Sergei Rachmaninoff
Piano Concerto No.3 in D minor, op.30

Media Review / Comparison

2015-01-11 — Original posting
2015-01-22 — Added larger-scale comparison list & YouTube links
2015-07-04 — Added link for music score
2016-07-31 — Brushed up for better readability
2017-03-24 — Added link to extra YouTube video
2019-06-29 — Added Wild / Horenstein to YouTube list


Introduction  / The Recordings

This posting is about Rachmaninoff’s Piano Concerto No.3. I currently have the following recordings, shown and discussed in chronological order:

The first two are also present in my LP collection, the others were added more recently, and as CD only — see below for details.

Background, About the Composition:

Sergei Rachmaninoff (1873 – 1943) completed his Piano Concerto No.3 in D minor, op.30, in 1909 in Moscow, before departing for an engagement in the United States. There, he gave the first performance under the direction of Walter Damrosch (1862 – 1950), followed immediately by another performance, in which Gustav Mahler (1860 – 1911) was conducting.

The Movements:

The concerto features three movements, superficially fitting the scheme fast — slow — fast, but the middle movement is full of virtuosic & fast passages / sections. Conversely, the last movement also has slow(er) sections. I won’t discuss the movements in detail. Instead, I will just outline the structure as seen in a superficial view in the score, trying to give an impression on the tempo changes in Rachmaninoff’s score (excluding rall. / accel. / a tempo for very short segments). Note that interestingly, Rachmaninoff has not added any pedaling annotations!

I. Allegro ma non tanto

The concerto starts with the well-known main theme, played by the piano:

Rachmaninoff: Piano Concerto No.3 D minor, op.30: mvt.1, score sample

Allegro ma non tanto — a tempoPiù mosso2Più vivo — 34 AllegroModerato — Allargando5 a tempo … 8Allargandoa tempo9Allegro10 — Tempo precedente11 AllegroTempo I (theme) — 12 Più mosso13Più vivo14AllegroAlla Breve, Allegro molto1518 poco a poco ritenutoAllegro molto, Cadenza19 Meno mossoModerato, CadenzaModerato20Tempo I (theme) — 2122Un poco più mosso23a tempopoco accelerando al Fine.

The above theme returns in the orchestra several times, the piano plays it just two more times: once after 11, softened, modulating to B♭ major, and then again after 20, in its initial form.

II. Intermezzo: Adagio

Adagio2425 Un poco più mossoPiù mosso — 26 Meno mosso — (short cadenza) — a tempo, più mosso27 Meno mossoPiù vivo28 Meno mosso (Adagio) — a tempo, più mosso2932 Tempo come primaPoco più mosso3338 Un poco più mossoL’istesso tempoattacca:

III. Finale: Alla breve

Alla breve (1/2 = 1/4)4042Più mosso4345 Meno mosso46Allegro molto47Scherzando (1/4 = 1/2)4851 Più vivo52Meno mosso53 a tempo54 a tempo55 Lento56a tempo come prima5759 Tempo I, Alle breve (1/2 = 1/4)60Più vivo6164Più mosso6566Più vivo67 … 69 Vivace7071 poco a poco accelerando73Cadenza74 VivacissimoUn poco meno mossoTempo precedente75 Un poco meno mosso7677 Più vivoa tempo79 Più vivoPresto.

The Interpretations, Overview

In order to provide a rating overview, as well as an idea about tempo relations both within an interpretation, as well as between the two recordings, I have prepared the little table below. Note that

  • The color coding for the tempo (blue = slower / longer, green = faster / shorter) refers to the average between the recordings.
  • The movements are full of tempo changes. Measuring all relevant tempi would be a large amount of work; instead, I have just measured the tempo for the first segment in each movement.
  • To complement this limited view, I have also listed the durations of each movement.
  • The times for Rachmaninoff’s interpretation have been adjusted for the omitted segments — by movement:
    • Allegro ma non tanto: 8 bars of the section “Tempo precedente” (prior to 11) — actual track duration 13’57”;
    • Intermezzo: Adagio: 6 bars (Più vivo) up to 28, plus the following 7 bars, annotated meno mosso (Adagio) — actual track duration: 8’40”;
    • Finale, Alla breve: from 45 (Meno mosso) through 46, up to the Allegro molto (29 bars); also, from the Meno mosso after 52 up to the a tempo at 54 (13 bars) — actual track duration: 11’22”.
  • Applause (Argerich, Wang) and dead time (Ashkenazy) were subtracted from the track durations for the last movement.
  • The ratings are subjective and relative, as usual (I can’t exclude that there are versions out there that I would rate higher than those with a rating of 5).
  • I can’t really rate Rachmaninoff’s own interpretation. He is a standard of his own. I still gave ratings, to provide an idea about the listening experience, rather than for the obvious historic / reference value of this recording.
Rachmaninoff, Piano Concerto No.3 D minor op.30 — comparison table

The Interpretations, Detail

The recordings below are presented in chronological sequence. Note that — as outlined above, Rachmaninoff plays the concerto with several cuts; all other recordings feature the concerto without these cuts.

Sergej Rachmaninoff, Eugene Ormandy, Philadelphia Orchestra (1940)

Rachmaninoff: Piano Concertos, Rhapsody — Rachmaninoff; CD cover

Rachmaninoff: Piano Concertos 1 – 4

Sergej Rachmaninoff, Eugene Ormandy, Philadelphia Orchestra

Masters of the Piano, CDO 3002 (2 CDs, mono); © 2001
Booklet: 4 pp., English

Rachmaninoff: Piano Concertos, Rhapsody — Rachmaninoff; CD, UPC-A barcode
amazon media link

Artists and Recording

We are lucky to have recordings of the composer’s own interpretation of all his piano concertos. He recorded concerto No.2 in 1929 (with Leopold Stokowski), the other concertos he did in 1940 and 1941, Concerto No.3 dates from 1940 (keep in mind that the composer was 67 by that time!). It’s mono only, of course, and there are limitations in how one could capture the sound of the orchestra. But this still gives us an idea of how the composer played his own works.

