Aleksandr Nikolaevich Scriabin: Sonata no. 3 in F# minor, Op. 23 - Played by the Composer
HTML-код
- Опубликовано: 5 фев 2025
- Since Scriabin was born in 1872 (Western Calendar), we thought we should upload his Hupfeld music roll recordings to the internet before his 150th anniversary expires. On 27 January 1908 the composer recorded fourteen rolls for the Ludwig Hupfeld company in Leipzig, initially published in the 73-note Phonola series, with serial numbers from M13426 to M13439. By the time of the Hupfeld General 73-note Catalogue of September 1912, the main classical series was already up to M14779, so one must assume that the Scriabin rolls were first published a good two or three years before that date.
Hupfeld was able to record notes, and presumably pedalling, as its pianists played, but It is not clear that it had any sophisticated method of recording dynamics. Certainly in its earliest recording correspondence, from late 1905 onwards, it asked pianists to send dynamic indications by mail, after their recording sessions were over. What that means - an annotated score, written explanations or descriptions - is anyone's guess. There was a dynamic recording patent taken out in 1906 in Leipzig, by Walter Bernhard, which looks vaguely similar to the written description of Hupfeld's practices provided by Ludwig Riemann in his 1911 book, "Das Wesen des Klavierklanges", but in 1906 Bernhard was working for Hugo Popper, an important music retailer in Leipzig who was facilitating recordings for the Welte-Mignon, manufactured by Hupfeld's main competitors. If you are interested in such details, you can find articles on Dynamic Recording for the Reproducing Piano in some of the later issues of the Pianola Journal, which are all available for free download at our website, on the Journal pages of www.pianola.org.
In Scriabin's case it is mainly rolls without automatic dynamic coding which have survived, and it makes sense to use those and to do the best one can to provide a sympathetic portrait of this most intimate of pianists. Once all twelve videos have been uploaded (the two rolls each of Sonatas 2 and 3 are each presented as one video), then they will be included as a playlist, in alphabetical order of title. Each video contains the same detailed scrolling credit at the end, but you don't have to read it every time! It seemed better to include it in each video, since one never knows how visitors to our channel will have discovered it.
Grateful thanks are due to a number of musicians, academics and enthusiasts, as follows:
The late William Candy, professional player-pianist for Hupfeld in London
The late Pavel Lobanov of the Scriabin Museum in Moscow
Anatole Leikin, Emeritus Professor at the University of California, Santa Cruz
Simon Nicholls, Co-Chairman of the Scriabin Association
Denis Hall and Adrian Church of the Pianola Institute
The rolls used for these video recordings all come from a later 88-note edition published by Hupfeld in the early 1920s, with different serial numbers - all listed during the course of each video. They currently belong to Rex Lawson, who was given them over forty years ago by Bill Candy, mentioned above, along with many other rolls from the Aeolian Company, Ludwig Hupfeld and others. Like many such rolls, it is important that they should remain available in some public way, especially as their current owners are failing to grow any younger! The pressure of age and available time is the main reason why these Scriabin videos are based around a single image, but we hope that the photograph of Aleksandr Nikolaevich playing to an intimate group of musicians and engineers will complement the composer's performance style, even though the recording session depicted is for the Welte-Mignon, and located in Moscow in 1910, rather than Leipzig in 1908.
Rex Lawson
Scriabin's music plays itself dynamically. The phrases imply the dynamics. This detached performance implies what is there anyway. This is great music and can survive detachment in the phrases..
Unglaublich faszinierend und packend! Danke!
One of the greatest recordings in music history, such a performance!!!
Спасибо вам большое за эти бесценные архивы, что вы публикуете! Это потрясает и будоражит дух!
The improvisatory feel of the sonata is absolutely magical, even given the limitations of the recording. Fascinating!
Incredible how different most interpretations are to this!
This is better especially that 4th movement 🔥
@@TheLifeisgood72 I tend to agree - I found the first movement a little unusual but I do wonder whether other performers listened to Scriabin himself playing (?) - a must, surely, if the recording is available??
@@billyfisher1539 Firdt movement has some weird rhythms but if he listened to himself it would probably get fixed
@@TheLifeisgood72no: this kind of rhythmic freedom was considered normal and desirable at the time, and can be heard in many other historic recordings. Robert Philip and others have written about this. Rhythms were not so much meant to be taken as literally notated, but were accentuated, reshaped etc for optimal expressive effect. This often sounds unusual to us today, because tastes changed over the course of the 20th in the direction of more literalistic interpretations. So what Scriabin is doing here is quite the lost art.
@@timothytikker3834 Even in his own day audience complained about his lack of rhythmic accuracy, and Scriabin himself did not consider himself an effective performer. The timing is off in many places in the first mvmt.
