What a nice concept, this compilation: Heifetz playing three double concertos, and each time partnering with a different instrument: violin in Bach, viola in Mozart and cello in Brahms. All these recordings were independently made and published, originally: Mozart's Sinfonia Concertante with Arthur Benjamin's Romantic Fantasy also for Violin, Viola and Orchestra (too bad that one is gone - it is now on Stravinsky: Suite Italienne / Benjamin: Romantic Fantasy & Others Recorded 1953-1967 (The Heifetz Collection, Vol. 31)) on RCA Victor LM-2149 (1956), Bach with Beethoven's Kreutzer Sonata on Victor LSC-2577 (1961) and Brahms alone on Victor LDS-2513 (1960).
Like Toscanini, Heifetz and those who partnered with him knew something about music that went lost after them: how to set it on fire. Heifetz first recorded Brahms' Double Concerto in 1939, with Emmanuel Feuermann and the Philadelphia Orchestra under Ormandy, and it was fiery (Heifetz Collection, Volume 5 (1939-1946)). In 1948 Toscanini also made a live recording with his NBC Symphony orchestra and its two first seats, Mischa Mischakoff and Frank Miller and it was, if possible, even fierier (Johannes Brahms: Symphony No. 3/ Double Concerto (for Violin & Cello)). But it was a thing of the times, really: when Oistrakh and Knushevitzky recorded it with Karl Eliasberg in Leningrad in 1948 (David Oistrakh Plays Doubles), and when Milstein and Piatigorsky recorded it with Reiner in 1951 and in Philadelphia (Piatigorsky Legendary Performers - Strauss (Don Quixote) & Brahms (Double Concerto)), it was a touch less urgent and fiery than with Heifetz and Toscanini, but still, it was urgent and fiery.
But the "revisionist" view of Brahms' Double Concerto started soon after, and can be traced back - it'll come as no surprise - to the Austro-Germans: : Schuricht in 1947 with Kulenkampff and Mainardi (Kulenkampff plays Brahms: Violin Concerto in D major, Double Concerto in A minor & cello), Furtwängler in 1952 with Boskovsky and Brabec (Furtwangler Conducts Brahms: Violin Concerto (Menuhin) & Double Concerto (Boskovsky, Brabec), but the latter is a live recording that was published only I the late 1970s). The Viennese émigré Bruno Walter (both in his 1954 recording, with Stern and Rose, and in his 1959 stereo remake with Francescatti and Fournier, but I'll send the product links in the comments section) and Galliera in 1956 (with Oistrakh and, already, Fournier) also established a less urgent view of Brahms, which didn't lack bite and explosiveness in the first movement's orchestral outbursts but lingered more in the lyrical passages, played the middle movement "adagio" rather than "andante", and insisted more on the "non troppo" of the finale's character indication than on its "vivace" part.
And that it was a thing of the times rather than of individuals, or of the generation or geographical origin of the performers, is amply shown by the interpretive evolution of both Oistrakh and Ormandy. While their first recordings had been fiery, their next ones (Oistrakh also did the famous recording with Rostropovich and Szell in 1969, and Ormandy a stereo remake in 1964, again with Stern and Rose) were spacious.
Naturally, something was revealed and gained in the new approach - more of the majestic grandeur of Brahms (and the more of it as the approach became more extreme and tempi dragged out in the opening movement, as with Furtwängler, Karajan, Bernstein and Harnoncourt), and also the "tender lyricism" of its slow movement, as one Gramophone reviewer once put it (to complain that Heifetz missed it) - but it did away with that essential element of Heifetz' Brahms - and, in my opinion, of Brahms' music: the fire, the passionate intensity. Not that the "old" approach lacked lyricism, either: first because you can be sure that all these guys, Heifetz, Oistrakh, Milstein, Feuermann, Knushevitzky, Piatigorsky, knew how to make their instrument sing. And also because, by refusing to linger on the lyrical passages but keeping them at a flowing tempo, they made the lyricism (to my ears) all the more searing and intense, with an undercurrent of pasion rather than tenderness, if not placidity.
Still, by the mid-1950s, a new paradigm - interpretive and listening - had been established. Then, Heifetz' stereo remakes came to be seen as too fast and ungiving: it is very telling to compare the Gramophone review from March 1962 of this recording, and the one from August '42 of his earlier one from 1939 (the Gramophone Archive is available online). The approach in both versions is strikingly similar, with infinitely better sonics in 1960, yet while the earlier version was hailed, the new one shot down in flames.
There are pluses and minuses of 1960 over 1939. The 1960 sonics are stupendous, with vivid presence and no tape hiss that I can detect, even over headphones. Incidentally, the CD layer of this SACD showed no improvement I could detect over the first CD reissue of the recording, Double Concerto, which already sounded stupendous. Other than the sonics, Heifetz has lost none of his powers and crystalline tone - it is even more in evidence here, thanks to the recording - and Piatigorsky is equal to Feuermann in tone and virtuosity, but again, afforded incomparably greater presence by the recording.