Still, one should keep in mind that the dynamics (in the piano part, as well as in the orchestra) may not be captured appropriately. Also, I don’t know whether the engineers had the ability to make corrections in the recording. I think we should take this as a live concert recording. Rachmaninoff’s recording is the shortest by a long stretch, though most of this is caused by skipping segments in each of the movements.

General Remarks

Still, Rachmaninoff’s playing is about as fast as technically (and humanly) possible. I don’t think we know whether the composer wanted to prove how well and fast he could still play his concerto, and/or whether he wanted to “set a standard” a benchmark, with the intent that other pianists ought to look at his interpretation as the only valid model interpretation of his work? I personally don’t think pianists should aim to duplicate his interpretation. After all, the composer had his own, personal skill set (just take the enormous span of his hands, covering 1.5 octaves!); plus, from listening through the 5 recordings here, I get the impression that this concerto allows for an exceptionally wide range of personal views. Many of these will make perfect sense to today’s audiences: there isn’t a single, “true” / “correct” interpretation.

At least for the piano sound, this is a very decent recording, given its 75 years of age! It’s amazing to see that Rachmaninoff plays this vastly faster than Ashkenazy — faster also than Martha Argerich (except for the last movement) and Yuja Wang! One limitation in the recording is that it does not reveal the pianist’s dynamic span and relations. It is possible that while Rachmaninoff had no technical issues with the trickiest of passages, but it seems that he used a limited dynamic span, rarely going below mf; also, the orchestra sometimes appears “just loud”. Its sound of course is not spatially resolved from the piano.

Notes on the Movements

I. Allegro ma non tanto

Duration: 13’57”
The beginning in the orchestra is fast. Rachmaninoff plays the main theme very much leaning forward, or is this just a matter of orchestra and pianist getting “in sync” / “in tune” with each other? This is only in the beginning, but he later does use a fair amount of (sometimes heavy) rubato, especially on a very short scale (a large amount of rhythmic freedom within a bar, where each beat may get its own weight & time). Some of this may be attributed to Zeitgeist, but one can still deduce that the notation only shows the skeleton of the composition. One can’t criticize this interpretation, of course, but one can for instance see how much he meant to have the (often partially hidden) melody exposed within a complex texture, and how much rubato is acceptable in the composer’s view.

Of course, the composer is free to alter the text as he sees fit. There are a few places where he might have deviated from the score (I did not try registering those). He certainly left out the 8 bars marked Tempo precedente, prior to 12. Also, his playing is exceptionally clear and transparent in general. He does not try obscuring any detail by excess use of the pedal. Rachmaninoff plays the standard cadenza. I’m not sure whether the alternate version would have enforced a slower tempo? The Alla breve section (starting between 14 and 15) as well as most of the cadenza are very fast, indeed! There are lyrical moments as well, of course, but less than in other interpretations. Certainly, he is far from Ashkenazy‘s romanticism!

II. Intermezzo: Adagio

Duration: 8’40”
One can certainly feel some Zeitgeist in the orchestra playing, with all the portamento in the strings and the excess vibrato, but thanks to the relatively “fast” tempo (by far the fastest in this comparison), the music does not sound overblown. Here we can hear what Rachmaninoff meant with cantabile, and what a good tempo for the melody is (and we are again reminded of the fact that Adagio means calm, not slow)! Rachmaninoff skips the 6 bars (Più vivo) up to 28, plus the following 7 bars, annotated meno mosso (Adagio). It is interesting to see that the composer makes fluent transitions between duplets and triplets in the right hand, after 30 (the left hand constantly plays triplets here). In the very virtuosic Poco più mosso section, the sound is concentrating on the piano, the dialog with the wind instruments is hard to follow.

III. Finale: Alla breve

Duration: 11’22”
The initial tempo is not exceedingly fast, but Rachmaninoff catches up in the second half of the movement: the composer’s playing is light, playful. Occasionally it is on the border of being a bit superficial in the articulation (maybe that’s caused by the excruciating technical demands in this concerto?). Rachmaninoff skips the 29 bars between 45 (Meno mosso), through 46, up to the Allegro molto. He further skips 13 bars from the Meno mosso, located two bars after 52, up to 54 (a tempo).

Overall Duration: 33’57
Rating (see above for details): 3.3 — Even though I don’t think this should be taken as benchmark, or as an exact model for an “ideal” interpretation, I still think this is a “must have” for people seriously interested in this concerto.

Vladimir Ashkenazy, André Previn, London Symphony Orchestra (1972)

Rachmaninoff: Piano concertos 1 & 3 — Ashkenazy, Previn; CD cover

Rachmaninoff: Piano Concerto Nos.1 & 3

Vladimir Ashkenazy, André Previn, London Symphony Orchestra

Decca / eloquence 460 487-2 (CD, stereo); ℗ 1972
Booklet: 2 pp., track listing only

Rachmaninoff: Piano concertos 1 & 3 — Ashkenazy, Previn; CD, UPC-A barcode
amazon media link

Artists and Recording

This is part of the complete recording of all four Rachmaninoff concertos that Vladimir Ashkenazy (*1937) realized in 1971/72, under the direction of André Previn (1929 – 2019). I already had this as an LP box — this re-release is split onto two separate CDs. In this concerto, Ashkenazy takes notoriously slower tempi than all others: his interpretation takes 46 minutes, compared to the 40 minutes specified in the score (and also the average of all performances in this post), let alone the 34 minutes in Rachmaninoff’s own interpretation.

