Maurice Ravel's 'Unutterably Slow' Pavane: A Tempo Mystery?
Vložit
- čas přidán 12. 09. 2024
- In this video, we delve into the fascinating world of tempo reconstruction, focusing on Maurice Ravel's "Pavane pour une infante défunte." We explore the history and beauty of this piece, including different interpretations and performances by various pianists. We discuss the tempo discrepancies between the original 1899 piano edition and the later orchestral version, examining whether Ravel intended a slower or faster tempo. We also touch on the complexities of piano roll recordings and their impact on understanding historical performance practices. Join us as we analyze, compare, and appreciate this melancholic masterpiece, providing insights that enhance our listening experience.
--
📱 check out our recordings here: ▶www.authentics...
--
🙋Join our Patreon community and help us create more content▶ / authenticsound
--
📩Stay informed! Join our mailinglist (yes we have one too!) 👉bit.ly/as-maili...
It started as a dance in 1899, and became a lament by 1910. Why? That is the interesting question, not whether Ravel transitioned from 'wholebeat' to 'single beat' in that decade! It might also be worthwhile finding out how the Pavane was actually danced in the 17th century and how people at the end of the 19th century thought it was danced. You have to dig a bit deeper than Wikipedia, perhaps look at what some of the experts on baroque dancing have to say.
It would also be interesting to know what the official line on the metronome was in the Paris conservatoire, where Ravel trained (on and off) in the 1890s.
There is a recording by the wonderful Marcelle Meyer, from the 1920s, playing the Pavane, not at 80, but certainly over 70. It has a dance-like feel.
By a happy coincidence, Ravel's teacher, Gabriel Fauré, also wrote a Pavane (for orchestra), in 1887. The tempo on the orchestral score is andante, molto moderato, quarter = 84. A Welte-Mignon piano roll from 1913 exists of Fauré himself playing a piano transcription. You can find it on CZcams. Judge for yourselves the tempo he takes - it is certainly not wholebeat!
(By the way, the Welte-Mignon rolls are accurate as to tempo, dynamics, and pedalling.)
It's an interesting topic. In IMSLP there are two other piano works of Ravel later orchestrated by him. Menuet antique, piano, 1898, orch. 1930, same MMs (first section 1/4=76, second section 1/4=80); Alborada del gracioso (in Miroirs), piano, 1906, orch. 1920, same MM (dotted1/4=92). Is the Pavane a special case, something like Chopin op.10 no.3, played much more slowly because it sounds nice like that?
If the wholebeat hypothesis is correct, then the Menuet antique and the Alborada were sped up in the orchestral version exactly 100%, but the Pavane was only sped up by about a third. Again, why was it treated differently?
There goes the whole beat Ravel.
Great video, very interesting
do you remember Ravel’s “boutade” about the princesse being “défunte” and not the pavane… “Pavane pour une infante défunte” and not “ pavane défunte pour une infante” he would have said… Listen to Marcelle Meyer, Perlemuter, Casadesus, Jacques Février…
If you play the part where Richter is sped up to 80 (9:50) with your playback speed set at 0.50 that would be 40 in single beat. It really sounds nice and does not seem too slow considering the title of the piece.
Thank You
I don't think Ravel changed the way he used the metronome, he just changed his mind about the tempo over the years.
The original pianoversion (1899) doesn't have a tempo indication exept the metronome mark, it just says " pretty soft but with a broad sound"
The orchestral version (Lent) is from 1911 and the pianoroll from 1922.
Isao Tomita performed this on his "Bolero" album, and the tempo is slow, slower than the piano examples you used.
I have found your video about Liszt's transcriptions of the Beethoven symphonies very interesting. Have you ever made a comment on Henri Herz's piano school book (published around 1840, I believe). It seems that he advocates for the use of the metronome in the same manner it is currently utilized (both in the French and American editions of the method). Have a nice day!
I see you have made a 'short' excerpt of this video in which you claim piano rolls are nearly useless for tempo . This is a gross simplification.
