Memorizing is hard work. And it’s time-consuming. But it’s worth it, because memorized performance always sounds better (assuming the music was solidly memorized in the first place). Memorizing this prelude and fugue is not to be taken lightly, and it must be constantly refreshed. I believe that the chromaticism of the prelude and the close strettos of the fugue just make this piece a memorization nightmare. Not even the Vierne 6th or the Clérambault first Suite was this tricky!
Well, that’s that. Below are some of my insights into playing the piece, memorized or not:
Dupré’s fingerings and pedalings are welcome, although I changed quite a few to fit my own hands and feet. I wonder what he would have thought. Sometimes he goes to an unnecessary amount of trouble with substitutions, whereas I just tuck a thumb or cross a long finger over another.
Graham Steed’s book has excellent insights into all of Dupré’s organ music. He knows of a couple corrections for the C Major that Dupré confirmed post-publication (measures 27 and 179, mentioned below). I see several additional opportunities for consideration.
Measure 12: The final eighths in the manuals will have to be released early to be repeated in measure 13. Dupré would have prescribed a sixteenth-note break, but if you’re playing slower, you might delay that to the thirty-second-note level with some success. So long as it doesn’t sound panicked. Ditto the second eighth note of 19 and the final eighth of 38 and 39 and isolated notes in the right hand of 52-54. Ditto lots of individual Pedal notes in measures 2, 4, 9, 26, 28, 40, 43, 45, and 65.
Measure 13: I believe the final A-flat in the left hand should be A natural, since there are two A naturals against it in the right hand, plus the A-natural quarter note in the Pedal. Crunchy chromaticism aside in this Prelude, outright atonality or the prolongation of what sounds like a sour note is probably a misprint.
Measure 27: According to Dupré/Steed, the final G of the right hand should be G-sharp. I concur.
Measure 33: Curious that Dupré does not reverse the Pedal couplers here like he did for a similar texture in measures 16 and 51. That’s probably because in 35, the right foot will be more melodic there than textural before. By 43, the Pedal is back to ‘normal’ in this configuration. In 60, the Pedal is once more melodic, but with the Récit coupler rather than Great. I doubt any of this is a misprint, but it is nevertheless noted, and the performer may have to do whatever it takes to preserve good balance. As we all know, the lower end of Pedal eight-foot flutes and bourdons can be notoriously weak in the U.S., so you may have to go into various contortions to keep the upper Pedal notes from being too loud and the lower notes from disappearing completely.
Measure 35: The right hand is on its own for legato here. I don’t try to disguise wide intervals with what my teacher Clyde Hollway called the lunging, “omigod” way of trying to achieve legato where legato is not possible. So I just lift the thumb proudly, note by note, until the intervals shrink back down to something more manageable. That is even more pronounced in measures 59, where the white/black key pattern makes detachment even more necessary. I’ll take a proud, gently detached sound over a panicked lunging-for-notes sound any day.
Measure 43: The first eighth (E-flat) of the left hand is not the same interval it was in corresponding passages in measures 2 and 15. I wonder if that E-flat should have been a G instead. If so, then the entire passage (42-43) would match its sibling measures 1-2 and 14-15, interval for interval. For the record, I have not changed that E-flat in performance; I just raise the question here. Misprint?
Measure 47: Right hand, sixth sixteenth: that should probably be F natural to make a more 'natural' resolution from the E natural just before it. (F had been flatted on the third sixteenth.)
Measure 56: The Pedal is engraved strangely. Okay, this is going to be painfully wordy: The half rest is positioned in the measure in such a way that it looks like the entire Pedal should take a rest for the second half of the measure. I happen to feel that the Pedal should do just that, but the note and rest values are incorrectly displayed in that case. Look just above in the left hand, where a quarter rest is employed at the same spot for a voice there. On one day, I thought that the half rest in the Pedal applied to the right foot, which had been resting since the previous measure. But if that were the case, then the rest should be located dead center of the measure left to right, which it is not. I looked again at Dupré's use of half- and whole rests throughout this Prelude, and he employs a whole rest when a voice rests for an entire measure, even when a measure is only two beats long (measures 1, 22-25, etc.). Therefore, I feel that the half rest given in 56 is indeed representing half the measure but should have been given as a quarter, and I feel that the low G should in turn be a quarter note. Furthermore, the sudden upstem of the low G there is better explained by the right foot no longer needing any notation at all. This will all create a nice decrescendo, dropping voices in left hand and Pedal, until we arrive at the gentle coda of 57.