Notes on the Movements

I. Allegro ma non tanto

Duration: 18’45”
Let me start with the positive comments first: Ashkenazy plays carefully, his playing is fluent. And he is the only pianist in this comparison who dares playing the more difficult, alternate (“ossia“) cadenza. It is possible that this choice imposed a limit to the tempo selection (who am I to judge whether that alternate cadenza is technically playable at the tempo chosen by, e.g., Rachmaninoff himself?). Ashkenazy’s tempi are substantially below those in any of the other recordings. But the same holds true for the other movements, so I doubt that the cadenza was the only decisive factor.

The main impressions from this recording are: this is a very (exceedingly) lyrical, soft, romantic interpretation. It is far from Rachmaninoff’s own, much drier, more technical view. The Allegro ma non tanto feels rather like an Andante. In direct comparisons, the interpretation sounds almost boring, and sometimes (e.g., in the Più mosso section), the piano may feel like dragging behind. I also think that Ashkenazy rarely really reaches p and pp in his part. The dynamic scope appears limited to mf .. ff.

II. Intermezzo: Adagio

Duration: 12’08”
To me, this is very slow, far too romantic & overblown, with lots of vibrato in the strings, even with some wind instruments, such as the oboes. The music from the orchestra feels like damp, hot air in a tropic summer night. From listening to the composer’s own interpretation, I don’t think this can possibly be what Rachmaninoff had in mind. At this tempo, the melody (e.g., in the violins, but also in the solo part) sounds vastly overstretched, certainly not fitting the cantabile annotation. Ashkenazy plays the solo with lots of warmth, with big gestures; the fast part is clear — at this tempo, it can’t be called cold, extroverted virtuosity, for sure!

III. Finale: Alla breve

Duration: 15’05”
This is maybe Ashkenazy’s best movement in this concerto, carefully played, not overblown, but also not light, the tempo comparatively slow in general (see the composer’s tempo!). To me, the main problem with this tempo & interpretation is that it makes this movement sound like a real earworm. Is the interpretation too uniform? In the rhythmically intricate passage after 44, unintended (?) acceleration leads to apparent, constant syncopation, the accompaniment by the solo flute is barely audible. The section 50 is much slower than Rachmaninoff’s interpretation. In the bars preceding 54, there are serious coordination issues with the woodwinds (oboes, clarinets).

Overall, I think that this interpretation of the last movement does not differentiate it enough from the middle movement (which also features plenty of rapid passagework!).

Overall Duration: 45’57”
Rating (see above for details): 2.7 — A carefully played, but exceedingly romantic, heavy, often overblown interpretation. The main reason I can see why people might want to look at this is the alternate cadenza in the first movement. However, there are tons of recordings around with this cadenza!.

Martha Argerich, Riccardo Chailly, RSO Berlin (1982)

Rachmaninoff: Piano concerto No.3; Tchaikovsky: Piano concerto no.1 — Argerich; CD cover

Rachmaninoff: Piano concerto Nr.3 in D minor, op.30
Tchaikovsky: Piano concerto Nr.1 in B♭ minor, op.23

Martha ArgerichRiccardo Chailly, RSO Berlin
Kirill Kondrashin, Symphonie-Orchester des Bayerischen Rundfunks (Tchaikovsky)

Philips 446 673-2 (CD, stereo); ℗ 1982, 1995
Booklet: 2 pp., English

Rachmaninoff: Piano concerto No.3; Tchaikovsky: Piano concerto no.1 — Argerich; CD, UPC-A barcode
amazon media link

Artists and Recording

This is one of the spectacular, early live recordings with Martha Argerich (*1941) that sparked the artist’s successful, worldwide career. The concert with the Rachmaninoff concerto took place in December 1982 in Berlin (the Tchaikovsky concerto was recorded in February 1980).

Notes on the Movements

I. Allegro ma non tanto

Duration: 15’46”
Compared to Ashkenazy’s (studio) recording, this one has more clarity and transparency, both in the soloist’s playing, as well as from the recording technique, i.e., concerning the orchestra and the balance between orchestra and piano. Martha Argerich’s playing is more eruptive, and she is using vastly more short-scale (intra-bar) agogics. Her playing is also lighter, more vivid than Ashkenazy‘s (as if she was using less pedal). She is observing the dynamic annotations much more closely. The “standard” cadenza is a better fit to her interpretation than the heavier alternate. the Moderato (orchestra, middle of section 4) is rather heavy. In section 6, the a tempo (solo) isn’t really, but then, that section is also marked espressivo, later dolce. Hence, Argerich’s agogic freedom is certainly legitimate.

II. Intermezzo: Adagio

Duration: 11’00”
Better tempo than in Ashkenazy / Previn‘s interpretation, but (for me) the orchestra still sounds extremely romantic (with lots of vibrato, also), reminding me of film music. However, the piano part is certainly excellent, very expansive, very expressive (e.g., in the bars preceding 27), with a fair amount of extroverted virtuosity, but also with really big gestures. In the Più vivo section (after 27), some of the triplets in the right hand are “washed out”, approaching duplets, and in the following, virtuosic Poco più mosso section, Argerich puts a lot of (dynamic) focus on the virtuosic “nonuplets” (27 semiquaver notes per 3/4 bar) section. Especially the bass notes in the left hand are hardly audible: to me a sign of showing off the virtuosic aspect. The music would have profited from a dynamically more balanced approach.

III. Finale: Alla breve

Duration: 13’58”, or 13’21” without the applause
A spectacular interpretation with olympic qualities, beating anybody else in mere speed, with Martha Argerich storming forward almost throughout the movement. The recording technique has its limitations here: the orchestra is sometimes almost covered by the piano. There are some extreme passages where Argerich appears focused on speed, showing / focusing on the outline of the solo part only, secondary voices in the solo part are barely discernible. The transition to Più vivo after 60 isn’t quite harmonic. But that may be attributable to the live recording. Yes, it is spectacular — but barely a model or reference performance, I think.