The situation is complex, as it is quite different for every brand. Welte rolls (who recorded Debussy and some Ravel) had a uniform speed. The company insisted that the rolls should run at three meters per minute, which gives 120 round per minute for the motor. They sold test rolls and the owner of the player had to check regularly whether his machine still gave the required speed.
Ravels Pavane roll was recorded for Duo-Art, which has different playback speeds, but always gave the tempo at which the machine had to be set. They also had extensive regulationn manuals and test rolls to check whether the player was still giving the desired speed.
The accuracy of this procedure can easily be tested by comparing rolls and recordings of the same pianist in the same work, like Saint-Saëns or Granados.
So, in the case of the roll under discussion: if the recording machine and the reproducing machine are correctly regulated, the tempo you hear is the one played by Ravel.
You are very suspicious of piano rolls (4:36). The high-quality 'reproducing' piano rolls were meant to be played back at the tempo they were originally recorded. So if your roll of Paderewski was marked 80, you set the tempo lever to 80 on your Welte-Mignon or Duo-Art or Ampico, and you would get the accurate speed - rather like getting a 33' LP and playing it back with the record player at 33 (for those who remember record players, much less gramophones). I see no reason for any 'huge grain of salt', unless there was a conspiracy to represent everything being played much faster than it was in reality. And what would be the point? People could go to a concert and hear Paderewski, then enthusiastically buy a Paderewski roll for their very expensive machine, and compare the live performance with the reproduced version. If the reproduction was very much faster, maybe twice as fast, as the performance, huge rats would have been smelled, but there is no evidence that they were. Yes, on some systems wrong notes could be edited out (as they are on vinyl), but often they weren't, but that has no relation to tempo.
Similarly, why should player-piano enthusiasts who lovingly record these early performers and put them up on CZcams wish to represent them as playing much faster than they really did? Is there an international conspiracy to conceal 'wholebeat' from the world - rather as there is an international conspiracy of scientists concealing the fact that the earth is flat and does not revolve around the sun? I just don't see the point. A lot of music was performed both in a rather different style to what is acceptable today, and rather faster, and both piano rolls and early recordings are good evidence.
Ok, Julian Dyer defense advocate. Some rolls are somehow “awkward” according to him. What in the world is that? He should strive to define that, for someone as in charge (starting by demonstrating the actual playing speed, and not playing as he imposes).
@@AlbertoSegovia. ???
One usually is blithely unaware of the fact that speeds of the roll are not always followed accordingly, especially by people respected on the matters. For an example: the Hupfeld rolls are considered head scratchers because they sound too slow when played out of their system. But the equivalence between systems is very understudied and practically unknown. According to more recent research the number 50 for H. means 2 meters per minute (maybe Ramet was wrong). And in a reproduction from Mr. Dyer of Grieg’s work, without using the Hupfeld system, he plays at 70 instead of the indicated 50! Which is a 40% positive difference instead of the hypothetical 22%. In another work, he plays it at 80 instead of the marked 50, a 60% difference. He does not expressly show what that difference should be, as it is common: research is needed and people usually take what has been recorded by pianists as reference for reproductions while ignoring any problems with the metronome. But given the efforts of Granados and Grieg, one has to be interested in unveiling the acknowledged mystery. Also Nicholas Cook is skeptical: he said that one must dispense the elements that are not helpful from listening to piano rolls. Not to mention the issues with editions by engineers, etc.
@@AlbertoSegovia. Thanks. There are people who have studied these matters in great detail, but I agree it is not generally well understood. Also, I think there remains confusion about the difference between a Pianola, where the player could adjust tempo and dynamics, and the reproducing piano, such as the Welte-Mignon. As far as I can make out, Julian Dyer is talking about a Pianola-type roll, so yes, there could be uncertainty about Grieg's tempi.
@@DismasZelenka “I don’t care” what he’s talking about. That question should be solved already. But papers insist on weird expressions like: they are too slow , but never saying what ratio of slowness. And people who attack the problem from Grieg’s MM don’t fully maje the connection with this controversy.
