The Virtual Jami Blog Progress of creating a virtual performance of Sorabji's Jami Symphony

This chunk of work has taken longer because it is longer and because I am working on typesetting Havergal Brian’s 3rd Symphony for the HB society for a planned future recording. I have now completed 69 minutes and may be close to finishing the 1st movement around the turn of the year. This video consists of two extracts being pages 164-173 and 181-200 of the MS.

As before there is a continuous description of the music included in the video but just to highlite the first section is a glorious polyphonic all singing section led by the chorus and strings and the second extract is dominated by woodwinds and brass making lots of noise. All very exciting.

Whilst I’m working on the next chunk, I thought I would document my current working process. I have three versions of the score; a PDF of the handwritten manuscript; a “clean” typeset version and a “performing” typeset version from which the virtual performance is generated. I work on chunks of around 10-15 minutes’ worth of music at a time very much dependent on finding natural gaps/pauses in the music. This next chunk is a long one at about 17 minutes. It contains a very lengthy, multi-divisi, choral and orchestral polyphonic singing section with the chorus in 12-16 parts and the strings in 20 parts. It concludes though with a fantastical virtuosic section for the orchestra with abundant rapid semiquavers, lots of untuned percussion and stentorian brass.

The first job is to add all the necessary bars into the performing score. As we’re talking hundreds of bars and the time signature changes frequently this is no easy task. And I can’t just cut and paste from the clean score into the performing score as the layout of each score is very different. The clean score, designed for a conductor, has upward of 60+ staves but the performing score has more than double that as I have to have separate staves for strings in 10 parts, strings in 20 parts, violins in 16 parts, chorus in up to 32 parts, staves for individual wind & brass as well as staves for tutti wind/brass sections and muted brass.  For instance, the clean score has a stave for trumpets 1&2 with the part occasionally dividing into two parts on the one stave and occasionally requiring mutes. In the performing score I have to have two virtual instruments for the two trumpets and a separate virtual instrument for muted trumpets and each virtual instrument needs its own stave.

Having created the correct number of bars I cut and paste each individual stave from the clean score into the performing score and with 60+ staves that’s a lengthy and tedious job but from then on I concentrate on the performing score. I go through every instrumental part separating any divisi passages (where there is more than one line of music on a stave an issue mainly for the chorus and strings) so that each line of music is on its own stave. I also check the range of individual parts (an issue mostly in the woodwind) moving bits too low for the flute, oboe or clarinet for instance into parts for alto flute, hecklephone or bass clarinet. I have to check ranges of tuned percussion as the way Sorabji expected octave displacement in Glockenspiel, celesta and xylophone doesn’t work the same in the virtual instruments which mostly have to be transposed down an octave to generate the correct pitch. I have to virtually rewrite the piano and harp parts to ensure they generate the right sounds from the virtual instruments and for the same reason rewrite the untuned percussion using special percussion staves designed to generate the correct articulations from the virtual instrument. Only then can I actually start on creating the performance.

Still in lockdown which is allowing me to be musically productive. As well as continuing with the Jami I have been working on the typesetting of Havegal Brian’s 3rd Symphony. That would be considered something of a symphonic beast if it wasn’t compared to the Jami and still acts as light relief compared to Sorabji.

I’ve now completed 54 minutes of the 1st movement and heres a couple of extracts. There’s quite a bit of solo violin in this chunk some of which isn’t explicitly marked solo some of which is. It made me think that there is no explicit solo passage for cello or viola in the whole symphony. The organ appears for two bars in this video and gardly appears at all in this movement whilst it appears quite a lot in the third.

The video inlcudes a description of the music.

 So here we all are in coronavirus lockdown. I’m lucky as I have a garden, the weather is good and I have Sorabji to while away the many hours. This has enabled me to finish the next chunk of about 14 minutes which takes me to 44 minutes and nearly the half way mark. I’ve chosen three extracts which are all described in the video. It’s probably just my fancy but there’s a premonitional nod to Michael Nyman and a look back (by couple of decades) to Richard Strauss and the stormy bits of the Alpine Symphony and a passage where I’ve employed metric modulation to determine the tempo, a technique developed by my illustrious namesake Elliott Carter. I could complain that Sorabji didn’t bother with minor details like tempo and dynamic markings but then that wouldn’t allow me to exercise my creativity. Enjoy.

