Beethoven's Piano Sonatas

[Mandryka]

Sorta cross posted from the Recordings That You Are Considering thread.  This disc is part of a series that will ultimately include 24 of the sonatas from a live cycle recorded a few years back.  But it's download only, at least for now.  I like me something to hold in my hands.

The first movement of op 111 is epic style, punctuated by moments of interiority. My impression was of thoughtful and nuanced playing. The second movement is stuffed with extreme rubato, mega-romantic, I thought it was horrible, horrible, horrible.

[Mandryka]

I've heard the EMI disc, and the sonatas are good, but nothing special,

His way of playing op 2/1 and 10/3 on Denon reminds me of Katsaris. I haven't heard any other Beethoven by him.

[Jo498]

I do no think there is a recording of op.54 by Gelber (unless on an old EMI LP but I doubt it), the additional x between 21 and 22 in Jens's scheme suggests a typing error.
The Denon discs are the following (confusingly not all of mine say explicitly "Vol. x", there may even be slightly different couplings around, e.g. opp.13, 27/2, 57 or so)

Vol.1 opp.13, 49/2, 10/1, 2/3
Vol.2 opp.27, 1+2, 28
Vol.3 opp.57, 31/3, 81a
Vol.4 opp.53, 90, 111
Vol.5 opp.31/2, 79, 101 (I lack this one)
Vol.6 op.2/1, op.10, 2+3

recorded from 1987-95

Hopefully, I find leisure to listen to some on the weekend.

[jlaurson]

I do no think there is a recording of op.54 by Gelber (unless on an old EMI LP but I doubt it), the additional x between 21 and 22 in Jens's scheme suggests a typing error.
The Denon discs are the following (confusingly not all of mine say explicitly "Vol. x", there may even be slightly different couplings around, e.g. opp.13, 27/2, 57 or so)

You are right on. It snuck in there, somehow, but wasn't on the list I had actually compiled. It's been mentioned now, but not actually added, on the list of "Great Incomplete Cycles".

[Jo498]

Someone above mentioned that several of Gelber/Denon are on spotify, so who's interested can try for oneself. With a little patience some of the discs are available fairly cheap either from CDjapan or similar sources or used from amazon marketplace, so it's no great expense to try one or two.
It seems MUCH harder to get the Annie Fischer/Hungaroton for acceptable prices... typical price from a marketplace seller for the Gelber CDs is around EUR 10-11, for Fischer 14-19. Ain't gonna pay that kind of money for Eastern bloc recordings

[kishnevi]

Someone above mentioned that several of Gelber/Denon are on spotify, so who's interested can try for oneself. With a little patience some of the discs are available fairly cheap either from CDjapan or similar sources or used from amazon marketplace, so it's no great expense to try one or two.
It seems MUCH harder to get the Annie Fischer/Hungaroton for acceptable prices... typical price from a marketplace seller for the Gelber CDs is around EUR 10-11, for Fischer 14-19. Ain't gonna pay that kind of money for Eastern bloc recordings

I have the Fischer set..it is available for about $90US on Amazon MP US,  and there seemed nothing wrong with the sonics. But I don't rate her as high as other people do.

[Jo498]

The complete Fischer set goes for ca. 90 EUR in Europe as well. Sure, that's cheaper than 140-180 for 9 single discs would be but 93 EUR is very expensive for such a set in 2016. I bought my first two sets in 1997: the incomplete Gilels for 90 DM or so (less than 50 EUR), and in 98 or 99 the Gulda/Amadeo for even less.
Sure, back then there were really expensive sets around but in general complete sets have become so much cheaper that the Fischer must count as rather expensive.

And I already have the Annie Fischer "Introuvables" with 7 sonatas and Vol. 1 + 3 from the Hungaroton. I would have to re-listen but I think I also never quite understood what was supposed to be so super special about Fischer; very good (if often gruff and humorless, hers is the most angry op.31/3 I have heard). I would snap it probably for EUR 50 because it is so highly regarded but not for more than 90.

[mc ukrneal]

The complete Fischer set goes for ca. 90 EUR in Europe as well. Sure, that's cheaper than 140-180 for 9 single discs would be but 93 EUR is very expensive for such a set in 2016. I bought my first two sets in 1997: the incomplete Gilels for 90 DM or so (less than 50 EUR), and in 98 or 99 the Gulda/Amadeo for even less.
Sure, back then there were really expensive sets around but in general complete sets have become so much cheaper that the Fischer must count as rather expensive.

And I already have the Annie Fischer "Introuvables" with 7 sonatas and Vol. 1 + 3 from the Hungaroton. I would have to re-listen but I think I also never quite understood what was supposed to be so super special about Fischer; very good (if often gruff and humorless, hers is the most angry op.31/3 I have heard). I would snap it probably for EUR 50 because it is so highly regarded but not for more than 90.