Pavane has always felt really fast to me....More like a dance. (A celebration for the death of the princess? ) I'd play this slower then Whole Beat tempo. (if I had the talent)
as far as we know, from the author's own words, it has nothing to do with death..
@@3r7s Ahh.. Thanks. I'd still like something revolutionary....not just another pretty tune.
It was a dance - for a dead princess, that is, a princess who died 200 years earlier. He imagines her dancing when she was alive. Ravel is supposed to have written "“Do not attach any importance to the title. I chose it only for its euphonious qualities. Do not dramatize it. It is not a funeral lament for a dead child, but rather an evocation of the pavane which could have been danced by such a little princess as painted by Velázquez” - but I can't find any proper citation.
I agree with the editing of piano rolls. When MIDI emerged the sequencers "Cleaned up timing" with quantisation which completely ruins the feeling of most music. The ambiguity of tempo is best resolved with common sense. The other advantage of whole beat practice is that it _doubles_ the precision of tempo markings. Personally wading through WTC book 1 and playing without metronome or damper pedal but straight from the heart and or spirit (breath rate and depth)- and the tempo varies day to day depending on my mood! How many musicians perform their compositions _exactly_ to the note and phrasing with no latitude for improvement?
About thoses early recordings, on cylinders or rolls, they wete very limit in available recording time. It might be an incintive to push the speed a bit ...?. Ritcher is amazin in his interpretation.
Before even watching... I'm more informed about Renaissance to Baroque music. A title sugesting that someone complained about a Pavanne being to slow, this sounds like a joke 😂
But it's probably true... Let's watch.
My thoughts before watching as well.
Hahah the first thing I thought of was Tomkins A Sad Paven for these Distracted Tymes 😁
as to the piano rolls.. i'm sure the people reconstructing these recordings at the institutes would be able to answer your questions.. i don't know if you ever tried to contact them.. it's research, after all.
you mention the Scriabin rolls.. as far as i know, Scriabin had two recording sessions, both for different companies.. and as far as the most recent reconstruction of his Hupfeld rolls are concerned, there's nothing fast (let alone insanely fast) about them..
I think the title of Pavane is misleading, the dance was a 16th-17th Century phenomenon, the days of its performance were long gone when Ravel wrote the piece so its "danceability" tempo wasn't a factor. He was harking back to the impression of a time, which though over had a spirit and atmosphere he wanted to touch and evoke. When asked about the Princess Ravel replied that one should see her as the one "in the painting by Velasquez". That painting is Las Meninas ( Ladies in Waiting) the Princess is the tiny figure in the painting , Margareta Theresa the daughter of James IV of Spain and his wife, Mariana of Austria.
pavane /pə-vän′, -văn′/
noun
A slow, stately court dance of the 1500s and 1600s, usually in duple meter.
The music for this dance.
A musical style characteristic of the 16th and 17th centuries.
The American Heritage® Dictionary of the English Language, 5th Edition
@@Renshen1957 Thank you I am perfectly aware of the nature of the Pavane, the point I am making and which you have missed is that Ravel wasn’t writing this as a piece to which the dance was to be performed. The title is to invoke an age and a sense of that age, which is the reason for the “utterly slow” question, dancers would have great difficulties attempting to dance at that tempo.
@@russellbaston974 I think you are partly right. The original quarter = 80 tempo is danceable; that seems to about the speed taken by early dance specialists for Baroque pavanes (there are several on CZcams). This is slow and stately compared to a Minuet, where quarter = 120 or faster is normal, but if you tried to dance a Pavane at quarter = 40 you would have difficulty maintaining your balance.
The French composers of the late 19th/early 20th century were very interested in early French keyboard music, Couperin, Rameau etc. They were well aware of the different dance movements. They were also fascinated by Spanish influence and Spanish dance music - the 'infante' is a Spanish infanta.
