because they so grossly distort the intended timeframe. Whatever else a thirteen- or fourteen-minute ‘Bolero’ is, it isn’t the piece Ravel wrote. Thus the splashy one by Zubin Mehta and the Los Angeles Philharmonic is out of bounds, as is the cool, coiffed reading by Christoph von Dohnanyi and the Cleveland Orchestra. This also rules out the powerful and highly characterized performance by Claudio Abbado and the London Symphony, notable for using players’ vocalizations to add to the chaos at the end, a theatrical touch originally invented by conductor Victor DeSabata. Even recordings coming in around fifteen minutes can be questionable. The famous RCA release from the early 1960’s by Charles Munch and the Boston Symphony is uniquely aggressive, but I suspect the fastidious Ravel would disown its reckless swagger. Virtually duplicating Munch’s pace, though with much more “sanitized” textures (and better recorded balances) is the 1980 Decca recording by Charles Dutoit and the Montreal Symphony. Still, though, it is too fast.

To find the real Ravel, we must discard these recordings and look for those falling closer to the fabled seventeen-minute mark. The recordings I know of in this range are as follows: Charles Munch and the Orchestre Nationale de France (the 1967 EMI recording), Herbert von Karajan and the Berlin Philharmonic (1966), Simon Rattle and the City of Birmingham Symphony, Daniel Barenboim and the Orchestre de Paris, Leonard Slatkin and the St. Louis Symphony, Riccardo Muti and the Philadelphia Orchestra, and Stanislaw Skrowaczewski and the Minnesota Orchestra.

It’s hard to tell from the Munch if it is an intentionally revised view of the work, or if it is just the declining grasp of the aged conductor which causes the performance to run over two minutes longer than the earlier one; it certainly lacks something in intensity compared to the Boston recording. The performance by Simon Rattle is rather grim, dogged and choppy. Barenboim’s earlier performance on Deutsche Grammophon is certainly… loud, and I have not heard the Slatkin and Muti recordings in many years, though I recall the Muti as being competitive, if a little inflexible.

That leaves Karajan and Skrowaczewski. Karajan is usually regarded as the standard for the piece, but I must respectfully demur. Karajan’s rendition strikes me as being an odd blend of cross-purposes. The melodic textures are velvety smooth and sensual, but the accompanying percussion is Teutonically rigid, more suggestive of a march than a dance. Furthermore, Karajan doctors the score, adding additional drums just before the violins enter. This wouldn’t be such a problem if the drums weren’t so prominent in the mix, but they are, and it becomes annoyingly clear that they aren’t well synchronized, perhaps a byproduct of Karajan’s baton technique. Furthermore, the recorded balance is such that near the end, the chords in the strings almost drown out the brass, which is a perspective that you’ll never hear in any concert hall in the world. In short, it’s an overrated performance. And at 16:08, it’s still too fast.

Thus we come to Skrowaczewski. It is among the slowest, at 17:22, but Skrowaczewski has the controlling hand to sustain this daring tempo. He sculpts the Minnesota Orchestra’s sound to a lean, incisive edge, a very different feel from Karajan’s velvet plush, and much more in line with what we know of Ravel’s neoclassicism. Each instrumental solo is characterized with an individuality that makes most other recordings seem generic. From the first tremulous flute solo, to the suave saxophones, to the wickedly witty trombone, it’s a performance that somehow manages to be full of personality without ever stepping outside the boundaries of Ravel’s aesthetic. What’s even more amazing is the way that Skrowaczewski never releases his grip, even in the closing pages. Whereas most conductors let the dynamics rip toward the end in order to relieve the built-up pressure, Skrowaczewski keeps everything in check until the final collapse, emphasizing the theoretical structure of the piece over purely dramatic considerations. Skrowaczewski is constantly aware of where the theme is and how it has to flow to the next part, and where that is in relation to the overall work. Few other conductors take the piece seriously enough to master its structure. By using such adept flow and control, Skrowaczewski’s pace never seems slow, even at seventeen plus minutes. Barenboim’s Deutsche Grammophon recording, by contrast, though only ten seconds longer, feels much slower. In sum, Skrowaczewski’s Bolero stands as the one and only performance we have which accurately realizes Ravel’s vision.

