Over the years I studied this piece with two teachers, and performed it in auditions and recitals, so I know it very well. I also have a recording of "La Cathedrale engloutie" but not at this website.
A few suggestions:
First, far better control over dynamics is needed! The opening is marked pp, but sounds more like mp. (These parallel octaves and fourths are best played from the shoulders with still hands.)
In the development, section A1, at measure 13, it's marked pp, but you seem to be at f! You're also holding tones through the rest in the second beat there. Observe the rest!
On page 2 you maintain the tempo, which I believe is the correct approach. (I played it about MM = 69). Much has been made of the Debussy piano roll where he takes a significantly faster tempo there; however, we cannot always trust what we hear on piano rolls or old 78 rpm records. All too often recording engineers from that bygone era asked artists to speed up the tempo in places (or even to make cuts) in order to fit the piece on the roll or the record. The recording technologies back then were very limited which led to unfortunate, undesired and misleading accommodations.
At the top of page 2, it's marked pp, but you're near or at f again. The highest dynamic on that page is f, but which requires a crescendo starting at the augmentez progressivement marking half way down the page. These premature loud dynamics detract a lot from the imagery of the piece. This is the legendary Cathedral of Y's rising very slowly up through the ocean which takes time. You have it bursting all at once through the surface of the water with loud dynamics way before their time. Bear in mind that the full apparition of the cathedral does not actually occur until page 3, section A2 at the Sonore sans durete indication. Until then, it has been rising in the waters little by little. By the time you reach the peeling bells at the bottom of page 3, you're already at ff, a dynamic that is not supposed to occur for three more measures. So, you have nothing left in reserve for that--you've already spent the ff too early.
Starting in the last two measures of page 1 with the scales in the Phygian mode and continuing through nearly all of page 2, there are very long pedals which are permissible; however in the last measure of page 2, that comes to a halt. There and at the top of page 3, in the first two measures (or actually starting on the 4th beat of the last measure of page 2), the left hand, which has been secondary, now becomes primary. So in playing those descending octaves, you must pedal for clarity in the diatonic scalar passages. That means pedaling every octave separately and individually. Right now you have one glaring, continuous and unmusical blur there. You also make a mistake I made where it is marked sff in the second measure page 2, that is, playing the last two bass octaves allargando or broadening them out. As another pianist quite logically pointed out to me, we're dealing with church bells there, so an allargando would be quite improbable. I believe he had a good point, and there should be no pause between those last two octaves.
Once you're at section A2 at the Sonore sans durete, your pedaling is not to best advantage. You're dealing with several things there. One is to produce full sonorities, which you do well. But you also need two types of pedaling there. For each pedal point (sometimes called organ points) on the contra-low C's, you should take those with the middle, that is, the sonstenuto pedal, using the left foot, and hold it down until the contra-low C is next sounded, whereby you again take it in the sostenuto and hold it down again until the next one. Pedaling the parallel triad chords there also needs a completely different technique with the damper (or loud or right) pedal. What works best there are partial damper pedal changes which we call half-pedal releases. Notice that the first chord in C in the right hand is consonant with the dominant G chord two chords away, but they are separated by a dissonant D chord in between. So, after the sostenuto pedal has caught the very low pedal point, you would depress the damper pedal in a syncopated pedal, ease the pedal up a bit to "spill" some of the C tonality with its overtones, play the D chord, release the pedal a bit more to spill the D, and finally allow the G chord to sound. That will give more clarity to the intervening D chord. Same technique with the following F, E and D chords, that is, spilling some of the E tonality through a half pedal release to manage the dissonance between the consonant F and D tones. As it stands, you have one huge blur throughout the parallel triads. It's really essential to pedal for clarity there with the method I've described.
Page 3 bottom, the transition featuring the major 2nds: The measure just before the last measure--you completely neglected to play it. Also, that line starts at p, then piu (or more) p, pp, piu pp, and the very last G# bass octave logically should be ppp given the long diminuendo there. To me you seem to play the whole line at mf. You really need to effect the diminuendo there.
Page 4: Here again, the dynamic is pp, but it sounds like it's being bashed out at f. It needs to be way quieter! There is no f until measure 13 on this page. Once you get to measure 13, you need to play the bass grace note, contra-G# and do an upward LH leap to play its next chord. You're now playing the grace note, then both the LH chord and the RH chord are being appegiated or rolled. It's not written that way in the score. (I'm looking at the Durand Edition which was the first edition.) The grace note should be played followed by solid chords in both hands in rhythm.
Page 4, measure 19: Look closely at the second half-note chord in the measure. You're playing it as it there were a bass clef in front of it, but in the preceding measure, the G clef has returned. So it's just a misread.
Last page, recapitulation: You're voicing the chords with the RH thumb which is correct. Good! But back to your nemesis again, dynamics. When you tapered off the preceding page, you were at pp. I thought, "Halaluja!" But then as soon as the right hand chords came it, it was back to loudness. Note that Debussy marks that page pp. The cathedral is now sinking and disappearing under the waves. All is anticlimactic now. The whole atmosphere has become subdued. You do a fine job with the left hand, keeping it in the background at ppp while playing it legatissimo to produce a quiet, tranquil effect as the cathedral gently settles on the bottom.
In the last line of the last page, the coda and reprise of section A1: Do not hesitate at all between the last chord of measure and the first chord of the next. Same with the last chord of the second measure approaching the first chord of the third measure. Keep it all moving. In the third measure the second chord is all whole notes. Be sure to hold them for full value. Right now you're rushing through the chord.
The important thing here is that Debussy's music is impressionistic and deals with imagery and atmospheric effects. Much of it has to be managed very quietly. If most of the playing is blaring loud dynamics, there is no way to create the atmospherics of the music. Watching the dynamics very carefully is key in order to succeed with this piece.
A tip: If you play this in a recital, "play" the rests after the C cadence at the very end. To do that, lift your hands slightly, be looking down at them with your eyes (do not look at the audience), and very slowly raise them off the keyboard and then very slowly withdraw them. That cues and prevents the audience from applauding too soon and breaking the magic spell of the music.
l hope this is helpful.
"Interpreting music means exploring the promise of the potential of possibilities." David April
Last edited by Rachfan on Sat Oct 23, 2010 9:31 pm, edited 10 times in total.