Mozart piano sonatas

[king ubu]

re: K 310 - love that one! Kempff is probably a favourite there, his Mozart disc on DG is pretty darn great in my book anyway!

And Schnabel ... I guess he's one of my own private deities 

[Marc]

KV 310 has always been a special favourite of mine.
I don't have that many recordings, but I like to listen to the performances of Ránki, Zacharias, Pires (Denon) and Van Oort.

[Cosi bel do]

Cosi,

I love how you are dissecting these sonatas and are making a path in the maze of the available renditions! Besides, it makes great reading!   

Bravo!

Well... Thanks ! It is really captivating indeed, I didn't think I would hear such different things in Mozart, actually more contrasting interpretations than in Beethoven. And if it is useful to others it is even better !

re: K 310 - love that one! Kempff is probably a favourite there, his Mozart disc on DG is pretty darn great in my book anyway!

I'll destroy him comment his performances with the other cycles too, when I start again from the first sonata.

KV 310 has always been a special favourite of mine.
I don't have that many recordings, but I like to listen to the performances of Ránki, Zacharias, Pires (Denon) and Van Oort.

You're a man of taste

[king ubu]

I'll destroy him comment his performances with the other cycles too, when I start again from the first sonata.

Alright, I'll wait and see what you'll have to say ... just enjoying Kraus' 1954 KV 310 as I'm typing this, btw.

[Mandryka]

The more I hear this sonata K.310, the more I like it. And it would be a perfect subject for a blind comparison.

Thanks to your many suggestions I listened to 19 versions. Most of them very enjoyable, but I have only one version to add to the 4 "best" ones I already cited yesterday (Lipatti 1950 studio + live, Gilels 1971 live in Ossiach, Richter 1989 live in London)... And it is Lubimov. I will not listen to the rest of his cycle before I finish the sonatas and start over, because I know I like it too much, but it was very pleasing to listen to it before jumping to sonata K.311. By the way, it's astonishing how it sounds at times like Gilels playing the fortepiano. It's not the first time similarities between the two Neuhaus students really catch my ear.
Among the versions that have been cited, Schnabel is of course very nice but lacking a certain tension and spirit very particular to Mozart (but I might change my mind when I listen to it again as it remains magically articulated). Pommier is neat and straightforward, not bad at all, despite a few mannerisms, and a lack of dynamic variety, leading to a very unambiguous reading. I already said how much I like the Ranki version, that I put just slightly behind my very favourites (but I could very happily live with it only if I had to), just as for Kraus and Barenboim. Also among these "4 stars and a half" versions, I count the live version by Sokolov in 2012 (very annoying that we don't get to know which performance this is), incredibly detailed but never mechanical or too demonstrative.
I'll listen to Arrau with the other cycles I keep for later.

And now the lesser recorded K.311 should not take as much time.

The comments about Pommier made me reflect a bit on how I like to hear Mozart played, so thanks for taking the trouble to respond. Now I wonder if frankness, straightforwardness, no ambiguity, isn't in fact the best way, because it's the strongest - the mixture of a strong clear direction and vision and plain-speaking honesty is maybe what I was getting at when I suggested that Pommier was noble and heroic in 310, the comment which intrigued premont. And I wonder if this isn't what Holden is hinting at in his advocacy of Würtz.


Gilels has some of this quality I think, though not so much in 310. You're much keener on his 310 than I am in fact. I'm really thinking of his 570 on Testament.

[Cosi bel do]

The comments about Pommier made me reflect a bit on how I like to hear Mozart played, so thanks for taking the trouble to respond. Now I wonder if frankness, straightforwardness, no ambiguity, isn't in fact the best way, because it's the strongest - the mixture of a strong clear direction and vision and plain-speaking honesty is maybe what I was getting at when I suggested that Pommier was noble and heroic in 310, the comment which intrigued premont. And I wonder if this isn't what Holden is hinting at in his advocacy of Würtz.


Gilels has some of this quality I think, though not so much in 310. You're much keener on his 310 than I am in fact. I'm really thinking of his 570 on Testament.

Well, I agree that frankness, straightforwardness is a quality in Mozart. But you can't play Mozart (or anyone else) without a certain tension, the score is rarely deprived of all ambiguity. I'd say the most unambigous and straightforward interpretation is Gould, and it is really bad and brutal...
The best "frank and straightforward" readings I heard until now are Ranki's. As usual everything he plays is supremely clear, always simple, with precise dynamics and very few and ultimately necessary inflexions. And the result is never cold or too formal (which makes him generally better than Uchida for instance, despite her qualities). That's not always what I prefer (among studio recordings only, Pires soft and modest piano, or the lively readings by Barenboim, are really pleasing, and Kraus' sense of detail and capacity to recreate the speech note after note is fascinating), but it is always almost perfect in a way.

I'm currently finishing sonata K.311 Comments coming in a short while.

