Perfect, Monica! I have to concede defeat here: I have found the indication on my scores too, but marked earlier than I the point where I noticed your speeding up. I must say I learnt this piece while I was studying and at not time did the teacher say anything about this indication. Come to think of her, she usually considered such effects in bad taste. (no half-baked pianist but a graduate of the Paris Conservatory) Anyway, I listened to all recordings on this site and none of them is consisitent: some do not accelerate, others do so only on the last two bars.

But then you see, I am one whose natural tendency is to accelerate and it is at times a struggle to keep time, so I suppose my ICC (Internal Censorship Centre ) chooses not to see such markings if it can.get away with it.

Excuse me again. As a penace I will play this piece three times in a row.

Why a sigh at the end ? Not on my score.

Bravo Monica! You did it!

Oh darn, I didn't know that 'sigh' was there!
Actually, it's really just me letting out my breath because I probably held it the whole time I was playing. I need to have 'breathe' marked in my score in a couple places, since I do the dumb thing of 'not' breathing when I am nervous and playing something fast. Good thing I got the piece down in around five minutes; it's not easy to hold one's breath that long.....
Thank you for listening, Didier.

Richard, okay you go play this piece five times in a row, but don't forget to hold your breath the whole time. Oh, and also you have to pat your head and rub your tummy at the same time too!

Monica, I hope you'll excuse my late arrival to the party. I offer some of my thoughts in our "virtual masterclass" here on PS (some of which may be controversial). As far as the etude nature of the piece I offer that it bears remarkable kinship in spirit to Czerny Op.740 (The Art of Finger Dexterity) No.17, possibly Op.299 (School of Velocity) No.26, but mostly (for me) the Moszkowski Op.72 (15 Etudes of Virtuosity) No. 6 in F major. IMO, the piece is best approached as an artistic etude. In fact, if I were sitting on an Oral Board of Examination for an advanced candidate in Piano Performance and Literature, and admittedly wanted to stew the candidate, I would ask if Schubert composed any Etudes. And when the candidate answered "No," I would follow with "What about the second Impromptu of Op.90?" The point being, not to be so limited that we forget to see the music at the cost of the title (as I said before, "A rose by anyother name is but a rose"). This fist step may be the most important and controversial, Do I approach the work by the nature of its title or its betrayed purpose? I think it makes a very big difference. I think it should be approached as if it were an Etude of your favorite etude composer (Cherny, Cramer, Clementi, Chopin, Henselt, Liszt, Moszkowski, Moscheles, Haberbier, etc.)

First, so that we're on the same page of music, the first note is in measure zero (0). As such the Ben Marcato ("B" music) begins at m.83, the return of the "A" material at m.169, and the "Coda" at m.251 (but this is false, as it is the "B" music; the real Coda, is not a Coda at all but a Codetta that is from m.276 to the end). That completes the overview.

I think the entire "A" section should be played without pedal, as the very nature of the thematic material is scalar and creeping chromaticism. This first section should have "sweeping" motion of dynamics that rise and fall and climb even as a roller-coaster does. With so many fast notes in the right hand, the notes themselves become less important and the work of the pianist is to shape the groups and phrases. The absolutely worse thing anyone can do with this work is to "turn the right hand on" and let it play like a machine gun. This balerina needs "sparkling delicacy" that is "supported" by the male dancer of the LH. For the section at m.26-, please voice the RH half notes fully, recalling with them the second-beat half notes of the LH material (if you don't think of it as such, you won't play it as such). Sometimes here the right hand is "required" to do some thumb/hand passing, or is it so? Have you thought that on occasion (m26, 28, 30, 32, etc.) you may play the down-beat first note of the RH with the LH? This way you don't have to do the crossing with the RH. Having come this far and noting the importance that Schubert gives to the 2nd beat, is it possible that it changes the way that we play the LH at the beginning? I think it does. For me, I see now (in analytical retrospect) that the sounding of the second beat is as significantly characteristic to the work as any other feature. Think about it. To wrap up the first departure and its return, I would make a big deal about voicing the Cb in the LH in measure 49 so as to carry it through to an audible resolution on the Bb in m.51.

Reagarding the Ben Marcato in m. 83, the most important thing I can say, is that Schubert failed supremely in the manner that he wrote his enharmonic modulation, and if you play it his way, you are starting off on the wrong foot. This is what I mean, the Gg major chord that serves as dominant to B minor is a huge sublimation! I would "change" the Gb major chord of m.82, to an F# major chord! Don't laugh, it makes all the difference in the world. Think about it. In this section I would use some judicious pedaling, but never allowing scalar melody to co-mingle the notes (meaning if you use pedal in m.85, you will need to change it on each beat) . The phrases here are four measures long and should be articulated with a break in between (just because he doesn't "compose" the [breaths with] rests doesn't mean there aren't any. I know you know what I mean.) At m.105 the RH should begin with fingers 2/5, no pedaling, always maintaining the prominent E# resolving to F# (which should be shortened so as to take a breath as the last note of the four bar group). (Last sentance is of course the same for m.149, etc.)