No, I think you are right that it is one of the higher priced ones. Even some recent ones like Guy and Lewis are quite a bit less than Fischer. With so much choice (and quality), there is no need to pay that amount if you don't want to.

[Jo498]

I'd be willing to pay at least three times as much for Fischer's as for Paul Lewis' 
But there are limits for everything...

[Todd]

Sure, back then there were really expensive sets around but in general complete sets have become so much cheaper that the Fischer must count as rather expensive.

90 Euros or dollars is mid-price.  Japanese and New Zealand cycles, now those can get expensive.

[Jo498]

Says the man who does not want to shell out for the Gelber discs...?

ca. 10 EUR is midprice for single discs. Boxes with re-issued or older recordings often went at lower prices per disc already in the late 1990s as I said further above.

Today a brand new two-disc set like Levit's late sonatas starts at about 10 EUR per disc and if one waits a few months for the first sale, one can probably snap it up for 15.

Nowadays I buy almost only used discs. For 90 EUR I do not buy 9 or 10, but probably 20 or more. But the hungaroton single discs are almost NEVER on sale. Or I could get Backhaus from Italy, one of the Kempff and still 10 EUR left or so.
I think I got the two Fischer Vols. several years ago at MDT when they actually had a (comparably slight, but substantial enough to induce me to buy) reduction for hungaroton. Probably the complete box was not reduced in price. I was less inhibited then and spent too much on CDs, so I wonder why I did not get the complete set then.

[Todd]

Says the man who does not want to shell out for the Gelber discs...?

That has nothing to do with price.  I know some Gelber, and the set is incomplete.  Therefore, it's low/no priority.  Were it a complete set, I'd buy it.  (Why go Gelber when I can go Kosuge?)  If it pops up in convenient cheap box form, then I may buy.  I'm not concerned with per disc prices.  Life is too short to worry about five or ten euros/dollars here and there.

[Todd]

Looks like Igor Levit is performing the entire sonata cycle at Wigmore Hall next season.  I doubt the performances get issued on Wigmore Live, but I wonder if Mr Levit is prepping to complete the cycle for Sony for 2020 or beyond.

Also of interest, he and Julia Fischer are teaming up to perform the complete Violin Sonatas at Wigmore in July.  If ever they recorded that cycle, it would be worth hearing.

[Todd]

Don't know if it has been mentioned yet, but Wilhelm Kempff's stereo cycle is now streamable at Amazon USA.

[Jo498]

In another thread we have been discussing tempo choices for several slow LvB sonata movements. Despite some obvious parallels between the slow movements of op.22 and op.31/1, both 9/8 "aria" style with some (or lots of) embellishments pianists usually differ in their tempi considerably. More precisely, very few pianists play the op.22 adagio "in slow 3" rather in moderate to slow 9, whereas even in slow readings of op.31/1,ii one can still "feel it in 3". Typical playing times for op.31/1ii are around 11 min, for op.22ii 8-9 min.

Now my request for those who own lots of Beethoven sonatas. I am looking for
1. fairly fast readings op.22ii; the fastest I know is Gulda/amadeo at ca. 5:30, but up to around 7 min playing time qualifies as "fast"
2. pianists who play similar tempi in those movements. My main candidate here is Brendel with ca. 6:40 in op.22 and about 11 min? in op.31. So note that a fast tempo in op.22ii is a "mainstream" tempo in op.31/1. A "mainstream" op.22ii at 8:30 would correspond to >12 min in op.31/1. So pianists need to be fast in op.22 and average in op.31 or average in op.22 and fairly slow in op.31/1.

Furthermore I am interested in very fast and very slow readings of the following slow movements (I put Gulda's timings in brackets as he is often a candidate for the fastest)

op.2/2 (6:32 slowest: Gould 8:43)
op.2/3 (5:50; slowest: Gould 12:45)
op.7 (7:02; slowest: Gilels 10:00)
op.10/1 (Gould 6:17, Gulda 7:18, Brautigam 7:12, Pollini 7:13; slowest: Gilels 11:45)
op.10/3 (8:02)
op.13 (Gulda 5:17; Gould 4:41, Pollini 4:38, Kocsis 4:38)
op.22 (5:29; slowest Gilels 11:18)
op.31/1 (7:55; slowest: Heidsieck 13:55)

In some of those examples we can clearly see how some pianists think that certain movements should be counted in a different unit. E.g. crotchets vs. quavers in op.10/1, dotted crotchets vs. quavers in op.22. Kolisch claims that the slow movements both of op.10/1 and 13 should go in crotchets but only the fastest readings do (kind of). Maybe he is simply wrong. But I still think Gilels in ludicrously slow in op.10/1 and op.22.
(And if one needs more evidence that Gould was mad... listen to these movements, some faster, some slower than anyone else...)