Overall Duration: 40’44”
Rating (see above for details): 4.3 — A famous & spectacular recording. Though, I think some of the fame stems from the artist intending to show off her abilities, aiming at the show effect, proving her outstanding virtuosic abilities.

Zoltán Kocsis, Edo de Waart, San Francisco Symphony (1983)

Rachmaninoff: Piano Concertos, Rhapsody — Kocsis, de Waart; CD cover

Rachmaninoff: Complete Work for Piano and Orchestra

Zoltán Kocsis, Edo de Waart, San Francisco Symphony

Philips 438 383-2 (2 CD / download only, stereo); ℗ 1983
Booklet: none, track information available from Deutsche Grammophon

Rachmaninoff: Piano Concertos, Rhapsody — Kocsis, de Waart; CD, EAN-13 barcode
amazon media link

Artists and Recording

Several times, a friend and former colleague of mine was raving about Zoltán Kocsis (1952 – 2016) playing, calling it “the very best (spectacular?) Rach 3 that he had ever heard”. I could not resist and was looking around for that recording. It turned out that the hardcopy CDs of Kocsis’ recording with all of Rachmaninoff’s works for piano and orchestra (realized between 1982 and 1984) is no longer available. Fortunately, Presto classical and Deutsche Grammophon still offer this as download.

So, here it is: the Rachmaninoff Concerto No.3 was recorded in October 1983. The recording location is not given. There is some sporadic coughing / “audience noise” in the second movement, but the track information from Deutsche Grammophon claims it’s a studio recording (apart from this concerto, all tracks in this set are described as being studio recordings made at Davies Symphony Hall in San Francisco). This was probably done at the ideal time in Kocsis’  career as a pianist. As far as I could see, he — sadly! — now is mostly conducting.

General Remarks

The sound of the orchestra is a bit muffled, certainly not nearly as well-captured as with Wang / Dudamel, but Edo de Waart’s accompaniment in general is very good (although the listener’s attention is typically entirely focused on the solo part!). My colleague may or may not be right — this recording definitely earns one superlative: it is faster than all others, even beating Rachmaninoff himself! Kocsis does not offer Yuja Wang‘s crystal-clear transparency. He nicely brings out main melody lines in Rachmaninoff’s complex piano texture. The accompaniment remains relatively compact, sounds legato. Kocsis  typically does not focus on aspects of polyphony by pointing out secondary melodies / middle voices (certainly not to the degree seen with Yuja Wang).

Notes on the Movements

I. Allegro ma non tanto

Duration: 13’55”
I think the tempo is slightly too fast, given that the annotation is Allegro ma non troppo: I feel pushed throughout the initial part with the main theme (even Rachmaninoff’s tempo is at the limit, considering the initial tempo coordination issues). There is enough opportunity to show fast playing throughout the rest of the movement. That initial “pushed” feeling settles after the first part, and I feel at ease in the Più mosso part and beyond. Here, Kocsis’ playing is not hammering, very fluent, very harmonic, rounded, controlled, full-sounding. Where he gets an opportunity (e.g., in 5), Edo de Waart lets the romantic feeling evolve, flourish. But for the most part, of course, the composition is dominated by the piano part.

A nice detail: in the Tempo I section after 11, the main theme returns in the piano, this time modulating to a major chord (B♭). Kocsis plays this with the same urge as the initial instance, but this time giving it a wonderfully soft touch. The fast tempo then resumes, but thanks to Kocsis’ astounding technical abilities, the music does not feel rushed, remains fully controlled. Kocsis plays the standard cadenza with amazing agility. Thereafter, the initial theme returns, this time mirroring the first instance. I would rate Kocsis the highest for the cadenza — overall, also Martha Argerich and Yuja Wang have definitive, unique merits — hence I’m giving all thee interpretations a top rating for this movement.

II. Intermezzo: Adagio

Duration: 9’56”
The movement starts a tad faster than Argerich‘s interpretation (but clearly not as fast as the composer’s own version), but equally romantic (lots of vibrato in the strings!), just about bearable in the Un poco più mosso at 25, but rarely overblown. Compared to Yuja Wang, Kocsis appears to use more pedal, the piano sound is softer, dense, rather romantic than extroverted. The sudden downfall, six bars after 28, leading into the a tempo, più mosso (ff) is rhythmically superficial, the demisemiquavers on the first two beats are hardly audible (unnecessarily over-punctuated). In the virtuosic Poco più mosso section after 32, Kocsis isn’t trying to show off, the piano part is far more harmonic and balanced. It’s excellent, though not quite as clear as with Yuja Wang.

III. Finale: Alla breve

Duration: 13’28”
Excellent, as the other movements, with a recording technique providing much better balance between piano and orchestra. The transition to the Più vivo after 60 feels more natural than with Argerich, but still might be prepared a little better in the orchestra? In general, the tempo feels more natural than with Argerich / Chailly, the piano part feels more balanced / equilibrated.

Overall Duration: 37’17”
Rating (see above for details): 4.7 — My colleague & friend was right: a spectacular version, among the very top. There aren’t many which will beat this. It’s hard to believe that this recording doesn’t have more visibility / better accessibility in the music market!