At the improperly labeled "Coda" (now clearly seen as just "B" material, making the auditor think it is a rondo or repeated binary form), again enter with an F# chord not a Gb chord . Here I will comment directly to your interpretation and say that just because the second four-bar group (m.255-258) is in eb minor, it in no way means you should draw back on the force of the statement. You do the prior phrase with nice attitude, but then retreat when it shifts to the closing key. I would not do that; I would maintain a strong character throughout. The only dynamic from here to the end is FF with several fz thrown in for good measure! My score (the Breitkopf & Hartel reprint by Dover) shows the accelerando beginning m267. Harmonically, he writes a hemiola for bars 280-281 (i.e. meter change to 2/4). Don't be afraid to play it that way: 1,2,3,1,2,3,1. You have a slight hesitation on the penultimate chord that spoils it for me (can you close the gap with editing?)

There you have it. Admittedly more about the work than your performance, but I think it is important. The big take away that I would tell you is, "Play it like a great etude and damn the torpedos!" Don't use any pedal if possible and think about the interpretive and "applied techinique" issues I listed above. Personally, I think you have the piece ready now to simmer, and when you return to it, you may have a masterful performance.

Wow, thank you, Eddy. You have surprised me! "Better late than never" is a something I believe in.

Here is my response:

I agree about bringing out the RH half notes at m.26. I did this better in another recording attempt but then thought someone may think I played them too loud. I should have stuck with my instincts there. Regarding using my LH for the 1st note in the RH: That's a very interesting idea and I went right to my piano and tried it, but alas I cannot reach the notes. I'd have to jump the 1st LH note to the 1st RH note and although it can be done very quickly, there is still a slight 'bump' and I'm not sure that would be acceptable.

Hit the C-flat more in m. 49 - that's good, I'll make note of that.

Ok, now about changing the chord at m.82 from a G-flat chord to an F-sharp chord. I can sort of understand where you are going with this, as it leads right to the key change at m. 83. But I'm not sure the ear could hear this difference. Unless you mean that one should sort of push forward on that chord? Except it's marked ffz and I thought I came down pretty hard on the chord...I guess I don't really understand what you mean. Plus, there is a rest on beat 3.

Regarding the phrasing in the opening lines of the B-section. I felt the length of the phrases to be 8 measures, not 4. Am I wrong? Are you right? Or is it subjective? I really do not know, since I've not studied this piece with a teacher.

M. 105 - yes, my fingering is marked 2 and 5 but I changed it to 1 and 5. I also mistakenly let go of the E-sharp too, but thought no one would notice.

Coda : My score at ms. 255-258 are marked mf, so that's why I backed off there and also at ms. 263-266. Did Schubert write in these dynamics? Interesting idea about playing those staccato chords at ms. 280-281 like how you describe. I'll try that. As to the hesitation - it's only just me trying to get my hands onto the next set of notes. I guess I could close it up with editing, but I would not feel right it. Is it really that bad?

Finally, when I came back to seriously practicing piano about 11 years ago (after not practicing for nearly 16 years), I felt that EVERYTHING I played was an etude, because each piece had some sort of technicality to work out or I needed to drill on the scale/arpeggio passages. So I think it's interesting that you some others here feel so strongly that this 'Impromptu' is really an 'Etude'.

Well, thanks again, Eddy, for the thorough and careful critique. You can now ask me to change your bio page again if you wish.

Regarding the magic I'm talking about at the enharmonic modulation (see my prior post), it is about attitude that sneaks its way into performance. This is very psychological and could never be proven "scientifically" but we're talking about art here. How real is it? For me it is VERY real! How real would it be to an auditor? That would be an interesting test. An educated audience is asked to listen to this work performed by ten pianists (half of which in their mind play an F# chord instead of the Gb. Would the auditors (who know that some will play F# major and the others the enharmonic) be able to tell a difference in the approach? I must admit, probably not. But that doesn't change my mind that when I play the first two notes of the Beethoven Op.57 (C-Ab) I do so in my head over a "silent" Dominant 7th chord, not the tonic chord! The difference is that I feel like I'm playing it lifting off the seat with infused energy, rather than sinking back into a comfortable chair. For me, the enharmonic modulation to the "B" section of this Impromptu, requies a REAL dominant to propel the music into it, not some sappy Gb major chord to sneak into it sideways. Having said all this, it may only affect those who are in a masterclass audience and getting their idea changed too, but I will continue to "live in my own little world."

The difference is editions (taken as legitimate) only raises the notion that we can take more liberties than we think we can in interpreting the work.

Hi Eddy,

There's a lot of good points in what you say. I just disagree with one detail:



This is a rather "hardcore" attitude! Would you play Chopin's opus 25 no. 11 without pedal too? Certainly you need to be a little discreet with the pedal in the Schubert. It's important to have enough clarity, and it would be a good idea to practice the piece entirely without pedal. But in performance, I wouldn't want it to sound too dry.