Moving onto the other apocalyptic piece on this recording, we come to ‘La Valse’, and stylistic issues become murkier. Here, Ravel the neoclassical French impressionist is writing about the Vienna of the romantic period. Thus a fair argument could be made for almost any stylistic approach, and many have been tried, from nervous to voluptuous. Ultimately, the most satisfying performances strike a balance between those extremes. Again, Skrowaczewski proves adept at pulling off the nearly impossible. This performance features the liveliness and nervous energy necessary to make it catch fire, but it also shows a voluptuousness not usually associated with this conductor. His rubato is natural and convincing throughout. It seems, for most of its length, to be a performance on track to greatness. But just before the end, Skrowaczewski seems to shy away from drilling the piece home with the destructive vigor required. Perhaps the work’s apocalyptic collapse is something that this strongly humanistic conductor cannot bring himself to enforce. Whatever the case, he fails to nail the coffin-lid shut, thus just missing a definitive performance. To hear the end of ‘La Valse’ in all its devastating glory, turn to Claudio Abbado and the London Symphony on Deutsche Grammophon. Abbado likewise manages to get a performance both nervous and voluptuous, and the early digital recording is less harsh than most from that period. For those who prefer a faster, more nervous and less decadent ‘La Valse’, there are classic recordings by Munch (RCA), Monteux (RCA, Philips), Markevitch (EMI), and van Beinum (Philips). For those who want a more measured pace, there are two recordings by Pierre Boulez (Sony, Deutsche Grammophon). And for those hearty souls who want to push decadence right over the cliff, there’s an extraordinary performance by von Karajan from 1971 with the Orchestre de Paris, which is intermittently available on EMI. It lasts for over fourteen minutes and is so wickedly lush, you might feel obliged to take a shower after you listen to it.

The remaining works on this disc pose considerably fewer interpretive problems. The ‘Rapsodie Espagnole’ has had no shortage of successful performances over the years, including one by Pierre Boulez and the Cleveland Orchestra on Sony stereo SACD SS89121. Boulez is a touch matter-of-fact in the opening “Prelude a la nuit” and in some of the work’s other lyrical moments, but he is surprisingly theatrical in the closing “Feria”, pushing the tempo thrillingly in the closing pages. Skrowaczewski makes much more fascinating work of the mysterious textures in the “Prelude”, but keeps the horses in rein at the end. The Sony SACD features reasonably good stereo sound from 1969, although it shows signs of the multiple miking that was to spread like a plague throughout much of the industry in the 1970’s and 80’s. But the original Vox recording on this Mobile Fidelity Sound Lab issue was made with fewer microphones, better balances, and greater care. For what it’s worth, the remainder of that Sony SACD is stereo-only versions of Boulez’ quadraphonic Ravel recordings with the New York Philharmonic, which have subsequently been reissued in a multi-channel release, which helps spread out the echoey acoustics of Manhattan Center. For those who like Boulez’ Ravel the multi-channel hybrid is the way to go. That does beg the question, however, which of Boulez’ approaches to Ravel? If ‘Rapsodie Espagnole’ shows Boulez in dramatic form, and the ‘Daphnis and Chloe Suite No.2’ finds him enjoying exquisite gossamer textures, one might expect that Boulez would lead effective versions of ‘Bolero’ and ‘La Valse’. Unfortunately, Boulez is so intent on eschewing all “extra-musical” considerations that he trudges through both pieces barely even suggesting the passion and power that they contain. Skrowaczewski may be reserved, but at least he is consistently in touch with the soul of Ravel’s music.

Skrowaczewski’s poise comes to good use in the Pavane pour une Infante defunte, helping sustain a tempo that is characteristically a little slower than the average. His tempo for the ‘Daphnis and Chloe Suite No.2’ is likewise broad, coming in at over seventeen minutes (unfortunately, the movements are not separately tracked). Here, Boulez is more formidable competition, as the work seems to engage his interest in complex textures, but Skrowaczewski is certainly no slouch in that department, bringing more life and less gleam to the music than Boulez.