Yuja Wang, Gustavo Dudamel, Simón Bolívar Symphony Orchestra of Venezuela (2013)

Rachmaninoff: Piano concerto #3; Prokofiev: Piano concerto #2 — Wang/Dudamel; CD cover

Rachmaninoff: Piano concerto Nr.3 in D minor, op.30
Prokofiev: Piano concerto Nr.2 in G minor, op.16

Yuja Wang, Gustavo Dudamel, Simón Bolívar Symphony Orchestra of Venezuela

DG 479 1304 (CD, stereo); ℗ / © 2013
Booklet: 22 pp. en/de/fr/es

Rachmaninoff: Piano concerto #3; Prokofiev: Piano concerto #2 — Wang/Dudamel; CD, UPC-A barcode
amazon media link

Artists and Recording

This is a recent live recording with Yuja Wang (*1987), realized in Caracas, in February 2013, under the direction of Gustavo Dudamel. Ever since her (still short) career really took off, technical “monsters” such as the concertos by Rachmaninoff and Prokofiev have been the core of Yuja Wang’s repertoire. Her technical abilities are fantastic! I like her interpretation of Prokofiev’s Piano Concerto No.2. Let’s see how she performs with the Rach 3 on that same CD. As Martha Argerich, Yuja plays the “standard” cadenza — and again, it is definitely the better fit for her view of the concerto. I’m sure she would have no problem playing the “ossiacadenza!. And her playing is again astounding!

Notes on the Movements

I. Allegro ma non tanto

Duration: 15’50”
Yuja Wang’s tempo is very close to Argerich‘s, though she uses a little less agogics, her playing is very light in the beginning, almost understating, almost as light as a Scarlatti sonata (!), up to 4, extremely clear, simple-sounding. In the crescendo and ff bars leading into the Moderato section, though, she demonstrates that she also can have a “steely” side, without ever losing her astounding agility. But she also has a strong lyrical side, e.g., in section 6, her playing never loses transparency — even lucidity. The latter is true also for her ff playing in 8 (Allargando — a tempo). In 12, she nicely brings out the “inner” melody line in the left hand, while keeping the right hand light and clear, and before 17 she takes the volume back almost to ppp.

Let me try characterizing her interpretation in this movement, in general terms: it is very different from Ashkenazy, it is not eruptive and uses less rubato / agogics than Argerich‘s interpretation. It does not try showing off virtuosity, while at the same time it is not weak, nor does it appear too smooth. Within the recordings I’m discussing, Yuja’s version is probably the most straight, internally coherent one, featuring both clarity and simplicity. It completely hides the tremendous technical challenges in this concert. Frankly: I don’t know what to criticize here. If you are looking for virtuosity / showing off, if you are looking for “keyboard thunderers”, if you want the alternate cadenza, or if you are looking for a romantic, “soft” interpretation, that’s definitely not the right one. But for me, it is an excellent and absolutely valid interpretation!

II. Intermezzo: Adagio

Duration: 10’38”
The tempo is between that of Ashkenazy and that of Kocsis. The interpretation definitely slimmer, far less overblown, more lyrical than (overly) romantic in general. OK, one can’t avoid romanticism in the Un poco più mosso at 25! The almost “Scarlatti-like” clarity and the dynamic balance in the piano part are exceptional, as is Yuja Wang’s ability to expose the inner polyphony in Rachmaninoff’s dense piano texture. Her playing is near-perfect. To me, this movement is also more cantabile than any other interpretation, in the true sense of the word. All this also applies to the fast part, starting at the Poco più mosso after 32: it’s not just very virtuosic, but remains clear, transparent, controlled, giving the left hand its proper visibility, too.

I could repeat my statements from the previous movement (and the orchestra part fits right into this picture!): I don’t see any “hooks” for criticism (there is a little mishap in the “velocecadenza, but that is negligible, given that this is a live recording) — excellent!

III. Finale: Alla breve

Duration: 14’20”, or 13’46” without the applause
Definitely excellent & truly outstanding: Yuja adds lightness and lyrical aspects to this movement. She dares lowering the volume to pp and beyond, where appropriate! The Più mosso section at 43 is excellent rhythmically. Unfortunately, one can’t really hear the flute solo (though that is marked ad libitum, so it may have been omitted). A nice detail in the orchestra: the brass sound in the ritardando leading into the Scherzando (prior to 48) is exceptional. In Yuja’s part, the downward sextuplet run in the left hand, 6 bars after 51 is truly stunning and beats all other interpretations! Also, the tempo change to the Più vivo after 60 feels more natural than in the other interpretations.

Overall Duration: 40’47”
Rating (see above for details): 5.0 — I was anxious to see how Yuja Wang would perform in this concerto, even though I have heard enough of her playing to know that she is far more than market hype. My fears were proven wrong: Yuja wins this almost hands down, with competition only from Zoltán Kocsis, though hers is a live recording, which makes this even more spectacular. Her abilities in light & transparent playing in the most difficult music, while at the same time being able to bring out the lyrical aspects without obscuring the texture of the piano score are unmatched, as far as I can see.

Addendum 1: Score

For this work I have used the pocket score by Boosey & Hawkes —Find pocket score on (#ad) —

Addendum 2: Tables

In the comments, the question of alternative recordings turned up, such as Van Cliburn’s legendary 1958 recording in Moscow, I started looking for such “other” recordings on YouTube. Quickly realized that one can easily get lost even just with this one concerto. I have located some 38 YouTube recordings, thereafter I gave up. Let me present the list of timings here, with an indication whether they play the “standard” cadenza in the first movement, or rather the original one, now marked with “ossia“.