Alexander, Just one point of disagreement?! I'm elated! To answer your question, no; that is, I would use pedal as I do also on the cascades of the trio section of the Chopin 3rd Scherzo. (But the RH of the Chopin you mention is not foremost, it is background to the LH chords.) I think a fine execution of the Schubert is possible without use of the damper pedal in the "A" section, but can also be accomplished with judicious use. I agree that it is best practiced without pedal, but legato not secco.

Hi,

Personally, I like the Schubert with judicious and careful pedal to assure clarity in light of the continuous neighboring and passing tones in the passage work. Otherwise, I feel it would sound altogether dry. To make a parallel point, back in the early 1990s (Chris calls these my "historic recordings" ), I recorded Chopin's Etude, Op. 10, No. 6 in E flat minor nearly entirely without pedal, as I believed that doing so was actually the whole point of the etude--playing a lyrical etude with legato touch with the fingers alone. I had too often heard other pianists cast an impressionistic-like haze over it that seemed more like a pall. But... if I were to relearn that etude today, I believe I would play it with pedal, albeit a very controlled pedal. It would be the same if I were to undertake this Schubert impromptu. Just my 2 cents.

David

I agree about the pedaling too and think it should be used, but like we all know - you have to be very careful or else you end up with a blur. But talk about applying the concept of etude to this piece; then not only is practicing playing 'finger legato' one of our goals, but also the goal of practicing the pedaling so that it's just right. It's a double-whammy! (for me, anyway) Boy, I think EVERYBODY should practice this piece and then record it so that you can listen back and see if you played things correctly. I really learned a lot from doing so (thanks to all of you! ) from the talk about clear runs, legato, pedaling, getting the triplets/rhythm right, and the list goes on. Most of you do not know this, but a few years ago we had someone submit this piece (and a couple other pieces) but it turned out to be a fake. He/she was not the player - the piece came off a demo-track from a digital piano. We booted him/her right off the site. Anyway, that's when I first heard the piece, liked it right away and knew I wanted to play it too some day.

Hi Monica,
first, let me say, I love clear playing (at least in 99% or so of all cases). And also concerning this piece, it´s the usual way to play the runs clearly, more or less without pedal respective some attentive foot kicks. That doesn´t mean, that an exception can be also convincing. (You know, artistic thinking can´t be hundred percent fix respective "stiff".)
In your version you use more pedal than I do in my newest version, btw.
And generally this piece cannot be considered as a real etude, I think that´s evident, at the best as an "etude" with musical respective artistic value like f.ex. the etudes by Chopin.
In summary you did a quite nice verison here, of course! I have the following positive remarks and suggestions of improvement:
The crescendos and decrescendos in bar 3-4 respective 11-12 and parallel places could be more audible (I hardly could notice them).
I don´t understand the accent on beat two in bar 38 in the bass-voice.
The ff is too weak (bar 68 ff.). Especially in Schuberts music the extreme dynamical contrasts are essential. They are expression of his inner disruption and his suffering, which was immense. They mirror the contrast between suffering respective resignation and sudden respective quasi eruptive attempts to fight against his destiny. Especially the coda is such a moment and should be played as fortissimo as possible (you play it much too softly and elegantly).
The cresc.-decresc. in bar 85, 89 and parallel places are missing.
I like your ideas of staccato at some places (which is not in the Henle-Urtext-score, btw), but it´s a nice change for the year and underlines the dancing rhythm of the 3/4-bar, like f.ex. in bar 101, 145, bar 154-162 (here I really love that, though its not in the score!)
On the other side you don´t play staccato on many places, there is a staccato-point in the score like f.ex. in bar 123, 124. (I also take some few freedoms concerning the staccato, btw).
The tempo of the reprise (A-section) still is too slow compared with your tempo before.

It´s a pity in my humble opinion, that you say, it´s the last time you have recorded that piece, Monica. I think, we are in a piano forum here, which sense it is to give us praises and suggestions of improvements, so that we always feel encouraged to work further on the pieces and to bring them up to a certain perfection.

Thanks, Andreas. Good that you liked my staccatos. Feel free to use my way in your own playing. Regarding the 'etude' issue - sorry, but I agree with the others that this piece is an etude. But really I've grown bored talking about it, and anyway I have already moved on to learning some new pieces.

I agree with your point but in this music I would not recommend 'as fortissimo as possible' on a modern grand. On a period instrument you'd probably need to do that to make the desired impact but on a bright powerful Yamaha I think it would be out of place. This was a discussion I had with my teacher about the many places where Grieg wrote ff or even fff, which on a modern grand we should treat more as f and ff respectively.

That is a very good point.

Then of course, if on a modern grand we can make more noise than on a 19th century instrument, there seems no reason why we shouldn't