Romanian conductor Sergiu Celibidache acquired a legendary status in his own lifetime, largely on account of his refusal to make recordings, but also because of his unusual theories about music and his extremely slow tempos. After his death a couple of years ago, his family decided to authorize release of selected performances in order to prevent them from becoming pirated anyway, and although they would never say so, they no doubt hoped to cash in on his legacy. Don't get me wrong: they should try to cash in on his legacy if they can. The only question is this: do these releases do him justice as an artist, and fairly represent what he was trying to achieve? By and large they do, and where his performances fail (as in the Beethoven symphonies), it's useful to keep in mind the fact that the old man himself never would have permitted them to be released in the first place.

But beyond the issue of Celibidache's intentions, there is the Munich Philharmonic to consider. These recordings belong equally to the orchestra, and even if they only demonstrated the players' dedication to their eccentric conductor, these discs would be worth hearing. The orchestra plays with 100 percent dedication, and often achieves astonishing results. This is particularly true of the Bruckner symphonies, where there are countless magical moments: the finale's coda in Symphony No. 4, the modified timpani parts at the end of the Fifth's "Scherzo", a granitic Ninth first movement, and all of the slow movements. The F minor Mass is another incredible performance, slower than humanly possible it would seem, but gloriously sustained nonetheless. These are magnificent performances, the only letdown being No. 7, markedly slower even than the Sony video production of several years ago, and here the reason must be the conductor's own frailty. He already was quite ill when he gave this performance in 1994.

Bruckner's symphonies, of course, generally sound better the slower you play them. The same can't be said of Beethoven, and here "Celi" (as he was fondly called by his orchestra) has a philosophical problem. He treats music of the Classical Period much the same way that Glenn Gould did: balancing the contrapuntal lines, smoothing out the contrasts, and eliminating all first movement repeats in order to present a linear progression utterly devoid of incident. That this runs completely counter to the spirit of the music--particularly Beethoven's music--is beyond question. The result is the most perversely dull interpretation of these works imaginable. There's only one symphony where Celibidache's approach works: the Sixth, which is generally relaxed anyway, and can stand the conductor's obsession with pure line and texture. Otherwise, he's completely out of his element, nowhere more so than in the Seventh, where he takes more than 16 minutes for the first movement, minus the repeat. (Klemperer, a notoriously "slow" conductor, takes a little more than 12 minutes.)

Celibidache's Brahms, on the other hand, though a little on the slow side, is largely as successful as his Beethoven is bizarre. The musical expression being predominantly lyrical and contrapuntal, the conductor is in his element, clarifying textures, letting the music breathe, and creating sounds as beautiful as we're ever likely to hear. The opening of the Fourth, for example, has a purity and delicacy that makes the movement's ending all the more tragic in contrast, while the finale is both swift and ferocious. And although the ending of the Second may not be the last word in physical excitement, it certainly doesn't suffer from lack of momentum or sheer musical energy. Celibidache's Schumann Second, though, which comes coupled to Brahms' "Haydn" Variations, is as dead as his Beethoven.

All in all, like most of the "mad genius" conductors (Stokowski, for example), Celibidache's work ranges erratically from the stunning to the grotesque. But at his best--in Brahms, and especially Bruckner--he's very special indeed.