He didn't play the last chord up one octave higher. He hit the last chord, but then played another chord up one octave to finish. I guess this was improvised. But he can do that he is the composer.
No! How dare he alter the notes by the composer. A total disgrace to Scriabin!
Fantastic
Yeah, This is the coolest thing ever
The piano roll starts slightly too fast and speeds up until movement 3, where it restarts at a good tempo(it's really two piano rolls) and then proceeds to speed up again until the end. The dynamics aren't accurate either as with all piano rolls.
Aren't accurate relative to what?
"as with all piano rolls." - once again, to make such a flat statement one has to establish the starting point. What is it?
@@gaiusflaminius4861Because this is a reproduction of a single performance which happened in 1908, there actually are “correct” dynamics that you’d want to hear which are the dynamics Scriabin played in that performance. The piano roll system in this case has its dynamics controlled by a pedal which the Pianolist operates, so it’s the operators dynamics and not Scriabin’s.
@@TheLifeisgood72 It may be inaccurate, but to state with certainty that something is correct you have to witness it. You haven’t attended Scriabin’s concerts, so you don’t have the firsthand experience to use it as a reference point. You realise that subconsciously, therefore you quoted the word "correct". The text is no aid either because composers often don't follow it. In the case of Scriabin, the perforations on these rolls show the articulation. Reproduction piano rolls, as opposed to pianola, have dynamics coded into them. The late systems such as Ampico, DuoArt and Welte T100 provide a high level of fidelity and precision, which becomes evident when comparing it to acoustic recordings where available. Acoustic recordings, for that matter, are less reliable than piano rolls.
So, the claim you made about “all piano rolls” doesn’t hold up.
wow
Very interesting, right from the horse's mouth!
16:38
Are the Triphonola rolls of this Sonata (and the other Scriabin's rolls) re-editions of the ones he made in 1908 or just a nomenclature differing from that of the Phonola catalogue? The relevant numbers are, according to Petrak's roll listing and as indicated in the image of the roll's tail, 54037 and 54038 as opposed to M13xxx.
Good afternoon!
At the age of 76, my rule of thumb about history is that it becomes less accurate the further away it is from its origins. That is particularly true in the case of perforated music rolls, where the willingness of young musicologists and their supervisors to accept uncorroborated information from coffee-table books, simply because they have an ISBN number, is both staggering and depressing.
I don't know how much RUclips allows external links to be included in these comments, but you'll find an article on dynamic recording for Hupfeld's player pianos in Pianola Journal 21, which, like all the other Pianola Journals to date is available for free download on pianola.org, if you seel it out.
It's my article, and I did the best I could at the time, but there really is very little to go on. It is clear enough from surviving photographs that Hupfeld transmitted (and therefore recorded) its note data pneumatically, which is not the wisest thing to do, since pneumatic communication runs at the speed of sound, whereas electrical contacts are as instantaneous as light.
However they achieved their dynamics, it doesn't mean that there was some magician who could take a dynamic of each note and turn it into a perfectly accurate roll. You would have needed a player piano mechanism that had separate dynamic decoders for all 73 notes of the Phonoliszt or all 85 notes of the Dea, and prior to dedicated MIDI recorders, which were eighty or so years away, there were no such mechanisms.
All reproducing piano companies had to do the best they could, by employing musicians who could somehow find ways of turning whatever dynamic information they could obtain, and converting it into the relatively simple two track coding (bass section and treble section) which is all the pianos of the time could reproduce. Given that many Hupfeld pianists were asked to provide suggestions of their dynamics after that had returned home, it doesn't bode well for Alexander - you'll need the Hupfeld sections of the Saxon State Archive for that, and I'm not about to try and provide dozans of links, because this reply will fill the entirety of the page!
In any case, Hupfeld's main activity as far as the likes of Scriabin was concerned was to provide rolls for the Phonola, which was the most affordable instrument in its range, and which was intended to be pedalled by foot. In Grieg's own diary he talks of recording for the Phonoliszt, and very specifically not the Dea, and the automatic dynamics for the Phonoliszt were a good deal simpler than those for the Dea.
This reply is really not the place for detailed musicological discussions, but with a small amount of detective work you can find my email address easily enough.
I hadn't replied to TheLifeis good72, because most player piano enthusiasts will know that any player rolls become faster, the more paper there is on the take-up spool, assuming that there is no pianolist to control the tempo lever. The Scriabin Sonata is on two rolls, so if you try to scan in some electronic way, then you will get the results that he describes. Since I was pedalling a notionally hand-played rolls on this occasion, I gave Hupfeld the benefit of the doubt and assumed that the tempi were the nearest we can get to Scriabin.
By far the majority of the world's music rolls were transcribed from sheet music to master rolls by musical staff, with the deliberate intention of involving a pianolist, and therefore providing just the notes. That's how Hupfeld started up, with its early 72-note Phonolas, and by the way I pity the advertising staff who had to try and persuade their fellow-countrymen and women that the hand-played rolls of the later Phonola were somehow better than both the Welte-Mignon, because they allowed the Phonolist to create his or her own dynamic performance, and also better than the Pianola (and many other US and European makes), because the hand-played Phonola provided the rubato of a musical performance, allowing the Phonolist to concentrate on their foot-pedalling!
Then we have the post-war transcriptions from 73-note Phonola to 88-note Solophonola, Duophonla and Triphonola. My best guess is that Hupfeld had not kept its master rolls in most cases, and that they used the time-honoured solution of pedalling the old 73-note rolls and using electrical contacts attached to the dynamic equaliser of a foot-operated Phonola. That's roughly the way that Aeolian wanted to create dynamics for the Duo-Art, until they lost their legal battle tih Wilcox and White. I note that Hupfeld took out a patent for a similar machine, in France in 1920 and in Britain in 1922, though I haven't found the German version. So I don't think there are any dynamics on the Triphonola that qualify as holy writ, unless Hupfeld was using some other, unknown, equipment for new recordings in the 1920s. The sustaining pedal is another minefield, since 73-note Phonola rolls only have a fairly coarsely printed series of asterisks. This whole area needs more sensible research, but I'm not aure we shall ever get it. There are occasional 88-note Solophonla rolls that appear to have more sophisticated sustainig pedal tracks. We need musicologists who are less keen on their careers and academic titles, and more concerned about digging out the truth.
Cheers, Rex - Probably more replies in due course, and thank you for being sensible.
Krass wie die Qualität der Klavierausbildung in den letzten 100 Jahren zugenommen hat. So würde man heute nirgendwo mehr eine Aufnahmeprüfung bestehen.
Different centuries, different styles of playing musical instruments. I once heard a ten-year-old girl, the granddaughter of a friend, play a small piece by Grieg, exquisitely. Possibly the odd wrong note, but only Beckmesser cares about those! Then she went through the music college system, and the next time I heard her she no longer played with the freedom that Grieg in particular allowed himself. His piano roll recording of Sommerfugl/Papillon/Schmetterling on Welte-Mignon darts impulsively about, no doubt just like the butterflies that he saw from his composing hut by the local fjord. Many modern pianists turn it into a nuts and bolts wing machine, and even the best (since the best will have taken the trouble to listen to Grieg) replace the impulsiveness with a tidier, scaled down version, full of perceived good taste and Steinway "D"s.
About fifteen years ago I gave a small Pianola recital at a friend's house in north London. I included some Grieg, though I can't remember what it was. Afterwards an elderly lady found me, and told me that she was Norwegian, and that her grandfather had been a medical doctor in Bergen. She said that Grieg had died in her grandfather's arms. I took her hand in mine. These piano rolls are something like the lady's memory - the nearest we have to touching and thereby thanking the musicians of past centuries.
Thank you for taking the trouble to comment.
You gotta remember the Piano Roll isn’t doing justice to Scriabin’s playing… there’s no balance, dynamics
@@TheLifeisgood72 Piano rolls are very poorly understood nowadays, just as the attitudes of human society in the early twentieth century are difficult for all of us to comprehend. I've been involved with piano rolls for something over fifty years, and I still find myself coming across new details of how they were made, how they were perceived and so on. Many people nowadays don't even seek to understand them in any great detail, and it is a bitter pill to swallow, accepting that they will be dealt with in a less exact way, as modern society changes. It's the same with all aspects of human society, in that the further away we are from a historical period, the more that history becomes a fairy tale.
Music rolls are portraits, not photographs, and furthermore they were not made with the intention of passing on musical performances to posterity. There may have been one or two pianists whose egos were flattered by the perception that they were recording for future generations, but the owners of the companies which developed these systems were in it for the money, make no mistake, and the market was aimed at those who were alive, not those who might be born in a hundred years' time. In the publicity of the time you find all sorts of lurid prose about the accuracy of the technology and the almost god-like attitude towards the pianists (and organists) involved, so if you are trying to understand the reality of those times, you cannot trust the advertisements. You also can't trust most of the books written about the subject, especially the modern ones, because there has been such a dearth of accurate information over many decades, and therefore those who write the modern books have very little on which to base their writings, especially because busy professors and musicians so often don't have the time to carry out original research.