I suspect what happened is that Ravel originally wrote it as a Pavane (not for dancing of course, but in that spirit) but, because of the beautiful somewhat melancholy melody it was performed more and more slowly, and finally the interpretation stressed the infante défunte rather than the dance, and Ravel came to accept this when he orchestrated it in 1910, and when he performed it on a Duo-Art piano roll in 1922. Regrettably, it was sentimentalized. (Perhaps Ravel himself regretted it: “Pavane pour une infante défunte” and not “ pavane défunte pour une infante”!)
Ravel might have thought the melody a bit more flowingly at first, but with a "quite gentle and slow sonority" (as the description says) - an Andante MM=80.
But then he changed his mind and slowed it down to 56, which he decided was the right tempo, more fitting to the memory for a dead princess.
If he had thought it very slow at first (MM=40), he would have written that straight away, and probably labelled it "tres lent".
There is, and has never been, such thing as "double beat".
If your double beat hypothesis was correct, and it probably is in some cases then there were composers whose intention was when they wrote M.M. q=100 to be 2 clicks of the metronome. And today I would think that all composers who mark M.M. q=100 to mean 1 q per click. Or BPM.
If that's true, then logically there should be composers who used both at some point. But why did they do that? And when? And under what circumstances?
According to this post, Ravel could be one of those "transition" composers that used both.
However, with Ravel's Pavane, we know two things:
1) Ravel once remarked "I wrote a Pavane for a dead princess, not a dead Pavane for a princess." So super slow may not have been his intent.
2) Vlado Perlemuter learned all of Ravel's piano works, played them for Ravel, and then recorded them. I cannot think of any better source for what Ravel wanted for the tempos of his works. You did not mention Perlemuter.
Everything you say about tempi versus everything most students of piano know by feel based on what they were taught are in stark contradiction, unless you add the caveat: Double-beat vs. single-beat metronome markings depend on the context. Many of those who have used the metronome to practice, especially to build up speed, know that sometimes you set it to single time, sometimes you set it to double time. Depends on the context.
If you accept both, that sometimes it's single beat, sometimes double, then everything falls into place. But you've taken an all or nothing approach to this doublebeat hypothesis, and that's why nobody is happy. But you've already invested everything into double beat, and there is really no going back on it. Kind of sad in a way. You've painted yourself into a corner.
I have read in a GPT that "Ravel’s use of the word “défunte” in “Pavane pour une infante défunte” doesn’t necessarily imply that the princess was literally “dead.” Instead, it evokes a sense of nostalgia, melancholy, and faded grandeur. The title suggests a beautiful, bygone era-a princess who no longer exists in the present but lives on through the music."
I've always liked Monique Haas playing Ravel. Though I haven't checked the tempo of her playing.
Glad you mentioned Scriabin, currently I'm studying his Fantasy Op. 28 and it seems odd to me that nobody play it at the tempi indicated by the composer on the score. Some of those tempi are unplayable and even if they were possible, those sections would sound ridiculous at that speed. Is there a chance that Scriabin was also a WBMP composer?
We have recordings of Scriabin playing and it's not whole beat
@@loren8888 daaaamn
yes
You are right that in the first edition (from 1899) the quarter note is marked with 80. But you did not mention (on purpose?) that there is also a "reprint" (also from 1899), that you can find on IMSLP. So why a reprint? Maybe there was an error in the first print? Maybe it was the tempo marking? At least it is not the case that the change was made only years later - no, it was changed in the same year!
It's more complicated. and IMSLP is a bit unclear. The first edition, with quarter = 80, was by the publisher Demets (1899). The re-issue was by Max Eschig, who took over Demets in 1922 or 1923. So Eschig simply updated the original MM to conform to the orchestral version, quarter = 54, in 1922 or sometime afterwards. 1922 was when Ravel's piano roll was made, at the slower tempo.
I don’t think it’s a mistake, but are we not allowed to disagree with the composer? Did not Rachmaninov say that Horowitz and others played his pieces better than himself? Maybe there are more ideal tempos for certain scenarios, it’s all opinion and subjective
Of course you can play any tempo you want. But if you want to come as close to the composers intention as possible, you have to respect the written tempo
So why aren't pitches "all opinion & subjective"?
