The recording quality here is probably one of the most realistic ever heard in a piano concerto. As I shall explain, the compliment is somewhat backhanded, and I can imagine that opinions, both on this and on the playing, will vary considerably.
From the outset I was sorely tempted to reach for the volume control, partly because of the piano's recessed sound-image and partly because the piano theme itself is a little shy—unusually feminine was my initial reaction. Listening 'blind' and noting the lack of clarity in the first semi-quavers and the cautious approach to the apex of the first short cadenza, I guessed that this was not a top-rank soloist. The mannered end of that same cadenza was a give-away, however, this had to be either Ashkenazy or someone who had listened closely to one of his three previous recordings (the version with Ormandy on RCA, with its curiously tinny piano sound, is deleted).
Ashkenazy long ago reached the stage where he can control and shape every nuance in this teeming piano part and keep poetry and structure in a satisfying balance. Some of his phrasing is uniquely beguiling—the swooning surge into fig. 4 is one of a host of treasurable details on the new recording and it is typical of his sensitivity to emotional ebb and flow. He has always had a special insight into the long plateau before the final peroration, and the spaciousness of the recording emphasizes how beautifully he floats the tone in lyrical passages and how intelligently he withdraws to let the orchestral contribution through. At the opposite extreme there is no lack of imperious grandeur; but the abiding impression for me, emphasized when Ashkenazy takes the larger cadenza as he does with Haitink and Previn (also on Decca), is that he has conquered the work with supreme effort rather than comfortably embracing its full scale (or ''swallowing it whole'' as Rachmaninov said of Horowitz). The climaxes tend to sound pumped up rather than naturally massive.
This is where the recording does not help. The apparently distant piano is how concertos sound in the concert hall. But there the visual aspect and the involvement of the audience help to bring the personality of the soloist into focus. In the empty Concertgebouw, however realistically captured, I get the alienating sensation of sitting alone, at some distance from the musicians and with only microphones, cables and recording engineers for company—one kind of artificiality substituted for another. It may be that, like wholemeal bread, one will adjust to this and come to despise old habits. But I have listened on four sets of equipment and have always found lack of presence, too much ambience, and insufficient clarity especially in the bass. Turning the volume up to get presence only made for an unbearably woofy bass and a slight but intrusive thumping, which I take to be the dampers descending in the strings. Clearly not a solution.
I also wondered whether Haitink himself could not have helped out with the definition of the bass in the first movement development, for instance. He does make some telling points, and the orchestra play magnificently. It is only when you turn to the USSR Symphony Orchestra (with Gavrilov on HMV) that you realize how far the Concertgebouw's brand of expressiveness is from the Russian soul. Interestingly enough the LSO produce more idiomatic playing for Fistoulari (Decca) than for Previn on the earlier Decca recordings. The violins' passionate outcry with the return of the first movement theme in the slow movement is a case in point.
Ashkenazy's reading in 1963, though less vivid in certain details, was just as well integrated and to my mind even more ardent and spontaneous than now. It is the only recording on which he plays the shorter first movement cadenza, which I think suits him better. It has been claimed, by Horowitz and Bolet amongst others, that Rachmaninov preferred this cadenza on artistic grounds. I doubt it myself, but certainly Ashkenazy, fine though his technique is, cannot bring off the big one with the authority of Gavrilov or the scarcely credible power of Berman (CBS 765597, 6/77—nla—and with claustrophobic recording quality). In pianistic terms both these Russians surpass all others (including Rachmaninov and Horowitz)—both have the ability to drench the air with sound, both have power in reserve where others are at full stretch, both have not just tone but 'voice' (so does Ashkenazy, but not on this scale). Stupendous though he is, Gavrilov is perhaps a fraction too keen to show who is boss. Berman, like the composer, is able to convey the impression of the music playing him, as well as vice versa, which needs just as much mastery and even more daring.
All of which suggests that the Rachmaninov No. 3 does not exist (in the sense that Richter's is surely the Rachmaninov No. 2). bolet's is a noble but stodgy account (again on Decca), brightly but artificially recorded. Earl Wild's has been much praised but impresses me more as a sporting than an artistic event (RCA); I fear I may have upset two admirers of his playing by allowing them to pour scorn on his playing of the first movement before identifying him. I agree with the consensus of opinion that Ashkenazy sounds rather self-conscious on the Previn version.
Personal preference is bound to vary. I would suggest, for the shorter cadenza and for a deeply satisfying but not overwhelming experience, Ashkenazy with Fistoulari; for the big cadenza and bigger pianism, Gavrilov and, if you can find a copy, the titanic Berman; for spontaneity and a sense of danger, as well as historical interest, Horowitz with Reiner (RCA—digitally remastered on GL85262—by no means so technically superlative as has been claimed) and the composer's own performance (now deleted)—both of these come with the disfiguring cuts. As recorded sound none of these can rival the new Decca, but all of them seem a lot more involving.'