In recent weeks during August, I have worked on a new rendition: the final movement of Gustav Mahler's 9th Symphony, "Adagio". I got inspired to do this after listening to the BBC Proms concert this summer, where Roger Norrington and the Radio Symphony Orchestra Stuttgart performed this work (a quite positive review is here ). This particular performance created a split response from the audience: some loved the "no-nonsense" approach, based on the mantra of Norrington that in Mahler's time there was "no vibrato", others did not like the dry sound. Initially I was also a bit underwhelmed when listening to it on the radio, but when watching the TV performance and seeing the player and the orchestra, it made more sense, and I enjoyed it.
I also know the legendary recording by Bruno Walter in Vienna, January 1938, which to my ears also is a bit in the "no-nonsense" category, with the tempi not too slow but moderately stringent.
These are the only two recordings I knew, and with these in mind I began working on the iconic slow final movement, which has the connotation of "farewell", "good-bye", "Leb wohl", although none of that is written anywhere in the score or in any notes by Mahler.
Since this is a slow movement, it has not so many notes - just 17 pages of score. There are some technical difficulties, those 1/5 notes need to be properly done. I could have attempted playing them live, but since in these segments there are several layers of different note duration fractions overlaid, I decided to do a brute quantization there. I must say that I admire the real orchestra musicians who are able to play these difficult passages.
The overall mood of this music is to be "light", airy, with a chamber-music-like quality to it, despite all the brass and woodwinds (4 flutes!).
My rendition is here:
https://virtualphilharmonic.bandcamp.com/track/mahler-9-adagio-4th-movement
Apologies for a few issues with clipping - it is actually not audible, but I see it in the wave analysis. Has to do with simultaneous instrumental attacks which I need to spread slightly to avoid that initial peak.
I hope you like this recording, and I hope that it does not spoil for anyone the enjoyment of this wonderful music - if you do not like my version in particular, there are plenty of excellent "real" orchestra recordings out there!
In this blog I describe my activities in composing and recording music. Stay tuned for updates on new releases, for comments, and for background info about my recordings and compositions.
Showing posts with label 4th movement. Show all posts
Showing posts with label 4th movement. Show all posts
Wednesday, 7 September 2011
Monday, 4 April 2011
New Rendition: Mahler, Symphony #1, 4th movement
In my pursuit of creating good renditions of the orchestral music by Gustav Mahler, I have completed another step: A new rendition of the 4th movement of Symphony No.1 is now online. A few days ago I did release an intermittent version, in which I put more emphasis on the strings, similar to what I had done with the "Blumine" movement, but more consequent: I added two additional layers for each of the string tracks, not only for the violins. This now creates a fuller sound and adds more "punch" to those staccato fp string attacks that are prevalent throughout this movement.
After I had put that intermittent version online, I addressed once again the tempo. Editing the tempo through the currently available methods in sequencer software is quite simple: I can simply edit tempo values at any instance, or can draw tempo curves over the time axis. While this editing is very easy, it is very difficult to achieve a consistent tempo flow which could be considered to be "musical". Because what would be the rule for it? What does it mean "allegro"? 120bpm? or maybe 125 bpm? This is currently all a matter of try and error... I put tempi in the tempo map, then listen to the result until it sounds pleasing enough. There are, however, more problems: leaving the tempo at a constant pace and having quantised music notes makes the rendition sound like a machine gun. Would be ok for Techno music, but is unsuitable for anything else. There are two ways of avoiding this: shifting the notes out of their strict quantisation, or drawing additional tempo values into the tempo map. Neither of these methods allows a natural intuitive musical play, because the process of doing this requires offline try-and-error. There is one "middle" ground which allows a direct entering of such tempo subtleties: during the recording process, play live. This allows a direct tempo interpretation as is done in a real live play of the orchestra resp. the instruments. I did try this method in the 2nd movement, but then I encountered errors when trying to homogenise the overall recording: because the notes were due to tempo fluctiations not exactly on the metronomic beat, it was very difficult to match properly the various tracks or deliberately place consistent tempo chances. Therefore in this 4th movement I have resorted to the method of drawing tempo values into the tempo graphics. This leaves the music notes within each bar quantised to their metric position, but then allows me to introduce tempo changes within bars, e.g. subtle accelerations and slight ritardandi at the end note wherever appropriate. The problem with this: I did it on one day, then a few days later listened, and found all those tempo changes inappropriate. So I had to go again through everything and change the tempo at all the locations where I thought it was not right. It seems that the judgement of how the tempo should be is quite subjective even by the same person, depending on the mood, on the time of the day, etc. So there is no absolute correct tempo map. But it would be nice if there would be some more overarching way of changing those tempo subtleties instead of having to change more than 700 bars of music individually each time the tempo does not seem to be right...