You can play the way you want, but we should at least try to understand what the composer meant.
@@MGJS71 Because the tune depends on the relation between pitches, and (to a slightly lesser extent) on rhythm. If you change those you change the tune. But if you change only the tempo the tune stays the same (within limits - too fast and it blurs, too slow and you can't hear it; think of bird song on the one hand, and whale song on the other: you have to slow bird song down to hear eveyrhing that is going on, you have to speed whale song up to realize there is a song at all).
As performers we shouldn't let our ego affect our playing, it's generally what the composer wrote that's important
There was a great interest in France of earlier with the Paris International Exposition, The International Exposition of 1889, which both Ravel Debussy attended in which the Taskin Family,’s Harpsichord of 1765 performed music…an outgrowth of the interest in Baroque an earlier music and the Gamelan music of the far east. Ravel between 1914 and 1917 would write Le Tombeau de Couperin, a suite for solo piano by composed between 1914. Pavanes were also written as commemorative pieces such as an early example of William Byrd’s Pavane after the Death of the “Earle of Salisbury” which the title so names the deceased.
Was Byrd's Pavan and two Galliards named for the 1st earl, who died in May 1612, or the 2nd earl? It was published as one of the virginal pieces in Parthenia. The collection Parthenia (music by Byrd, Gibbon and Bull) was dedicated to Elizabeth, daughter of James I, on her betrothal to Frederick Elector Palatine, which was in December 1612.
A similar question about Sir William Petre, for whom there is a Pavan and a Galliard. The elder Sir William Petre died in 1613, but his son was also Sir William (he was a knight of the shire). Ambiguities everywhere!
You could cite Anthony Holborne's Countess of Pembroke funeral pavan, and John Downland Sir Henry Umpton's Funeral, for associations with death.
@@DismasZelenka To my recollection, the First Earl, Robert Cecil, of the fifth creation of the title….The Second Earl of Salisbury dies 45 tears after Byrd.
The first creation of the Earl of Salisbury, the original first Earl of Salisbury dates from the early part 12th Century.
The numbering system of the Earls of Salisbury is complex, based on whether one of the intervening Earls was either a subsequent Earl by a restoration of title or a subsequent creation which in there’s an Earl of Salisbury.
The title could be inherited through a daughter if no male heir were available…heraldry also allows or sometimes permits bastards to receive the title.
Another source credits the Pavane of Ravel to the success of Fauve’s Pavane in Paris, for Ravels Pavane and other ancient Dance Suite works by other composers, but Faure’s was written a decade after Saint Saens use of Gavottes. Saint Saens was transposing J S Bach Cantatas for Piano as well as an edition of Rameaus Harpsichord works. Similarly, Saint Saens was an organist of renown and his Organ Symphony was dedicated to the then late Franz Liszt(who composed organ works). Saint Saens stated the Organ Synphony was in imitation of Liszt style of composition. First performed at the inaugural performance of a new and-very large organ in England which sadly no longer exists. Brahms and Chrysander had published collections of F. Couperin’s Harpsichord Suites as well the works of Handel. Something for Wim, Brahms owned a clavichord, the previous owner was a Wolfgang Mozart. As to Brahms, his last compositions were Chorale Preludes for Organ. Mendelssohn also wrote works for Organ, too. See what a Bach revival will cause.
@@Renshen1957 I wish we had heard more from Wim about the interest in early French keyboard music, including suites of dances, in French musical circles in the late nineteenth century; I wonder if there was a parallel interest in the actual dance steps/movements among choreographers? It was of course part of the Francophone resistance against the overwhelming influence of the German schools of music.
The seventeenth-century 2nd Earl of Salisbury, William Cecil, was 21 when Parthenia was published (1612) and had just come into the title. People did not have protracted adolescences in those days - he had already married in 1608, at the age of 17. Was the Pavane and Galliard(s) named for the first or the second earl? Just on the evidence of dates, non liquet.