[George]

op.22 (5:29; slowest Gilels 11:18)

... But I still think Gilels in ludicrously slow in op.10/1 and op.22.

I just checked Schnabel's Op. 22/2 and his is at 9:01. I don't think it sounds too slow at all. Given that, Gilels is actually closer to Schnabel than Gulda. I will agree that Gilels's style takes some getting used to, though. At first, I didn't like it at all. But it quickly grew on me with repeated listens. I wish he had lived to finish the set.

[Jo498]

As I have only 7 or so recordings of op.22 (it is not a favorite piece and also not so likely to be included in anthology discs), I am not sure but is seems that most pianists play it around 8-9 min. > 8min is as we discussed in the other thread, usually too slow for "in 3 dotted crotchets".
So Gulda and Gilels are statistically similar far of the average/median but my point is that the median is suspect because I think the piece should go "in 3". Brendel is a minute slower than Gulda but one of the few who apparently also think that way. So does Pollini (6:15).
Of the stats I have so far there seems a gap: Either pianists play it "in 3" (up to ca. 7 min) or they are around 8:30-9:30 (most, the only one close Gilels is Arrau at ca. 10:40). There is very little between 7-8 (Kovacevich 7:13, Serkin 7:50).

opp.2/3, 10/1, 13 seem somewhat similar cases but with 8ths vs 16ths and quarters vs. 8ths as counting units.

[Jo498]

There are some of Gilels Beethoven sonatas I like a lot (often the live Melodiya/Brilliant even more than the DG). But I think there are some movements where the tempi are way off and it does not work (some kind of do despite the extreme slowness). E.g. unlike amw I think a "majestic" slowish first movement of op.106 can work well although I agree that it is clearly wrong.
Independent of tempo I find Gilels also too serious in some of the more humorous pieces.

[amw]

Of the stats I have so far there seems a gap: Either pianists play it "in 3" (up to ca. 7 min) or they are around 8:30-9:30 (most, the only one close Gilels is Arrau at ca. 10:40). There is very little between 7-8 (Kovacevich 7:13, Serkin 7:50).

FWIW I remembered Kovacevich as my favourite performance of Op. 22 and re-listening the other day after I have no idea how many years confirmed that impression.

I also took the time today to retry 22/ii with a metronome set to a slower speed (eighth = 108), and compared it with the one I played following my own instincts (eighth = 148)—which was not a pleasant experience, lmao—and concluded that the best tempo for listening would probably lie somewhere in between, but closer to 108 than 148. So I guess that means recordings in the 5:30-6:00 range. 108 feels pretty slow, but that's partly the metronome (no flexibility whatsoever), and partly that a better pianist could make it more interesting.

(I also tried 31/1/ii with a metronome, though I sort of ignored it a lot of the time because the little cadenzas threw it off. The metronome was set to eighth = 120, and my average tempo comes out to 119, unsurprisingly. I think that's a good & very "typical" tempo for the movement, basically about the same as Pollini, Goodyear and Komen among others.)

[Jo498]

What I tried to point out was that there seem to be very few recordings going for op.22 "in 3" at all. And except for Gulda they are usually closer to 7 min than to 6 (so closer to 100 than 108, which is still in the same range) and very few at the slowest end of what could still be counted in 3. Apparently most players prefer what is closer to andante (ca. 72-82) in 8ths to adagio in dotted quarters. So in a sense Arrau and Gilels bite the bullet and get closer to adagio (in 8ths) than those in between.

I have 3 discs of Kovacevich's EMI series (not the one with op.22). The sound is not so great ("hard and brittle") and he is sometimes a little brutal. The single discs are mostly out of print and somewhat pricey and I don't really like the sonata op.22 enough to get another one although both he and Pollini would certainly be candidates for me. And I actually like Gulda here, I think he is too fast in the finale but the adagio works quite well for me. Similarly with op.10/1 I listened to last night. The adagio has a nice flow despite being at the faster end of normal but the first movement at 80 bars/min or even faster was too breathless for me (most play it around 70). (Then there's Gould who goes through 10/1i at about 112 bars/min and rushes through the thing in 2:40 without repeat...)

Generally, it seems that Gilels and some others take in some of these movements the position that even "small notes" (like 32nds) should go rather slowish, although the melody at the beginning of a movement quite clearly goes in quarters or dotted quarters.

Another outlier from Gilels is the variation movement in op.14/2 where he (Gould as well) ignores the alla breve marking and plays it almost at half the speed of most others (8:05 vs. typically around 5 min.)