The best current reviewer of classical music recordings isn't in the print media and doesn't, as far as I know, have a website. He's Simon Roberts, an Australian who lives in Philadelphia and works as a lawyer (yes, a Philadelphia lawyer), and in his spare time posts to the usenet newsgroup rec.music.classical.recordings. Roberts isn't a guide to everything; he doesn't care for most French or Russian or American music, which means that most of his posts are about German music of the 18th and 19th centuries. But if you go to Google Groups advanced search, enter Simon Roberts in the "author" field, and do some searches on posts he's written on the great German dudes (Bach, Handel, Mozart, Beethoven, Schubert, Haydn), you'll find that he has a broad knowledge of available recordings and exceptionally interesting things to say about them. His posts have alerted me to some great recordings that were ignored by most of the standard classical magazines (things are a little easier now that classicstoday.com reviews a lot of less-widely-distributed product). For instance, there was a recording a few years ago of Haydn's opus 33 string quartets by the Apponyi Quartet, a group that made only two recordings and may not even exist any more. The recording was not reviewed in Gramophone or the Penguin Guide, and I'd never heard of it until Roberts proclaimed on rec.music.classical.recordings that these were "the best performances of any Haydn quartets, HIP or
otherwise - remarkably alive and imaginative." I ordered the recording from Germany (the only place it's still available), and found that Roberts was absolutely right: even with all the good recordings of Haydn quartets, these stand out for terrific playing and willingness to make Haydn sound bold and brash rather than just cute and "elegant."
That kind of sums up Roberts' taste in recordings in general; he likes performances and recordings that take chances, are exuberant and extroverted and not afraid to make a lot of noise. He doesn't have much use for "understated" or "subtle" performances. I like that, because first of all it means that a recording he recommends will almost never be dull; this puts him at the opposite extreme from many major classical-music critics, and from the people who run the major labels. (There was a guy who posted on rec.music.classical.recordings, a fellow Canadian, who was involved with A&R for the Philips label. The guy turned out to be a loon, a rabid anti-Semite, and rude to boot... but perhaps worst of all, he turned out to have the blandest taste imaginable, preferring artists who were so "subtle" as to be dull and denigrating artists who actually produced attractive or entertaining sounds. No wonder the "major" labels are collapsing, when they put such people in charge.) His favorite conductors include Toscanini, Scherchen, Klemperer, the younger Bernstein (before he left the New York Philharmonic), Harnoncourt, Rene Jacobs, Frans Bruggen -- guys who take a lot of chances and make music sound excitingly rough and tough.
But what's most interesting about reading a lot of Roberts' posts about orchestral music, and what I barely heard about before I started reading him, is his emphasis on balances. You hardly ever hear about balance in the Penguin Guide or Gramophone; a conductor is praised for generic stuff -- "shaping," "sense of line," etc., but there's rarely much reference to the specific things the conductor makes the orchestra do: does he make the trumpets stand out? Does he go for a string-heavy balance, as Ormandy did, or does he emphasize the woodwinds, as Klemperer did? And so on. Balancing is perhaps the most obvious way for the conductor to shape the way the orchestra sounds, and yet you rarely hear about it. From Roberts, you hear about it all the time; he's particularly concerned with the timpani, preferring that they should play out and make a real racket. (This is particularly important with composers like Haydn and Mozart and the baroque composers, who didn't always include timpani parts; when they did, therefore, you have to assume that they wanted the timpani to be heard and have a real impact.) When you listen to conductors' recordings this way, you can not only learn more about their styles of conducting, but find that their styles are sometimes the exact opposite of the way the "standard" critics describe them.
An example is the late Georg Solti. Based on his brass-heavy, nervously energetic performances of Wagner, Strauss, and Mahler, Solti is mostly considered a high-energy conductor, given to "driving too hard" and letting the orchestra make too much noise. But Roberts pointed out that in pre-Wagner music, in Mozart, Haydn and Beethoven, Solti was the exact opposite of the stereotype; his problem was undue restraint in balancing, making the brass and timpani play softly and daintily and going for a generically pretty sound. This is definitely true of most of Solti's recordings of those composers (though his first recordings of The Magic Flute and Cosi Fan Tutte, both with Pilar Lorengar, are still excellent), and you'd never know it from reading the reviews in the standard record-review journals and books. It took a lawyer posting on the internet to point out that a conductor's problem was not that he made too much noise but that he didn't make enough -- and that's because Roberts describes what specifically happens in a performance or recordings, whereas many critics don't describe much of anything.