Timing Table

Rachmaninoff, Piano Concerto No.3 d minor op.30 — YouTube comparison table
Rachmaninoff, Piano Concerto No.3 d minor op.30 — YouTube comparison table

Notes on the Above Table

I did not try rating these recordings — in fact, I only listened to a few of them at full length. An in-depth discussion (such as for the recordings above) would be a huge amount of work, plus, it would be unfair, given the very poor sound quality in some of the older recordings (Gilels’ 1949 recording is certainly affected here!). Some notes on the table first:

  • I sorted the table by year. Within the same year, by the artist (first name) — for a list sorted by the pianist’s last name see below.
  • In the column “Source“, “CD” indicates recordings in my collection — those discussed above. “YouTube” is obvious — but not all these are videos: some use stills and are transfers from LPs or radio recordings, etc.
  • The column “mvt. I cadenza” is obvious — “standard” is the now “regular” cadenza, “ossia” is the alternative, actually the original cadenza.
  • The color coding is the same as in the table above, except that the average timing here is based on all 38 recordings.
  • In the entries labeled “CD”, the timings are taken from the table above, with the same corrections (no applause, Rachmaninoff’s timing corrected for the skipped sections).
  • I have not tried making a selection of “valuable” recordings. These are all across the board, from the 15-year old Dimitris Sgouros to students, to performances at competitions to “professionals” to “super-virtuosos” to “post-virtuosos”, old and new.
  • There appears to be an inflation of recordings over the recent years. That is mainly because competitions now are often to be recorded, and such recordings tend to be posted on YouTube, as welcome (self-)promotion

Interactive Timing Table

This is an interactive table of the recordings, initially sorted by the Pianist’s last name, with links pointing to the YouTube recording (where available). Note that for the CD recordings above, the YouTube recordings may differ from the CD. With some artists, there are several recordings — if so, I have preferred postings with the complete Rach 3 in one video posting, even if that meant stills only rather than video (e.g.: with Lang Lang, there is a video that lacks part of the second movement). If I have forgotten key recordings, please let me know — both tables can be amended.

Last update: 2020-12-27
Andsnes, Leif OveBringuier, Lionelossia2009
Argerich, MarthaChailly, Riccardostandard1982
Ashkenazy, VladimirPrevin, Andréossia1972
Bachauer, GinaSherman, Alecstandard1957
Bachus, Brett???ossia1992
Berezovsky, BorisInbal, Eliahuossia1991
Bronfman, YefimGergiev, Valeryossia2012
Buniatishvili, KhatiaJärvi, Neemeossia2011video no longer available
Charles, JoshuaNeely, Davidossia2009
Cliburn, VanKondrashin, Kirillossia1958video no longer available
Gavrilov, AndreiLazarov, Zdravkoossia1976
Geniet, RémiAlsop, Marinstandard2013video no longer available
Gerl, JaminaHostetter, Paulossia2013
Gilels, EmilKondrashin, Kirillstandard1949
Giltburg, BorisAlsop, Marinossia2013
Horowitz, VladimirMehta, Zubinstandard1977
Indjic, EugenAltrichter, Petrstandard2011
Kern, OlgaConlon, Jamesossia2001
Khozyainov, NikolayBlendulf, Danielossia2017
Khozyainov, NikolayParisotto, Marcoossia2014video no longer available
Kissin, EvgenyOzawa, Seijiossia2012
Kocsis, ZoltánWaart, Edo destandard1983
Lang, LangTemirkanov, Yuriossia2013
Larrocha, Alicia dePrevin, Andréstandard1971
Lugansky, NikolaiAbbado, Robertostandard1998movements #2 and #3
Matsuev, DenisSlatkin, Leonardossia2013Rach 3 @ 33'43"; video no longer available
Mechetina, EkaterinaNeuhold, Günterossia2013
Nakamura, HirokoKondrashin, Kirillstandard1980
Oborin, LevStokowski, Leopoldstandard1961
Ozolins, ArthurJansons, Marissstandard1982
Paik, Hae-SunPletnev, Mikhailossia2007
Prats, Jorge LuisJärvi, Paavoossia2013
Rachmaninoff, SergeiOrmandy, Eugenestandard1940
Sgouros, DimitrisLópez-Cobos, Jesúsossia1984
Sokolov, GrigoryKitayenko, Dmitriossia1978
Trifonov, DaniilDouglas, Barryossia2013video no longer available
Volodos, ArcadiPletnev, Mikhailossia2007video no longer available
Wang, YujaDudamel, Gustavostandard2013
Weissenberg, AlexisOzawa, Seijistandard1978
Wild, EarlHorenstein, Jaschastandard1966

If you have an opinion on this posting / page, please rate it (1 = dislike, 5 = best)

Average rating: / 5. Vote count:

Be the first to rate.

We are sorry that this post was not useful for you!

Let us improve this post!

Tell us how we can improve this post?

AboutImpressum, LegalSite Policy | TestimonialsAcknowledgementsBlog Timeline
Typography, ConventionsWordPress Setup | Resources, ToolsPictures, Methods

27 thoughts on “Rachmaninoff: Piano Concerto No.3 in D minor, op.30”

  1. I’ve had the Rachmaninoff discs for fifty years and always found the performances disappointing. He seems to me to be “mailing it in,” doing his duty as a twentieth-century composer to leave a recording of his work by his own hands, and not much more.

    I’m also been aware of the Ashkenazy version for some time, prefer it to the composer’s own. But the only version that’s grabbed me is Cliburn’s in Carnegie Hall. I also heard him perform it live at an outdoor summer concert in New York. Both experiences gave me goosebumps and continued to do so for years after.

    Having said all this, I think I agree with your own assessment of these four performances, at least as far as the ranking goes. I sometimes wonder what Avdeeva would do with Rachmaninoff, but I don’t feel any urgency about finding out. I think there is other music I’d rather hear her tackle first.

    Thank you for this, Rolf.