[Cosi bel do]

To clarify how I rate different versions of 310, I let you have a look at my personal evaluations. I absolutely don't mean to give a definitive "note" to each interpretation, which would be very trivial, and I frequently revise my evaluations when listening again. This is just an indication of how I rate each one depending on all kinds of aspects, technical quality, interpretative insight, even (but as marginally as possible) the sound of the recording.

Of course this is something I do without pretention, and if I compared versions more carefully it is possible a version rated 3 would be 3,5 or 2,5, I mean, this is not the point (I also always write, whatever I'm listening too, what my impressions are, which is in the end more important). And about the stars, I use that system because it's visually more efficient when I'm reading my files again, it allows me to identify more quickly which versions I disliked or really adored (and to quickly compare new versions to my previous "best" ones, etc.).

What I wanted to show this for is that it can be seen I quite like almost all versions, except two. I do not deny Pommier's or Hamelin's qualities, but I also don't think what they do is perfect or exceptional, and I see more aspect that can be criticized that in Sokolov, Ranki, Gilels etc.

[Mandryka]

I'm currently finishing sonata K.311 Comments coming in a short while.

Re 311, there's a very interesting,dark, one on from Bozhanov, who played it in the 2009 Van Cliburn competition. It must have been a sonata which Landowska loved -- she recorded it twice. I believe Rudolf Serkin recorded it, but I've never found the recording -- can anyone help?

[Moonfish]

Somehow my mind went to a mental image of a box set focused just on Mozart's K310  (like the Stravinsky Rite of Spring box)!!   

[Jo498]

The Gulda is not really a studio recording. Those tapes are in mediocre sound, backups taped during preparations for recordings that never came to pass. As Gulda apparently played most or all of the sonatas in several concerts around that time I wonder why there are no live recordings from Munich or so around 1980.
(There is one disc worth of studio quality recordings from the same time that was published by DG long after having been recorded, but years before the "Mozart tapes" with K 570, 576 and the c minor fantasy and another one on Amadeo (1970s) with K 331 and 333.)

It has been a while that I heard them and as I got them in volumes first, I know the first volume of the "Mozart tapes" better, but I found them fascinating despite the problematic sound. But they cannot be a first choice because of these problems. I would probably recommend the single DG disc in any case.

Overall, for some reason I do not like most of these pieces as much as I probably should (with a few exceptions like 310 and 576); for me they are a comparably uninteresting part of Mozart's oeuvre. I gave away the complete recordings by Badura-Skoda (Eurodisc, modern piano) and Endres (Arte Nova) some time ago when I wanted to reduce my collection before a move. So now I have the Gulda tapes, Goulds idiosyncratic (ranging from terrible to interesting), a few single discs (Schnabel with 310, one disc with Zacharias, two sonatas with Schiff as a filler to a concerto) and a bunch of other discs on historical instruments: one of Staier's, two of Hakkila's and the "late" Sonatas with Immerseel.
As the Kraus (both, I guess) have been highly recommended to me for years from several sources, I am interested, but not willing to spend the money asked for the big boxes. There is one cheap disc on Sony (unfortunately again of usual suspects like K 331 I am not terribly fond of) I could try.
Or I might try one of Klien's VoxBoxes.

[Marc]

[....]
You're a man of taste

Thank you, sir.

[....]

I take it that the huge differences in timings are mainly caused by the inclusion/omission of repeats.

[Cosi bel do]

Yes, repeats are the reason for these timings differences.

I've just had a lot of fun doing a video as I like them, with a bunch of interpretations put together in the first movement of K.310. You won't find all of them since it is quite a short movement (I omitted/forgot Haebler or Gulda for instance) but the purpose is not to include every version. I still found contrasts fascinating.

http://www.youtube.com/v/-pVuxIJ1VTo

[George]

Yes, repeats are the reason for these timings differences.

I've just had a lot of fun doing a video as I like them, with a bunch of interpretations put together in the first movement of K.310. You won't find all of them since it is quite a short movement (I omitted/forgot Haebler or Gulda for instance) but the purpose is not to include every version. I still found contrasts fascinating.

http://www.youtube.com/v/-pVuxIJ1VTo

Superb work!! Thanks for sharing that.

[Marc]

Yes, repeats are the reason for these timings differences.

I've just had a lot of fun doing a video as I like them, with a bunch of interpretations put together in the first movement of K.310. You won't find all of them since it is quite a short movement (I omitted/forgot Haebler or Gulda for instance) but the purpose is not to include every version. I still found contrasts fascinating.

http://www.youtube.com/v/-pVuxIJ1VTo

Great fun!

[Cosi bel do]

I relistened a 20th time to K.310 and Hamelin's live recording in Montreal in 2011. I'm afraid I can't consider it is more than a very trite performance, not technically perfect (articulation is vague in many agitated passages, there's too much pedal...) and not really structured, which results in flat moments when you'd expect surges of tension. Of course it still is kind of clean and enjoyable, but not exceptional either, and quite forgettable.

