Graham Vick's production plays on bare, semi-abstracted yellow-ochre sets (designer: Richard Hudson) that make their point when we reach the desert scene in Act IV but are otherwise just neutral space. Its emptiness gets filled - working the chorus hard in Acts I and III, camping up Act II - but with laboured routines. The embarcation scene is amateur, with desperately over-acting whores. The death is nothing. If Vick does care about his characters, it doesn't show.
But then, he doesn't have the best material: it's a bizarrely dead cast. Only Patrick Denniston's des Grieux makes any impression, and that's more physical than vocal. The Lescaut, Geronte and Edmondo are all workaday stuff, not Glyndebourne quality. And the Manon, a young Romanian called Adina Nitescu, only previously heard in Britain as a Cardiff Singer of the World contestant, is appallingly miscast. An old-time Eastern European diva in the making, she presents with an attractive, fragrant, softly textured voice but with the personality and sexual allure of suet pudding. Staring blankly into space or straight at the conductor, there is no engagement with her fellow singers. Every move is painfully premeditated. And as one half of what is arguably the least convincing instant relationship in modern opera, she fails to ignite it with the smallest spark of passion.
All in all, it's a substandard show, and I'd avoid it but for one thing. The conductor. John Eliot Gardiner has passed through many lives and much repertory since he made his mark in period performance, and though few could have guessed that he would make his Glyndebourne debut with Puccini, it turns out to be yet another example of his inability to put a foot wrong. He spring-cleans the score with bracing and cathartic vigour, galvanising the orchestral playing - brilliantly delivered by the LPO - and gives the whole performance credibility. It needs it.
Fortunately Owen Wingrave, which opened at Glyndebourne on Thursday, is everything Manon Lescaut isn't: an imaginative, purposeful production (Robin Phillips), beautifully delivered by a fine cast with an utterly superb young artist in the title role. Technically it's a revival, of a staging from the Glyndebourne Tour two years ago. But in the process of transfer to the Festival there have been changes to the sets and lighting, generally to make the scenes more site-specific. And in doing that, it comes closer, I think, to answering the peculiar needs of a demanding piece.
Like Manon Lescaut, Owen Wingrave is a secondary work in its author's catalogue. Britten wrote it for television in 1971, suspicious of the medium but still attempting to exploit what broadcast fantasy could do, with short scenes that dissolve into each other and sometimes overlap. When it transferred to the stage in 1973, the scenes were wheeled across an empty stage on trolleys, and it didn't work - helping to fix the opera's reputation as a piece to leave alone. And the fact that it proved musically a transitional piece, fine-tuning Britten's ear for Asiatic, gamelan- style sound-colours in preparation for their definitive statement in Death in Venice, encouraged critics to downgrade it - undervaluing its strengths, exaggerating its flaws.
Listing the negatives, it does pull its punches (especially in the final death scene, which needs more musical profile); it is monothematic theatre, selling its message (pacifism), a touch too hard; and the characters do, in the process, come dangerously close to caricature. But for me, the integrity and beauty of the score - which flows alchemically, decanting menace into calm - wins through. And this Glyndebourne staging proves what I've always believed: that second-division Britten is still first division by the standards of anybody since. The ensemble singing could carry more bite in the attack, the tempi could be faster, but no matter. The performances are powerful, articulate, persuasive, with especially fine contributions from Steven Page as the tutor Coyle, and Gerald Finley, whose Owen outclasses even that of Benjamin Luxon in Britten's original cast. It comes with paradigm conviction and assurance; and the voice is glorious: firm-toned but rich, with effortless projection. I've seen nothing better on an opera stage all year; and with Ivor Bolton's clean, deliberate conducting, it's an obvious contender for the next round of awards.
It's unsettling to visit a historically great music capital and find its greatness more a matter of history than you thought. But such is Budapest: a handsome city that retains the bricks, mortar and memory of musical prestige - a stunning opera house, some jewel-like concert halls, Kodaly and Liszt commemorated in its street-names, Bela Bartok on its bank notes - but has a real struggle to keep the tradition alive.
If there's a solution, it lies outside the country, in the conscience of the rich West; and few years ago a well-connected Swiss composer/conductor called Robert Bachmann took it on himself to establish a Budapest Whitsun Festival, with the intention of attracting quality, cash and worldwide interest to the city. The 1997 season ran last weekend; and the promised highlight was the world premiere of an epic score, by Bachmann himself, whose performance requirements read like an attempt on the Guinness Book of Records. Called Uluru: a Fractal Symphony, it featured 120 separately notated instrumental parts, massed voices, pre-recorded tape, and assorted soloists including someone "believed to be among the world's best didgeridoo players" according to the programme book. Obviously not Rolf Harris. There was also an element of theatre that involved ceremonial processions through the auditorium and Mr Bachmann making a grandly papal entrance, arms outstretched, to the accompaniment of a gargantuan crescendo. And it all amounted to a media event of magnitude, played out for television and before a flown-in, diamond-studded audience.
What the glitterati made of the piece I can't imagine, because its ideas were complex: an attempted musical response to novel concepts in geometry. As a "fractal" symphony, it builds a non-developmental structure out of many parts, each one of which contains a sort of DNA code for the whole and therefore is the whole. In miniature. The parts are freestanding, flexibly ordered so that no two performances should be the same. As for the name, Uluru is the aboriginal word for Ayer's Rock in Australia: an iconic object of spiritual significance in much the way that Bachmann sees his score. And yes, it all makes fascinating theory.
But in practice, in Budapest's cramped Franz Liszt Academy amid the cameras and diamonds, Uluru amounted to little more than the bag of tricks musical avant-gardists played with in the early 1960s and abandoned before the decade was out. The DNA element was simplistically banal: a brutal barrage of arpeggios, scales and trills with nothing in the quality or handling of the material to compensate for its failure to develop. There was no sense of coherent architecture across its 50-minute span. The tape element was crudely interpolated (Stockhausen achieved subtler effects 40 years ago). And the sheer pretension of the piece was laughable.
Mr Bachmann has plans for Uluru. As an infinitely adaptable, perpetual work-in-progress, it will be enlarged into a "fractal opera". It has already been recorded, privately. It's on the Internet. And Bachmann is talking about performances at the South Bank Centre with the LPO or RPO. No doubt he has the money to make these things happen. No doubt he sees Uluru as a cult-score for the new millennium. In truth, it's just an Emperor in not so new clothes.
'Manon Lescaut' continues on Wed & Sat, and 'Owen Wingrave' on Tues & Sun, at Glyndebourne, E Sussex (01273 813813).