I’ve now completed the 2nd chunk of the 1st movement which takes me to 30 minutes in and I can offer you two contrasting excerpts.

The first half of this video opens onto a brief passage (MS pages 39-44) for 10 part strings alone which leads to one of those magical, whole orchestra, bi-tonal chords (quiet on this occasion) where the lower orchestra is ‘C’ major (+ harp 2 ascending gliss) and the upper orchestra is ‘C#’ Major (+ harp 1 descending gliss). This is followed by the most extended acapella choir passage in the whole symphony at about two minutes. An extended passage of beautiful 8-part counterpoint is followed by 16-voice, 4-part block harmony counterpoint which resolves onto an ‘E’ minor to ‘C#’ minor cadence. Unison trombones & tubas lead to some deep and ominous rumblings in the contrabassoon and double basses.

The second half opens on an unusual passage scored just for wind and brass (MS pages 53-66). Brass soon drop out replaced by strings as the wind play an extraordinarily virtuosic passage which looks (and sounds?) a bit like the virtuoso wind writing at the beginning of the Rite of Spring. Was Sorabji ‘cocking a snook’ at Stravinsky who he vehemently derided? Probably just my fancy. This is soon followed by an extended passage where the strings are divided to their maximum 28 parts, 16 violins and 4 each viola, cello & Doublebass. An extraordinary mosaic of every playing technique known to Sorabji; rapid runs; tremolos; glissandos; pizzicato; harmonics; trills; rapid arpeggios; spiccato and melody. Such passages are incredibly difficult to render and balance. I have always emphasised the melodic and there is melodic contribution from wind and brass. At the close of this passage chorus and wind rejoin and the whole orchestra quite rapidly moves to a huge climax on an ‘F#’ minor chord through which flutes, oboes and 2^nd violins (with some necessary help from me) wail a crying lament which soon collapses in on itself.

I’ve had a fair amount of spare time to spend on the ‘Jami’ recently and have completed the first large chunk of the 1st movement which takes us to around 16 minutes. This video consists of two excerpts, the first glorious polyphony moment and the first major climax.

MS pages 15-20

Starting at about 7 minutes into the symphony is the first emotional climax when the whole orchestra and chorus sing glorious, tonal, polyphony. It opens with a descending trombone line and soon leads, via high violin harmonics to a crescendo of descending triads in the lower strings with a rising gong, then rising crotchets in upper strings and wind, a trumpet fanfare, and a bass led cadence into ‘F’ major led tonality.

MS pages 29-36

The second half of the video. starting at around 13’20” into the symphony is the first major climax, and it’s a doozy! It is preceded by quiet descending and ascending parallel tonal chords in the strings before kicking off a massively complex section for the whole orchestra including percussion. Too complex to describe in detail there are more notes in these 7 pages than in 7 Haydn symphonies. Towards the end there are huge descending parallel chords, a regular feature of the symphony. The massive final chord is rooted on ‘F’ and is largely ‘F’ minor with most of the other scale notes added.

So here is the wonderful opening of the 1st movement up to the first reprise of the main theme, just over 5 minutes. Interesting things compared to the third movement? Much less use of the organ; no Tam-Tam or tenor drum at all. But this opening contains many ideas clearly carried forward.

I had thought I would take a break from Jami after the 3rd Movement. I have lots of other smaller projects gathering dust on the shelves. But, relative to the small but concentrated Sorabjiverse, there has been a very positive reaction to the completed 3rd movement and I am much encouraged and buoyed up so couldn’t resist getting straight on with the 1st Movement.

Whilst most people were recently listening to the 3rd, I listened to the 1st for possibly the first time for many years, maybe even a decade. I did the 2nd movement first in 2012 because it was short and I was learning my craft. At that point I realised my skills and my hardware may not be up to the task of the whole symphony. The 1st movement was very much my favourite and I didn’t want to tackle that until I knew me and my hardware could cope. My thinking was, if I could do the 3rd movement, the longest and most difficult, then the 1st and 4th would fall into place.