It would be great if there was a tool with which one could add/modify the tempo directly as the music is playing, and then record the result right away. A slider might work, but it does not truly reflect the concept of "energy" in the musical motion... I am working on a software that would enable a different concept, but that software is work in progress only.
In any case, this is my "interpretation" of how I thought this movement should sound on 3.April 2011. Maybe in a few days I would post another version...
After I had put that intermittent version online, I addressed once again the tempo. Editing the tempo through the currently available methods in sequencer software is quite simple: I can simply edit tempo values at any instance, or can draw tempo curves over the time axis. While this editing is very easy, it is very difficult to achieve a consistent tempo flow which could be considered to be "musical". Because what would be the rule for it? What does it mean "allegro"? 120bpm? or maybe 125 bpm? This is currently all a matter of try and error... I put tempi in the tempo map, then listen to the result until it sounds pleasing enough. There are, however, more problems: leaving the tempo at a constant pace and having quantised music notes makes the rendition sound like a machine gun. Would be ok for Techno music, but is unsuitable for anything else. There are two ways of avoiding this: shifting the notes out of their strict quantisation, or drawing additional tempo values into the tempo map. Neither of these methods allows a natural intuitive musical play, because the process of doing this requires offline try-and-error. There is one "middle" ground which allows a direct entering of such tempo subtleties: during the recording process, play live. This allows a direct tempo interpretation as is done in a real live play of the orchestra resp. the instruments. I did try this method in the 2nd movement, but then I encountered errors when trying to homogenise the overall recording: because the notes were due to tempo fluctiations not exactly on the metronomic beat, it was very difficult to match properly the various tracks or deliberately place consistent tempo chances. Therefore in this 4th movement I have resorted to the method of drawing tempo values into the tempo graphics. This leaves the music notes within each bar quantised to their metric position, but then allows me to introduce tempo changes within bars, e.g. subtle accelerations and slight ritardandi at the end note wherever appropriate. The problem with this: I did it on one day, then a few days later listened, and found all those tempo changes inappropriate. So I had to go again through everything and change the tempo at all the locations where I thought it was not right. It seems that the judgement of how the tempo should be is quite subjective even by the same person, depending on the mood, on the time of the day, etc. So there is no absolute correct tempo map. But it would be nice if there would be some more overarching way of changing those tempo subtleties instead of having to change more than 700 bars of music individually each time the tempo does not seem to be right...
It would be great if there was a tool with which one could add/modify the tempo directly as the music is playing, and then record the result right away. A slider might work, but it does not truly reflect the concept of "energy" in the musical motion... I am working on a software that would enable a different concept, but that software is work in progress only.
In any case, this is my "interpretation" of how I thought this movement should sound on 3.April 2011. Maybe in a few days I would post another version...
Wednesday, 8 September 2010
New version of Mahler 1, 4th movement
Yesterday evening I decided to give my recording of Mahler's 4th movement an overhaul. The initial recording from 6.July had been done in quite a rush, trying to finish it before the anniversary on 7.July. Hence there were a few details that were not perfect. Especially the tempo had its problems: in my opinion now it was too slow.
So I revised the tempo and accelerated in many parts. There could still be some more acceleration, to keep the tempo more consistent in pace.
I also reduced the violin attack in a quiet passage between bars 166 and 220. Now the melody sounds much smoother.
Overall I think that I improved this recording. It is accessible on its web page in two versions: as MP3 and WMA. It also is automatically inserted into the streaming of the complete symphony which is accessible through this page.
Some pages on this web site do not load very quickly, due to the scripts for creating the links to the music files. I will have to switch this eventually to a database-oriented version.
So I revised the tempo and accelerated in many parts. There could still be some more acceleration, to keep the tempo more consistent in pace.
I also reduced the violin attack in a quiet passage between bars 166 and 220. Now the melody sounds much smoother.
Overall I think that I improved this recording. It is accessible on its web page in two versions: as MP3 and WMA. It also is automatically inserted into the streaming of the complete symphony which is accessible through this page.
Some pages on this web site do not load very quickly, due to the scripts for creating the links to the music files. I will have to switch this eventually to a database-oriented version.
Subscribe to:
Posts (Atom)