Measures 69-70: The left-hand stab accents are a bit jarring. Was Dupré just asking for an early release? If so, he might have used a staccato, as he did at the beginning of 69. And I don’t know how much “accent” one should be expected to get from such a light registration. Was Dupré asking for only the top note to be released early and the other two notes of the triad to move legato? Who knows? I’m still experimenting.
The Fugue is a nightmare for finger/voice independence. Make sure you’re obeying Dupré’s staccato and legato markings exactly.
Measure 109: Third beat, I move the right hand to the Swell as instructed, but I move the left hand to the Positif, for greater relief of the subject. I’ll move the left to the Récit for the downbeat of 115.
Measure 115: The soprano is legato, but not the second soprano. This legato lingers all the way into the downbeat of 123, including the manual change.
Measure 124: The pedal is still legato, which suggests a full-value second beat in this measure, against the staccato second beat of the manuals. So far, that hasn’t sounded ragged to me. But depending on the organ, that low G might 'stick out' unnecessarily if it isn't shortened. Similar mismatches occur in the B Major fugue, so it’s not unheard of. But in that piece, Dupré’s exacting markings and performances are testament that they were correct.
Measure 134: The left hand note values are incomplete. The final E-flat should be an eighth note.
End of measure 137: Note that both soprano voices are now legato, all the way to the rest in 143.
Measures 149-157: Legato here, staccato there. Watch Dupré’s markings, especially for the isolated staccato voices. And release legato voices exactly on rests, not early.
Measure 158: The downbeat is still legato, which means full value to the rest.
Measure 159: Second and third beats: notice that Dupré has written a long note value there, asking for a later staccato release.
Measures 167-173: Hardest part of the piece to memorize, if you ask me. The voice independence is a nightmare.
Measure 179: I move the left hand to the Positif on the downbeat of 179, rather than 180, so that I don’t have to negotiate a potentially awkward manual change after sextuplets. Third beat of 179: I move the right hand to the Positif, one beat earlier than instructed. Since the fugue subject begins on the 3rd beat, it just makes sense there. Curiously, Steed/Dupré says that the left hand should remain on the Great there to the end of the piece. That doesn’t make sense to me until 190.
Measure 189: I move my right hand to the Great a beat early. Although that is not a subject statement, moving on the third beat of 189 would retain the third-beat start pattern.
Measures 197-204: I find it dangerous for my hands to work so hard playing all those octave-spread chords. Those bouncing/traveling chords would all be hard enough to play, anyway, even without the Pedal scurrying up and down. I redistribute the notes and leave the hands in one position. That introduces issues of finger independence but it removes the issue of hitting cracks with so much quick repositioning from chord to chord.
Measure 218: Dupré didn’t have that high C at St-Sulpice. If you don’t, either (and if you’re in the U.S., then why not?!), then solve it any way you like.
The Great upper work and reeds, the Pedal upper work and reeds, and the manual 16s are not indicated at the beginning of the Fugue. Other than pedal coupler adjustments and a few (really short) hairpins and some dynamics indicating box position, Dupré does not specifically call for registrational buildup except by way of his fff in 196-197, which to the French meant simply ‘full organ.’ But surely he would allow some sort of buildup along the way from his opening registration and indicated box position of mf. I suggest some growth at the third beat of 137 (which may have to be reversed somewhat in the Pedal for 161-173). Depending on the size of the instrument, I suggest further growth at the third beat of 173, the second beat of 184, and the third beat of 189, plus Dupré's required full organ at the third beat of 196.
Measure 196: Unless the 32s muddy the Pedal scales, I don’t see the point of holding those in reserve for later. In any case, I have full organ on by 212. I feel the pregnant pauses don't need further buildup, since they are the only pauses in the entire piece, other than the break between Prelude and Fugue. We'll get the message!
Measures 213-215: People add their own interpretation of the pregnant pauses. Graham Steed says not to lengthen those nor the quarter notes in 213-214. Dupré even adds a breath mark in 213, to remind of the staccato directive. If everyone always followed Dupré's and Steed's instructions, perhaps everyone would sound alike. They would certainly sound like Dupré himself, who was the ultimate technician and micro-manager of every note. Then again, not everyone plays this piece, and so the risk of monotony in interpretations is low. And with the acoustical joyriding these guys engaged in during the final measures of countless pieces, further revving of that particular engine is irresistible! But I don't give in -- I keep it all in tempo, and I hold the final chord for a countable two measures and a downbeat (but who's counting?).