It is with the ‘Daphnis and Chloe Suite’ that we come to the crowning glory of this multichannel SACD – when it was originally recorded in 1975, the end product was a quadraphonic LP consisting of two front channels and two rear channels for ambiance. Elite Recordings used four Schoeps M221B microphones across the stage in an omni-directional pattern for the two front channels, and two widely spaced cardioid Schoeps CMC60’s facing away from the stage for the rear channels. Producers Marc Aubort and Joanna Nickrenz elected to place the offstage wordless chorus in the rear of the hall so that the voices would float in from behind the listener, demonstrating the possible glories of multi-channel sound. As we all know, the home reproduction equipment in the bad old days was never quite impressive enough to convince the public at large to go for quadraphonic sound, but now that we have digital surround sound, this stunning recording can finally be heard properly. The wash of sound from the orchestra is amazing enough in a piece like this, but having the voices float in languorously from somewhere behind you is simply magical. Anyone wanting to demonstrate the value of multi-channel sound will love this disc.

The original Vox LP was a famous sonic document in its own day, and it sounds better than ever in this new Mobile Fidelity Sound Lab incarnation. It completely supersedes the earlier MFSL regular CD release (MFCD 802) from the 1980’s. Even those who don’t have SACD players would be well advised to obtain this release, because it is a hybrid disc, and even the regular CD layer is much more vivid than the older issue. At first, one might think it is just at a higher level, but careful comparison shows that it is much richer in texture than the original CD. MFSL’s GAIN2 (“Greater Ambient Information Network”) processing is presumably responsible for this. The MFSL engineers have gone back to the original master tapes to analyze and process them in order to obtain more sonic information, and it proves to be a rousing success. For detailed information on this fascinating system, investigate the technology section at: http://www.mofi.com/

The stereo SACD layer of this release also offers special benefits to those who prefer stereo sound and may have listened to the original quadraphonic disc in stereo: The pure front channel mix. After all, listening to the quad LP in stereo would have meant having a certain amount of additional sound overlaid from the rear channels, whereas modern hybrid discs have everything sorted according to which layer you select. But the real star here is the multi-channel mix. Since this was originally a quadraphonic recording, the multichannel mix we have here is 4.0 (2/2.0 – two front channels, two rear). It is close enough to capture solo instrumental timbres so vivid you can almost see them, yet at the same time, it is spacious enough to feature some hall ambiance. The sound quality of these recordings in Minneapolis’ Orchestra Hall are rightly famous, and although many other companies have recorded the Minnesota Orchestra there since then, few have done as impressive a job (one exception being Reference Recordings). One especially interesting acoustic anomaly is discussed by Marc Aubort in the program notes: Although the hall worked well for recording purposes, there was one spot toward the back of the stage where you could clap your hands and hear a distinct bounce-back echo from the hall. As it turned out, this spot was exactly where the snare drum went in the orchestra’s seating arrangement. So if you listen carefully in Bolero as volume increases, but before the textures thicken, you can hear the echo shadowing the player like a sort of “ghost drum”. On the original LP, this quirk is lost in surface noise and on the previous MFSL version, it appeared that the drum sound was for some reason not in focus. Now the quirk can be heard for what it is. Fortunately, it is a minor anomaly, and only adds to the distinctiveness of these recordings.

There is an interesting bonus track that will intrigue surround-sound listeners. As noted above, there were originally four microphones across the stage to pick up sound for the front channels. These feeds were mixed on a Studer 169 console leading to the Scully 4 tape machine. If the feeds had been committed to tape separately, it would presumably be easy to create a 5.0 mix for the present release. But the challenge of experimentally creating a 5.0 mix piqued the MFSL engineers, and they created this bonus track of the “Danse generale” from ‘Daphnis and Chloe’ by running the master tape through an AGM TSS1 processor, which used ambisonic algorithms to redistribute sound over the three front channels. How well does it work? Well, there are advantages and there are drawbacks. The big advantage is that pulling some of the left and right track information to the center creates a more definite location for the sound of the woodwinds of the orchestra, and that focus seems to give them a touch more presence and warmth. But the pulling of information from the side tracks also means that the sound (and placement) of the string sections are pulled toward center, too, slightly exaggerating the size (or shape) of the various sections as laid out on stage. This pull toward the center also has a tendency (if heard following the 4.0 version of the movement) to create the aural illusion that the sound stage is being narrowed. Careful comparison with the 4.0 tracks prove this isn’t the case, but the pull to center is strongly noticeable. While the added warmth and definition of the winds is welcome, I would have to say that I found the 5.0 mix just a little too aggressive to be comfortable with. Listeners less familiar with the original recording may not find themselves bothered. It is an interesting undertaking, however, and if it can be refined in future usage, there may be considerable applications for its use.