  2. Thanks for your comment, Thomas!
    I think that the way Rachmaninoff played & recorded his own concertos is a question of personality and attitude — whatever these were, the recordings still give a rough idea about how he saw and characterized his own composition, so they remain a useful legacy (had he been worried about leaving behind a definitive / authoritative testament / documentation, he would probably have played the concerto without the cuts… ).
    How would you characterize Van Cliburn’s interpretation in relation to the ones discussed above? I suspect also Gilels’ interpretation would be interesting to hear (that must be a mono recording). Horowitz’ technical abilities were phenomenal, but in view of recent recordings I think he has been (and still is) overhyped: I’m sure he was brilliant in the virtuosic sections, but I don’t really see him addressing the lyrical aspects in this composition (his fan community might beat me up for this)…

  3. If I say that Cliburn’s is the one for Avdeeva to “beat,” and that only she is capable of doing so among living pianists I am aware of, I think you will understand what I mean.

    I confess to not liking speed, especially if it smacks of showing off, as it does for me in the case of Horowitz. I suspect that composers play their own work too fast simply because they know it so well and they lose track of what it sounds like to other ears. I suppose the same could be true of some performers, e.g. Kocsis, but then it surely must be a conscious decision on their part.

    Ultimately, as you well know, it’s all about the experience we have listening to the performance if we really are there for the music and not just for the acrobatics or the technique. And Cliburn, like Avdeeva, achieves what I could only call, half-facetiously, a religious experience for me. Those few chords toward the end of the piece sum up everything else in it (Rachmaninoff has a similar moment in all his concertos). And that moment isn’t just musical, it’s the sort of thing that make great art great art.

    I listened to Cliburn’s Moscow performance of the Third last night on YouTube (my turntable being on the fritz)–probably the one following his winning the competition (wasn’t Gilels the judge who asked Kruschev if it was alright to award the first prize to an American?). It’s not as magical as the Carnegie Hall performance, but it’s close.

    • Thomas, based on your input, I have started looking at other recordings on YouTube, but soon — as to be expected — got lost pretty soon! Still, I have added a more extensive list of recordings without discussion, and I’m providing video links to 38 recordings.
      Yes, Van Cliburn is impressive, I also listened to Gilels (sadly very muffled in sound), other recordings that I listened to from beginning to end include Gavrilov (interesting, good!), Horowitz (horrible), Sgouros (amazing for a 15-year old), Kern (not all that bad), Charles (OK, that’s a student), Buniatishvili (can’t hold the water against many others). Others that are likely interesting include Bachauer, Oborin, Larrocha, Weissenberg (horrible sound), Sokolov, Berezovsky, Luganski, Volodos, Andsnes, Indjic, Kissin, Bronfman, Trifonov, Matsuev (plays Rach 2 & 3 in the same concert!), …
      But taste is subjective, and even just by mentioning names one may start a controversy!
      Last note: when comparing recordings with different cadenzas (standard vs. “ossia”), one should probably try *not* considering the cadenza …

  4. Thanks again, Thomas — I’ll have a listen soon (didn’t even think of checking YouTube, sorry!) — may take me a couple days, though…
    I see that there is a second video from 1966, and there’s Gilels, though from 1949. One problem I see forthcoming is that such old recordings can hardly capture the full detail of the piano part (even if the solo instrument is very much “in the front”), in particular: if the recording sounds muffled, it will be hard to judge a pianist’s clarity, which is what I particularly like about Yuja Wang’s playing. I had a quick listen into the first movement in both Gilels’ and Cliburn’s interpretations — and the sound is softened considerably by the recording, making it hard to hear legato vs. non-legato in faster passages. But I can definitely say that both these versions provide a much more satisfactory experience than the composer himself (or Ashkenazy, I think)! I hope to get back to this next week. Interesting: Cliburn plays the alternate (= original) cadenza, Gilels the (now) standard one…

  5. There is another recording with V. Ashkenazy, E. Ormandy and the Philadelphia Orchestra. Recorded in 1975 or 1976, digitally remastered and published on CD RCA Papillon Collection in 1987.

    • Thanks a lot for your input, Eugenia — I have added this recording to the lists — it’s an interesting, very clear performance!

  6. You’re welcome Rolf. However, the link you provided for the update is another pianist’s performance. Here is the correct link:
    And thank you for this very interesting review on one of my favorite piano concertos.

  7. Ah — thank you so much for the correction! I have of course listened to the correct concerto version, in order to take the times for the table—but then, when it came to take the URL, I must have waited beyond the end of the video, and the YouTube list already switched to Ekaterina Mechetina’s recording…

  8. Excited to find this, my favorite, many thanks Rolf for putting it together! I saw Behzod Abduraimov play the Rach 3 with the ossia cadenza this last spring in San Diego. The symphony accompaniment and the hall acoustics did not do his talent justice unfortunately, but I had a great seat up front to see his hands. I am curious if you would kindly share an update as to which you find to be the best ossia cadenza performance to-date?

  9. Hi Susannah,
    I’m glad you like my review (there are many visitors, but I rarely get feedback!). As to your question: unfortunately, I have not evaluated and rated the cadenza separately, and I did not want to rate YouTube videos of questionable quality. Furthermore, I have since picked up concert reviewing and am very busy with that, so I won’t be able to pick this up again in the near future. Among those reviewed so far, Argerich, Kocsis and Wang — all selected the standard cadenza. I do have, though, two more recordings that I have not reviewed yet, and in both cases (Denis Matsuev, 2009 and Leif-Ove Andsnes, also 2009) the soloists selected the ossia cadenza (at least in the YouTube recording, haven’t checked the CDs yet). My expectations are very high in both cases, even though the recordings date almost 10 years back. My first bet would be Matsuev at this point, though I was also very impressed with Andsnes in a recent concert performance (not Rach 3, not even Rachmaninov, though). As a little compensation for the lack of proper evaluation, I have added a column “Cadenza” to the above, interactive table, as the spreadsheet with the timings is hardly very helpful in this. The judgement on the “best ossia performance” also depends on personal preferences, i.e., how do you rate power and / vs. clarity? clarity of the orchestra / recording? Are you looking for a “particularly Russian” performance? etc.
    Best wishes,

  10. Much appreciated Rolf! I guess I go firstly for power (to me the orchestra is more fluff to a great pianist) and have difficulty with poor recordings – or the people who sneeze in the middle of the performance… but that is another issue 😉 I love emotional music that moves me (sadness, anger/excitement, or passion). Lived in Rome for a while right next to La Scala in Milano and before in Rome on a piazza of a famous violin and cello repairman so heard great artists playing their instruments with such talent (even when they felt nobody listening). I’ve been looking for the Rachmaninoff himself recording to hear his interpretation (even if shortened). Good luck with your new concert endeavours; sounds fun!