In K.311, I compared 11 versions. Kraus and Ranki are clearly more refined, detailed, lively, and also technically very perfect, which is important because clarity and articulation are essential in this short sonata where Mozart, in 1777, gives the impression that he writes in his own style of 2 or 3 years before.
Apart from these two, Uchida and Bozhanov are actually very close. Uchida still a little formal but not as cold as she sounds in some other sonatas, and with good ideas. Bozhanov very detailed, imaginative, fanciful, and he strongly reminds what Barenboim usually does in Mozart (only a few technical flaws really prevent him to compete with Kraus and Ranki).
Barenboim is excellent in movements I and III but, as in many other sonatas, the second movement was obviously played at a different session, and playing and recording are less satisfying, less subtle (that's an endless source of puzzelment).
Landowska would be one of the greatest readings if the performance was complete (I listened to the 1938 recording, partially lost during the war). Her phrasing is really breathtaking.
The most eccentric version is by Carlo Zecchi, a 1954 Westminster recording, really fun to hear even if far from definitive or entirely reliable.

Updated list of favourites :

1 (K.279): Lili Kraus (1954), Daniel Barenboim (1984-85)
2 (K.280): Lili Kraus (1954), Clara Haskil (1961)
3 (K.281): Lili Kraus (1954), Emil Gilels (1970), Daniel Barenboim (1984-85)
4 (K.282): Samuil Feinberg (1953), Lili Kraus (1954), Sviatoslav Richter (1989, live), Andreas Staier (2003), Elisso Virsaladze (2013)
5 (K.283): Lili Kraus (1954), Sviatoslav Richter (1966, live), Dezső Ránki (1978-79), Dezső Ránki (1997, live)
6 (K.284): Daniel Barenboim (1984-1985)
7 (K.309): Sviatoslav Richter (1968, live in Prague), Daniel Barenboim (1984-85)
8 (K.310): Dinu Lipatti (1950, studio in Geneva + live in Besançon), Emil Gilels (1971, live in Ossiach), Alexei Lubimov (198?), Sviatoslav Richter (1989, live in London)
9 (K.311): Lili Kraus (1954), Dezső Ránki (1978-78)

[prémont]

Or I might try one of Klien's VoxBoxes.

I think you should, sooner or later. I find him mandatory.

[Cosi bel do]

I'd still want to make a last comment on K.310, and about Maria Yudina's 1951 performance.

Except the tempo, nothing in her reading can justify to compare it with Gould's, and doing that it is quite unfair to her performance. Yes, going so fast is almost suffocating. But in her case it has nothing of a pure demonstration and, on the contrary, her playing is incredibly detailed, subtle. You can actually hear under her fingers choices quite close from Gilels', only done faster, with high risk (and a few small small mistakes). But the result is not a cold and absurd distortion as with Gould, but a unique sense of urgency.
The second movement gives a rare sense of lyricism and a spontaneous, quasi-improvised style which is always the sign of greatest mozartian artists.
And the third movement is again everything but cold, it's a breathtaking stampede where the median "trio" part is dancing as it never has.
Of course it is not a reference recording, it is mostly very personal (and not technically perfect at all), but it is still a great performance and a more interesting way to play Mozart than most other versions I heard (I mean all the good, well-behaved, a little artificial renderings that are quite pleasant and forgettable at the same time).

Currently in the middle of my K.330 comparison.

[Mandryka]

Currently in the middle of my K.330 comparison.

I hope you'll get a chance to hear Brendel's very late recording

[Mandryka]

I'd still want to make a last comment on K.310, and about Maria Yudina's 1951 performance.

Except the tempo, nothing in her reading can justify to compare it with Gould's, and doing that it is quite unfair to her performance. Yes, going so fast is almost suffocating. But in her case it has nothing of a pure demonstration and, on the contrary, her playing is incredibly detailed, subtle. You can actually hear under her fingers choices quite close from Gilels', only done faster, with high risk (and a few small small mistakes). But the result is not a cold and absurd distortion as with Gould, but a unique sense of urgency.
The second movement gives a rare sense of lyricism and a spontaneous, quasi-improvised style which is always the sign of greatest mozartian artists.
And the third movement is again everything but cold, it's a breathtaking stampede where the median "trio" part is dancing as it never has.
Of course it is not a reference recording, it is mostly very personal (and not technically perfect at all), but it is still a great performance and a more interesting way to play Mozart than most other versions I heard (I mean all the good, well-behaved, a little artificial renderings that are quite pleasant and forgettable at the same time).

Just thinking of the first movement, dynamically Yudina is more nuanced than Gould, though at a cost of losing some of the momentum. In terms of voice leading I think that Gould has interesting ideas. And I'm not convinced that Yudina makes the repeat interesting - I'd need to listen again to it. Yudina gives the impression of breathlessness; Gould gives the impression of an unstoppable vector.

Generally the sort of strong dynamic contrasts that Yudina gives it may not be stylish for this sonata, just a romantic aberration.

Yudina in the 3rd is very imaginative. I can't remember anything about either Gould or Yudina in the second movement!

[Cosi bel do]

I hope you'll get a chance to hear Brendel's very late recording

Will do (now that you mentioned it ).