San Francisco Classical Voice

These two brothers are frighteningly good. Renaud, the violinist, ought to have a sideline advertising whatever rosin he favors, because I have never seen a bow attach itself so neatly to the string and stay there as infallibly as that. Gautier, the cellist, sometimes lets his vibrato get too fast or too wide or both for comfort, but like his brother he's cabable of paring the vibrato down to near-zero when he wants to. Both are fearless players and marvelous technicians, and this is all music in which there's no cover at all.

Of the four works that the Capuçons programmed Tuesday night, Maurice Ravel's 1922 Sonata is by some distance the most often performed. In a way, that's a little surprising, because the piece doesn't really go out of its way to be listener-friendly. One observer, the composer and critic Roland-Manuel, memorably described the Sonata as "bristling with virtuosity and a lyricism which spits like an angry cat." It's not clear that this was meant exactly as a compliment (in the same sentence, Roland-Manuel calls the piece "one of the most significant — and least flattering — works in Ravel's new manner"), but it's an uncommonly good description of Ravel's prickliest work.

The Sonata's raw material is an alternation between major and minor, one so calculated as never to allow the listener to know which (if either) is the "real" mode. A phrase will use one note (making for a minor triad) on the way up an arpeggio, and the other (making for a major triad) on the way back down, or vice versa. If that sounds abstruse, the bedazzlement and mystery it affords the music are not. The whole piece is really something of a later, drier gloss on the Piano Trio of eight years earlier, with all the richness and softening of the piano's colors taken away and the remainder honed to a knife-edge. The slow third movement, for example, is a later, sparer sibling of the Trio's "Passacaille," and the second movement has all the energy of the Trio's "Pantoum" without any of the elfin mirth. The finale's last peroration, with its buildup of closely overlapping repetitions of the main motif, manages to be dense and economical at once in a way that I don't believe Ravel had tried before or ever did again.

The Capuçons' approach to all this was rhythmically flexible to the point of positive danger, or rather would have been had they not been in such thorough control. As it was, the effect was of a thrilling spontaneity that I strongly suspected wasn't any such thing, but that certainly contrived to sound like it. The finale, especially, was full of rubato, but also paced with rare expertise, so that what you perceived wasn't "players pulling the music around," but a natural, in-the-moment give and take between the players, culminating in a visceral run to the finish that would occur naturally to anyone in the heat of the moment in this music, but probably wouldn't then be executed quite as cleanly as that.

Erwin Schulhoff's Duo preceded the Ravel, which in a way was a pity, not just because it was written three years later, but because its first movement seems to allude to the Ravel more than once. Had the order been the other way around more people might have noticed. Schulhoff is one of those composers cursed by being remembered primarily as a victim of the Nazis, though in his case the Nazis may have been so obsessed with his characters as a Soviet citizen and a Communist that they hardly noticed he was also a Jew.

Here, this piece was in the company of the very best of its genre, and did not sound at all out of place. The smooth, even-handed counterpoint of the opening and the clusters of harmonics at the end of more than one movement are what suggest Ravel, along with a few actual melodic lines in the first movement that seem to have come straight out of the French composer's Sonata. The "Zingaresca" second movement is one of those left-hand-pizzicato spectaculars, rather like (again) Ravel's own Tzigane.

Martinu's second Duo, following after intermission, is one of his last pieces, from 1958, and one that ought to be played more. I hesitate to say this, as a violist, but this is a better work than the much more frequently played Three Madrigals for violin and viola of a decade earlier, and not least because it's a few minutes shorter. There's the same ecstatic lyricism, even a lot of the same textural devices. (The interlocking/overlapping-two-parts thing seems to have been a Martinu favorite in sparse textures.) But this piece manages to be rich and sweet at once, without knocking you out with fat or sugar overload, or overtaxing your patience either. I don't know whether 1958 was the last year anyone could write music so innocently happy, but I'm glad it was possible as late as that.

Ending the program proper was Zoltán Kodály's Op. 7 Duo, from 1914, which is likely the best-known violin/cello duo apart from the Ravel. If the Ravel puts much the larger burden on the cellist, who has to drift around in violin range as though that were trivially easy, the Kodály assumes a flamboyant violin virtuoso to whom the cellist is not exactly backdrop, but not quite subject to the same demands either. Certainly Renaud Capuçon didn't seem unhappy in his tenths, his octaves, his multiple harmonics, or his little stratospheric flights into Gypsy Virtuosoland in the central movement.

The encore was the Handel-Halvorsen Passacaglia, which might possibly contain just enough Handel for USDA approval, though it's emphatically a "Johann Halvorsen-added product." The Capuçons played it as brilliantly as Perlman and Zukerman did the violin/viola version in a concert from the late 1970s that's since become legendary, though I can't shake the sense that they cut a variation or two. Maybe they did.