But the major minus is that Wallenstein's first movement isn't as fiery as Ormandy's. It has to do less with tempo (the movement is less than 20 seconds longer) than with phrasing and articulation: Ormandy's Philadelphia strings play with considerably more bite in the orchestral outbursts than Wallenstein's RCA Victor Symphony's, who lets some of the thickness of the "new" approach creep in, and you can hear it as soon as the first introductory orchestral tutti. That said, all is relative: it is something that might strike you if the versions of Ormandy and Toscanini are very present in your aural memory; in its own right the version of Wallenstein is one of considerable power. But the 1960 Finale is as great interpretively as 1939 - and with incomparably better sonics.
Many listeners attuned to the "other approach" might feel ruffled (and the Gramophone reviewer back in March '62 certainly was!) by the performers' urgent tempo in the middle Andante: compare their 6:35 to Walter's 7:52 in 1959 with Francescatti and Fournier and Ormandy's 8:09 (in his 1964 remake with Stern and Rose). I find it entirely convincing in its own right (which doesn't mean that I don't find the slower approaches also convincing). One can of course discuss endlessly about what Brahms had in mind with his "andante" marking, but that's not the point. The Gramophone reviewer complained that "the lyrical tenderness of Francescatti and Fournier [was] entirely missing". Spot on! Rather than that "lyrical tenderness", it is an atmosphere of aching passion that Heifetz and partners convey, which IMO is entirely true to Brahms' message.
Make no mistake. Despite my small reservations about the first movement, this is Hall of Fame Performance, Great Recording of the Century, Desert Island version. It is the recording of Brahms' Double I'll keep if I have (Heaven forbids!) to keep only one. Thank the God of music that Heifetz was still around and with undiminished powers to re-record this in great stereo, and that, unlike Oistrakh, he never mellowed or relented (as far as I am aware there are no stereo recordings, even live, by Milstein). Thanks to that we now have this version which stands out among many as a truly UNIQUE testimony in great stereo of an art of interpreting Brahms which, sadly, now that the interpretive models have evolved in another direction, seems to have gone lost (and this is no figure of speech: of the more than 20 other versions I know from the stereo era, only one, Gielen's, recaptures something of the intensity, passion and drive of the early versions). Sadly, because it said things about the musical and emotional message of Brahms which I strongly believe are true and essential to this composer. The "other", grander, more majestic rendition of Brahms' Double Concerto, which has now become standard, for all the beauties it may have to offer, is simply, in my opinion, a "revisionist" view of the composer, one that does not let you hear anything close to what Brahms intended - and one also that, for me, generates much less excitement and emotion. The lesson Heifetz and partners offer on interpreting Brahms deserves not only to be remembered, but given a new lease on life.
The 1956 Mozart Sinfonia Concertante with William Primrose is also a great one, and Izler Solomon and whatever was the RCA Victor Orchestra must take part of the credit: apparently a chamber-sized ensemble, they play with great zest in the outer movements, an enthusiasm that matches the soloists', and a superb transparency of textures, with the all-important oboes and horns never covered and adding great timbral character. There is, as ever with Heifetz, an urgency in the playing, but one that I don't find so much rushed as enthusiastic and ebullient in those outer movements. The middle movement is for once taken as the "Andante" Mozart wrote, not as the romantic and solemn "Adagio", kin to the Masonic Adagio & Fugue, everybody seems to want to turn it into, but by way of paradox that very urgency lends it an undertone of inquietude which I find very romantic indeed (in the way Don Giovanni was the opera preferred by the Goethe, Berlioz and the romantic era). I find it outstanding. And here, thank again the God of music, the lesson is starting to be heard. I've just heard the recent version of Carmignola and Abbado: the tempi and approach are strikingly close to Heifetz' (Mozart: The 5 Violin Concertos). So maybe there is hope for Brahms also, when Carmignola and Abbado record it. The 1956 recording is stereo and vivid, only fault being a very soft electronic buzz that you can hear over headphones in the moments of silence.
The 1961 Bach Double with Erick Friedman suffers from acoustics that are far too resonant, making the orchestra sound like a large one recorded in a church and loosing the chamber-like intimacy of the Concerto. But Heifetz' brisk tempo in the first movement is great for avoiding the kind of pedestrian squareness associated to Bach interpretation in the 1960s (for instance with the Oistrakhs father and son, or Menuhin and Ferras), and truly points to the future of Historically-Informed Performances. Likewise the fastish Largo is vibrant without the kind of sentimentality that derives from slower tempi, and one thing that those who accuse Heifetz of always playing too fast don't seem to hear, is his unique art of singing, the sinuousness of his phrasing and throbbing vibrato (and that's certainly a feature that links him to the past). Heifetz (and I am assuming that he is playing the upper part) is also admirable for not trying to put the spotlights constantly on himself, but letting his young colleague take or share prominence when the music demands it. Sargent's conducting is not particularly imaginative or illuminating.
Did I say that this was Hall of Fame status, Great Recordings of the Century, Desert Island Disc? Sure I did! But I'll never repeat it enough.
P.S. As already mentioned, the CD layer of this SACD shows no significant sonic improvement over the previous reissues. On the other hand, the first copy I bought of the SACD would not read on my computer, which means that I couldn't rip it onto it or onto my ipod either, and only with some difficulty was my computer able to read the second copy I purchased. So if your usage is to read CDs on your computer or rip them onto portable devices, I'd suggest you favor any of the previous reissues.