  11. What would you think of putting Earl Wild on your Rachmaninoff 3 list? Royal Philharmonic Orchestra,with Jascha Horenstein, (recorded the four R. concertos).

  12. I highly recommend Alicia de Larrocha’s recording with Previn. Who would have thought that a pianist who stood 4′ 9″ tall and could just reach an octave would produce such a wonderful, clear, full sound, giving every inner voice its due? But in a way, the thick writing of this concerto is quite similar to the Granados Goyescas, at which she was the undisputed master. One pianist told me he ranks a the difficulty of a piece by how much time he needs bring it back to performance level, if he hasn’t played it in a season. Rach 2: two weeks; Rach 3: 2 months.

  13. I would like to thank you for your magnificent review of Rachmaninoff’s Third Concerto and for the mark you give to Yuja! I think she really deserves it!
    Important news: it seems that in her current concerts, she plays the original (heavy) cadence! And I’m sure she can make it light! I look forward to hearing it.
    Finally, I wanted to tell you that “Chinese Piano” will be published in Spanish this fall! Thank you again for everything you write.


    • I’m glad you like my review! Thanks for the news in Yuja’s choice of cadenza in Rach 3—with her technical mastership, I’m sure she has no problem tackling also the more strenuous “ossia” cadenza! I wonder whether I’ll have a chance to hear her with this (since I gave up reviewing for Bachtrack, my focus has shifted towards solo and chamber music recitals…)!
      And of course, I’ll be happy to amend my book review for the Spanish edition!
      Best wishes, -Rolf

  14. Dear Rolf,

    A friend tells me that in reality Yuja did not play the ossia cadenza: this affirmation was a mistake from the critic!! Yuja herself wrote it to this friend. Hopefully she will play it in the future… By the way, she played Rach 3 at the Proms of London and received two ecstatic reviews, including these words:

    “But the rhythmic nuances, and daring rubato flickers and pivots, revealed an artist honestly in search of the piece’s inner truth, rather than some meretricious acrobat content with a parade of flashy stunts. She made us forget the intrinsic difficulty of this famous finger-punishing ordeal and see instead its evolving landscape of emotion.”

    And, what I always thought, but that it took critics a long time to recognize:

    “Wang has a very public face – but, in this work anyway, a beautifully private musical voice.”

    And in conclusion:

    “The headlong dash to the close was thrilling to behold, the Prommers ecstatic in their acclaim for this exceptional pianist, who appeared to wipe away tears of happiness as she left the platform.”

    In other words, your opinion about her Rachmaninov is largely shared!

    When my book is published in Spanish, I will be happy to give you the reference!
    Thank you for including the link to my website. I’m a little ashamed of its artisanal character (I did it myself): yours is beautiful, and I wouldn’t be surprised if you did it also by yourself! Congratulations!


    • Dear Étienne,
      thanks for the update! I don’t think the “ossia” cadenza is a “must have”: artists should choose the one that suits the character of their interpretation (shame on that critic for not recognizing which cadenza Yuja played — they are easily identifiable in the first bars…). People who want to demonstrate power, force and stamina will of course select it (Matsuev, of course), while others who might want to highlight the lyrical / romantic aspects in the concerto will and should select the standard one. As I stated, I have no doubts that Yuja can master either cadenza — but I would not call power and force the primary characteristics of her playing, hence…
      Website: sorry for not having included your personal Website so far — I corrected that now. I suspect that I searched Google for your name, and back when I wrote the article, your Wikipedia(.fr) entry may have shown up first.
      I don’t compare other Websites with mine — there’s definitely nothing to be apologize for here — I really like the esthetics of your site! Yes, I did mine completely by myself, too (with the help of WordPress, of course) — but my objectives are entirely different, and so is the volume & the structure. Of course, I don’t want/mean to promote anything but the Web contents themselves, so esthetics is secondary: given the sheer volume (1.5 million words), the primary focus is on functionality and accessibility (be it only for my personal use of the Web contents, which serve as my personal concert and review archive). I still call my site “Blog”—and it does function as a blog — but unlike typical blog sites posts don’t gradually “retract into oblivion”, but most (all but a few) blog posts receive a limited, but more or less constant stream of visitors. And so, I spend a disproportionate amount of time keeping *all* blog posts readable and accessible. As you may have noted, I have taken a (roughly) three-months “leave” from concert visits and writing new blog posts: I spent the past three months (!) re-building all of my 500 blog posts (no changes in layout and content, except for minor corrections, and adding outlines to >200 posts so far), in order to keep the contents editable & quotable, and to make it even more usable to readers… For sure, I’m not running out of work! Actually, I barely have time left to review CDs from my library — I get plenty of free concert tickets & invitations, and on top of that, there’s an ever-so increasing flow of CDs that I receive for reviewing. I was hoping to deal with some of this over the past months, but then, I blocked myself with the Web maintenance / restructuring… I don’t let myself get overwhelmed, though: I’m still highly motivated and enthusiastic in writing about music…
      Best wishes,


Leave a reply—comments are welcome!

%d bloggers like this: