I thought of this topic while waiting for the elevator and listening to part of Grieg's Holberg Suite in the orchestral version(the part that used to be the theme for the I Remember Mama TV show). This piece would be a case where I think the orchestral version is better.

So I'm interested in pieces that you feel either work very well or very badly in the orchestral transcription(and why if can explain this).

One I remember hearing and being shocked by was Schumann's Carnival. I thought it sounded terrible, but maybe because one get's so accustomed to hearing a particular(in this case piano) version it becomes hard to judge a transcription independently?

I've always felt that Mussorgsky's "Pictures at an Exhibition" was too big for piano, and that the orchestral version is more satisfactory. This could be due to the fact, however, that one is inclined to hear the orchestral version more than the original piano version, and one thus becomes accustomed to the varied orchestral timbres that bring more colour and life to that version.

On the other hand, as it were, orchestrations that don't work :
- some Debussy piano pieces transcribed for orchestra - Clair de Lune, some of the Preludes, Golliwog's Cake Walk, etc., just don't work for me because the whole essence of these pieces is pianistic.

- Don't get me started on orchestrated versions of Chopin! "Les Sylphides" is an abomination, and only someone who has no understanding of Chopin's music would conceive of the pieces in it as suitable for orchestral transcription. But that's just my bias!

Regards,

One orchestration I always enjoyed is Dvořák's Op 46 G minor Slavonic Dance. This may seem a bit generic, but the use of different instruments just adds so much color and character to the music, and those full-out boisterous double G-major/minor chords are so much fun. They seem more controlled coming from an orchestra, whereas in the 4-hand setting, it's a fine line between 4 hands generating a nice, full ff, as opposed to it sounding like "banging".

Daniel

I once listened to a transcription of Debussy's masterpiece La Mer for 2 pianos. Objectively it was probably OK, but this is such a symphonic work with a masterful use of orchestral elements that the piano piece sounded puny and frankly disagreeable..

In general, I don't think music by Debussy and his Impressionist contemporaries admits well to being forcibly removed from it's intended instrument. Debussy wrote La Mer for orchestra for a reason; same with his piano music.

I think part of it is how deeply embedded in our consciousness the piano version is. For a crazy example, orchestrating some of Beethoven's piano sonatas would make better music than quite a few symphonies that were written for orchestra in the first place. But I still wouldn't accept it, because the piano sonatas are so much "part of the world" in their original form.

Which Sonatas do you have in mind?

I never understood what made Liszt arrange the wanderer-fantasie for piano and orchestra. Imho it is total travesty and done by a man who loved the original work, quite a paradox!
Ravel's arrangement of Mussorgsky's pictures at an exhibition is more succesful but I'm afraid I will never like it as much as the original which I acquainted at first hand
I can't think of any other examples right now but there are quite a few arrangements the other way round - orchestral works transcribed for orchestra - and I'm not a fan of most of them either.

Well, there have been enough bad symphonies foisted upon the world that any one would do. Let's say Op. 57. I'm not saying it would be good to orchestrate it - I'm saying it would be less bad than bad stuff, that's all.

I understand what you are saying, but I have heard probably about half of the orchestrations of the Debussy preludes that Colin Matthews has done, and they seem to work quite well, at least to my ear. Of course, it is strange to be hearing the music in such a different form, and I can understand why some people wouldn't take to it.

With Ravel, it's sometimes hard to say what the intended instrument was, since he provided both piano (usually 4-hand or two piano) and orchestral versions of some pieces. I can't think of another major composer with such a high percentage of works that appear in both an orchestration and some sort of piano version, both done by the composer. I'm sorry that he didn't orchestrate Gaspard, though. That would have been very interesting. Judging from the recording I have, the Marius Constant orchestration of it is not to my taste.

Can you expand more about why you think the essence of those particular pieces is pianistic?
(I haven't heard the orchestral version of any of them so I canlt comment on whether I think they're any good).

One of the great virtues of the piano sound - although it's occasionally considered a limitation in some contexts - is that it cannot sustain indefinitely, nor even briefly, without immediate decay. Listening to the opening chords of the original version of Clair de Lune, for example, where the essence of the sound is the decay of the overlapping d-flat major chords, following the initial pp attact, one realizes that violins, even con sordino, cannot create this mystic suspension and its gradually diminishing sound. In an orchestrated version, the sound of violins sustaining those chords leads to a total loss of the magic - or even mystical - quality of that opening.

Conversely, if a bass note is plucked by a harp, a cello or a double-bass, it decays too quickly and doesn't sustain long enough, yet a sustained bowed note doesn't have the same effect as the pedal-sustained bass of a piano.

It's for reasons such as these that give a unique sound to piano works that I feel that orchestral transcriptions of some piano works fail completely.

Regards,

Excellent points, all. That would have been interesting if Ravel had orchestrated Gaspard.

You mentioned Carnaval having been orchestrated--to which orchestration are you alluding? I've found the surviving fragments of Ravel's orchestration of Carnaval very interesting and compelling!

Unfortunately I don't remember. I only heard it once on the radio.

Very interesting and I never thought about that. Very recently, I've started playing Lullaby for Our Daughters by Mike Garson.(The music will be available at his website fairly soon)

http://www.youtube.com/watch?v=qW__f8K0THQ

I think that just like and for the same reason you mention about Clair de Lune, this would not sound as beautiful transcribed for orchestra. Do you agree?

While skimming through Howat's new book, "The Art of French Piano Music", I just found something that seems quite related to your thought re Clair de Lune:
"'Pagodes' suggests that it was the gamelan tha showed Debussy how to embrace the piano's intrinisic percussiveness in creating a gamelan-like carpet of sound, its varied reverberation and tonal decay working in the music's favor.

Debussy had heard the gamelan first in 1889.

I'd be interested to hear transcriptions of Debussy's Preludes. The second book strikes me as more orchestral than the first, and I'd expect all of them except for the first two ("...Mists" and "...Dead leaves") and the last two ("...Alternating thirds" and "...Fireworks") to sound fine when orchestrated. For example, when I play #3 ("...The Puerto del Vino") I hear mostly strings, with the mysterious melody played by something like a bassoon, or a funky oboe that can go lower than a regular one. With #4 ("...The fairies are exquisite dancers") I hear woodwind with subtle string support. I might ask for the opening quintuplets to be played by 3 different instruments, one for each beat, flowing from one to the other (say, flute, oboe, clarinet), as 3 different fairies. The "sans rigueur" and "caressant" melodies would be played by violins. #5 ("...Heaths") = strings, with woodwind support for the playful bits. #6 ("...General Lavine - eccentric") = brass and woodwind, no strings. #7 ("...The terrace for spectators of moonlight") = string orchestra with woodwind for the descending, winding lines. #8 ("...Undine") = strings and woodwind, with a harp for the last page. #9 ("...Hommage to S. Pickwick Esq. P.P.M.P.C.") = full orchestra. #10 ("...Funeral urn") = string orchestra, with woodwind chiming in for the mysterious winding bits. I'd imagine such arrangements sounding as good as the solo piano originals.

I'd also imagine quite a few of Albéniz's Ibéria sounding good in orchestral versions. And I've read that Respighi orchestrated some of Rachmaninoff's Etudes-Tableaux with the composer's permission and approval. I haven't heard any of these yet, sorry! For now, all I can say is that I prefer Pictures at an Exhibition with the orchestra, for the variety of timbres and bigger sound over the piano, and that I also prefer Ravel's own orchestration of Pavane pour an infante défunte to the piano original, as to me it sounds more sustained and tender and sad when played by an orchestra.

I heard the a minor one(with the opening triplets). I can't remember how much I liked it compared to the piano version.

Is there an orchestral arrangement other than Ravel's 4 pieces?

I'm with you there. Utterly dreadful tosh.

I've heard them; nothing really striking- just sends one back to the originals. Reportedly Rachmaninov was first approached to do the orchestrations, but -with good reason it would seem- shrugged it off.

And I must be one of the few on the planet who actually prefers -by a substantial margin- the piano original of Pictures.

Regarding Pictures at an Exhibition, I've been working on it for 10 months now, in sort of a love/hate relationship. I think there's plenty of cause for Horowitz' rearranging it for piano because Mussorgsky wasn't a great pianist and clearly didn't write like one. I think Ravel's transcription works really well for some of the pictures, but not for all of them. In their orchestral guise, I particularly dislike Gnomus & the Ballet of the Unhatched Chicks, and I'm not wild about The Marketplace at Limoges or the more wild parts of Baba-Yaga. The main problem I see is that most orchestral instruments take longer to generate full tone (unlike the piano, which instantly generates full tone). Because of this, quick, sprightly sections like gnomus and chicks in their shells tend to lumber while each note is held long enough for sound to resonate from the more sluggish instruments - these sections thereby lose their character completely. On the other hand, big, proud sections like the various promenades, bydlo, and the Great Gate of Kiev are at least as good if not better in orchestral version. One of my long term projects I've been considering once I perform Pictures in its original guise is a piano recomposition to address the deficiencies I see in the work. I think this approach will be more effective than orchestration.

Regarding Beethoven Sonata orchestrations, when I took orchestration during undergrad, the final project was to orchestrate some short piano work to be performed by the student orchestra on the last day of class (sight read of course, since there were 20-some students in our orchestration class). I orchestrated the second movement of Quasi una Fantasia, and a pianist friend of mine orchestrated the 1st movement of Sonata No. 9. Of course, we both preferred the piano versions, but it was an interesting exercise, and I think Beethoven is especially suited to this, because of his "theme fragmentation" sort of approach to composition. Another student orchestrated Csikós Post by Hermann Necke, which was probably the most successful transcription of the class.

The Weber Invitation is as good as Berlioz' transcription. Hindemith's Symphonic Metamorphosis took Weber's incredible harmonies out of the parlor and into the symphony hall.

I always think of Mozart's sonatas being played by a string quartet. It helps me bring out the singing qualities to his melodies. In terms of orchestral transcriptions, I haven't heard very many besides Pictures. Orchestration really highlights the majestic quality in the Great Gate of Kiev.

Brahms wrote the 2-piano version of the Haydn Variations first and then wrote an orchestral version. Now I don't if he was already planning to write an orchestral version while he was writing the first version, but this is definitely a case where the orchestral version is vastly superior IMHO to the original.

You can hear the Debussy preludes as orchestrated by Matthews - they have been recorded by the orchestra that commissioned them. I know that the Belgian composer Luc Brewaeys has also orchestrated them, but haven't heard his version. Danish composer Niels Rosing-Schow has orchestrated over half of them for a small ensemble, but I haven't heard these, either. I have no idea how many more versions there might be.

Albeniz' Iberia has been orchestrated, too. Apparently Albeniz himself orchestrated one of the pieces, and may have composed some or all of them with eventual orchestration in mind. They certainly seem like it. I especially like the orchestrations Carlos Surinach did. Earlier, Enrique Arbos has orchestrated a good deal of Iberia, and Surinach completed the set, which has been recorded . Someone else named Peter Breiner did the whole suite, and a recording of it is available on Naxos; I haven't heard it, so can't comment.

Your mention of the Rachmaninoff reminded me that I have a recording of orchestrations of his Corelli Variations and the Trio élégiaque. The trio was turned into something like a piano concerto. Somebody should do that to the Brahms Piano Quintet, I think.

Which takes us to a whole other genre, that of orchestrated chamber music. Of course, lots of string quartets get played in string orchestra versions, because it a pretty simple thing to do. Other, less obvious arrangements can be interesting. Schoenberg did an orchestral version of the Brahms G minor piano quartet; lots of people like it, but it sounds sort of clunky to me. I can imagine it could done better. Lucian Berio did a gorgeous arrangement of the Brahms f minor clarinet sonata, which retains the clarinet solo part and orchestrates the piano part.

I think the whole thing was orchestrated for a Diaghilev ballet, with various sections done by Glazunov, Rimsky-K, Liadov, and A. Tcherepnin.

Interesting. Is there a recording?
Thanks.

I have recordings of both the piano version and the orchestral version, and I came to the same conclusion several years ago. Some movements work better with orchestra, mainly the ones based on the main theme like Promenade and the Great Gate at Keiv, while others work better on piano, like (as you pointed out) Gnomus.

@wr: Thanks for the info. The Debussy and Albéniz recordings are joining my "would be nice" CD list.