Very fascinating video! However, I personally have trouble with the idea that whole beat was being used as late as 1899. I think that the early composers of the metronome era used whole beat, but then as the 19th century progressed, performers were playing faster and faster, and basically metronome marks were ignored, and after maybe 1850 or 1860, people didn’t even know how Beethoven, etc. used the metronome. But since the performance standard was really fast by then, they assumed single beat. The commentaries on metronome marks that you have quoted from the mid-19th century onward seem to indicate a single beat interpretation. If there are conflicting metronome markings in the late-19th century, I think it is probably because of conflicting ideas (maybe even in the composer’s own minds) about how the music should be played.
Luigi Cherubini was one of the earliest users of the metronome, and on some of his religious music he gave a metronome indication and the overall time. For instance, O salutaris hostia no.2, 79 bars of 3/4, at quarter = 58, four minutes. This is not wholebeat! I don't think the understanding of how to use the metronome changed at all since its first invention - it was never 'wholebeat' - but, as you say, composers could change their minds, and performers could decide on different tempo interpretations. It is not a rigid situation.
@@DismasZelenka I agree that there is not a rigid solution. And in a sense, there is some de facto “whole beat” use of the metronome today. In 4/4 music, teachers will sometimes have their students set the metronome so that each tick represents an 8th note. Furthermore, I have an edition of Bach’s Inventions and Sinfonias in which the editor (Willard Palmer) compiled a table of tempos used in various recordings, and in a couple cases, he inadvertently listed the tempo in whole beat!
@@DismasZelenka Interesting example re Cherubini. I would think that religious music lends itself to single beat more than whole beat. The concept of whole beat revolves around subdivision, which becomes relevant in music with a lot of busy passages with fast notes. I’m not familiar with Cherubini, but picturing traditional religious music I know, indicating a tempo in whole beat would be kind of silly.
@@Graph1159 If you mean that religious music tends to be slow, wouldn't the same go for slow movements in non-religious music? And there are fast movements in religious music too. Take Cherubini's Mass in C Minor (1816), a work very much admired by Beethoven (he asked for it to played at his funeral, and it was). It starts with a Larghetto sostenuto, in cut time, half = 50 (so very slow). But then it also contains an Allegro maestoso in cut time, half = 88, and a movement in 4/4 Tempo a capella, Poco Allegro, which starts half = 120, and becomes piu Allegro, half = 132. Are all these in single beat, or is the Larghetto in single beat and the Allegros in whole beat? In that case, the Allegro maestoso would be slower than the Larghetto sostenuto!
@@Graph1159 In very slow movements you can certainly count the subdivisions. It is useful to practice fast movements slowly, possibly counting the subdivisions, but that does not mean you should perform them at those slow speeds. As for Bach, there were great disagreements about the tempi in the nineteenth century, as there still are today - sometimes as much as 100% or more faster or slower. Are you sure Willard Palmer was listing a 'wholebeat' tempo, rather than just a fast tempo?
1. Joke: this means that this is Ravel’s “La plus que lente”?
2. Ravel’s « Introduction el allegro » M.46 commissioned in 1905 has q=40 (Très Lente) in edition Durand’s from 1906. “France Illustration, le monde illustré” says in 1953: “se plaisent même à trouver dans la répétition fréquente des deux fragments mélodiques (…) d’une impression visuelle reçue par Ravel” ; “écrivait-il alors à son ami Maurice Delage, on ne voit que des moulins, tournant, tournant au vent”. The mills sound, to my opinion, too fast, or crazy as in a postmodern film, if one takes a too fast allegro in the 3/4 in respect of the initial marking of 40 or the other “less slow” 50. What is allegro here? One can check videos of mills spinning on Holland.
The internet is fantastic :p
(The magazine says that Ravel was inspired into this work by his travel along a river in holland.)
I could bet that a MM for that allegro would be very well around 80 (or 60), and not 172,