Arguably, the man who invented the Ampico reproducing piano, Charles Fuller Stoddard, was the greatest liar of the original period, or certainly the most successful, because he knew how to colour his accounts so that they seemed very believable. I have high-resolution scans of a "Salesmen's Manual" for the Ampico, in which staff are told how best to lie about the competitors' systems. But Hupfeld was not so different, and it glossed over the methods of recording by means of careful advertising. A good selection of the correspondence from the Hupfeld company to its recording artists has survived, thanks to the desire for preservation that the communist regime in East Germany upheld. It's very clear that in the early days (1905 to about 1910), Hupfeld had no method of recording dynamics, because they asked their pianists to send them letters detailing the dynamics that they wanted, after the recording sessions had taken place and the pianists had returned home. The replies have on the whole not survived, but how on earth is a written letter, of less than a hundred pages, likely to contain a detailed analysis of the individual keystrokes of a complete recording session? In any case, pianists of that time had no personal means of measuring the speed of their individual keystrokes - the whole ethos of the time was far less exact and much more emotional.
The only dynamic recording patent taken out by Hupfeld dates from the early 1920s, and it consists of a method whereby a player-pianist's pedalling could be converted into levels of suction that could in turn be used to mark up a wavy dynamic line on a music roll. Two foot-pedals and an equaliser with a spring would result in a very wavy line indeed, as the illustrations on the patent pages testify, and any similar mechanism derived from the expression box of a reproducing piano would produce a similar result, whereas the printed lines on Hupfeld rolls are straight lines with corners, which would be impossible to create by anything other than an editor's decisions. There was a very similar patent awarded to the Wilcox and White Company in Meriden, Connecticut, for use with the Artrio-Angelus system, but that company failed and was bought out in the early 1920s, so in all probability Hupfeld reckoned it was safe to apply for European patents at that stage. The Aeolian Company, for its Duo-Art, originally wanted to use a similar system, and a photo from 1915 has survived at the Library of Congress of the main roll editor (W. Creary Woods) using a push-up Pianola secured to a second recording piano as Paderewski played in the foreground. There was a long drawn out litigation process between Aeolian and Wilcox & White, which Aeolian eventually lost, at which point Aeolian changed to the system of rotating dial knobs that appear in many photographs from both their New York and London studios.
I do the best I can with these rolls. I don't do it to become famous, rich or loved, to be better than anyone else, or indeed to enhance my career, but I do it in order to share the music and musical styles that I love. In most cases you can find other RUclips videos of the same rolls. I was lucky enough to have Simon Nicholls, Co-Chairman of the Scriabin Association, to guide me as I played them through, and in particular his wisdom over the thorny problem of Hupfeld roll speeds was invaluable. Hupfeld doesn't generally print exact roll speeds on its issued rolls, but instead it gives a speed range, between two numbers that are ten apart, and there has as yet been no definitive explanation of the paper speeds that these numbers represent. Simon and I decided that it was unlikely that Hupfeld would have used different roll speeds for rolls recorded in the same session, so we adopted a speed that we felt best matched all the rolls. The Poème, Op. 32, no. 1, is faster than the Welte-Mignon recording, and in the case of Welte there is generally not the same difficulty with overall roll speeds, though it is important to note that some longer Welte rolls run at 2.5 metres per minute, rather than 3 metres. But musicians play differently as the mood takes them, and the grouping of friends around Scriabin, as he recorded in Moscow, might have made him feel more intimate, in comparison to a less familiar recording studio in Leipzig, after having had to negotiate aggressively over recording fees. I note that Lifeisgood72 has used the Pierian CD track of that Poème on his channel, but has speeded it up by a little under 10% without altering the pitch, or has acquired it from someone else who has done so, and I'm curious to know what motivated that change. As I said, music rolls are portraits, and a portrait can sometimes create a better emotional atmosphere than a photograph, as you can see on page 21 of Pianola Journal no. 21, available free at www.pianola.org/journal/journal_vol21-22.cfm.
@@pianolainstitute What a great reply. A very interesting and enlightening response as a fellow piano roll enthusiast.
@@pianolainstituteThanks for the great reply. I think the 3rd and 4th movements are the perfect tempo, but the 1st and 2nd are a bit too fast. Maybe that’s why the other person said it sounds like an amateur plying. But the 4th movement is absolutely amazing, one of the best piano recordings on record.
There is often a case for composers' not playing their own music; this one clearly is. Of course, maybe the piano roll itself is at fault.
On concert tours he only played his own pieces
a shallow mediocre performance not worthy to be criticized by any scholar in music or piano.
Its literally the composer himself
someone calling the performance of your OWN piece "mediocre" has to be insane lmao
@@realityproof_0891 @GabrielDarioChaconMadrigal I haven't finished listening to this performance, so I can't judge it yet. However, there are composers known to have been mediocre performers of their own works, e.g. Busoni and Ravel.
@@realityproof_0891 @GabrielDarioChaconMadrigal I can't judge this performance because I haven't heard all of it yet, but there have been cases of composers who were mediocre performers of their own works, e. g. Ravel. Also, Busoni, despite being a great pianist, has a terrible recording of his own transcription of Bach's Chacone.
found rachmaninoff's alt account