After 18 months I realised me and my hardware were not up to the task, so I stopped. 5.5 years later I could afford to build a beast of a computer and the software had improved immeasurably. I had also learnt much from listening to performances of Sorabji’s mightiest works from both Jonathan Powell and Kevin Bowyer and developed ideas about interpretation. The completed 3rd movement, I think, justifies continuing with the project and so I have got straight back on the wagon and plowed into the 1st Movement. I shall update as things progress.

Here it is, the complete 3rd movement

Although I completed the performance in November 2019, in an attempt to encourage greater engagement, I decided to make a commentary which runs for the whole 2+ hours. This turned out to be more time consuming than I had envisaged.

The commentary consists of my factual/prosaic description of the music as it progresses intermingled with text about, and quotes from, Sorabji as well as a technical description of how the performance was created. Permission was obtained from professor Roberge to use material from Opus Sorabjianum and from Alistair Hinton to use text from Mi Contra Fa and Around Music.

I have always envisaged presenting the complete performance of the symphony alongside a film of some kind. Perhaps a collage of text/film/pictures associated with Sorabji and this symphony.

Just using text is not ideal, but a start, and I quickly realised I needed to be far better organised about what text and how placed and this aspect will need re-visiting when the whole symphony has been done.

One moment of serendipity has encouraged me though. The placing of Sorabji’s description of the music in Szymanosvki’s 3^rd Symphony at one of the most ravishing moments in the Jami Symphony is goosebumpingly uncanny. 1:17:04 in the video.

I would be really interested in any comments about any aspect of this performance, good or constructive. What wonderful music. On to the 1^st Movement next.

I have stitched together 5 moments from the opening 21 minutes one fading out and in to the next.

The opening is loud and complex with every section setting out its stall. It foreshadows, by some 2 hours, the end although that is very, very loud and very, very complex. This opening statement ends with a strings full stop followed by a brief acapella choir section.

We fade back in to the culmination of a lengthy melody in the violas and then violins with full orchestra rising to a solid harmonic climax on G major with a typical descending, close harmony, parallel major triad movement in the bass instruments. This leads to a sudden halt followed by a not uncommon quite pulsing rhythm in the strings over which the 1^st violins play yet another variation of the main theme of the whole symphony.

The next excerpt starts with a beautiful gesture from the Sopranos and Altos echoing each other in contrary motion (who’d have thought!) and is followed by a long melody in the tenors which also peaks on a lush major chord, this time E major (with additives) and an extra lush violin glissando.

More singing in the next excerpt but this time in 32 parts, a huge joyful section for the chorus. After a climax the 32-part chorus sing acapella quietly with a hint of main symphony melody in the sopranos.

The final part of this selection leads to a big climax but is preceded by a most unusual section for just woodwind, brass and percussion. The brass marking the beat whilst the woodwind scurry around like a swarm of insects, the chorus and strings can’t help but finish it off though.

It’s interesting to note that neither the organ nor the tam-tam play in this opening section, neither joining the foray until around 25 minutes in.

Chronologically, this is the 2nd of 8 excerpts. The whole 3rd movement is likely to come in at over 2 hours and this excerpt is 0hr 21m to 0hr 31m.

Its notable for several significant trumpet solos including the quotation from a symphony by Chausson. There are two lengthy sections for full chorus, orchestra and organ. After the second, at the end of this except, is one of my favourite moments when unison cellos and basses lead into a long, quiet, sustained B major chord spread across multi divisi strings which is at first disturbed by movement in the lower strings but then settles back and over which 4 sopranos distantly intone a variant of the main motive of the whole symphony. Stunning.

That leaves the final chunk of the 3rd movement to do which is the 1st 20 minutes. That’ll take a while what with the start of the new music season and several band contests and orchestral gigs coming up but I am hopeful I could have the whole of the 3rd movement finished by Christmas.

So I’ve completed another retrospective chunk of 16 minutes plus and can offer you a most generous 11 minute plus portion. My thinking at the moment is not to post everything and not to reveal the whole movement at least until it’s finished and then maybe not then as I’m contemplating some sort of “World Premier” performance of the whole work, maybe on line via YouTube, maybe in a performance venue, maybe both.

There’s a nice bookend to this extract which begins and ends with a bass oboe solo. At 23″ the strings all expand to 4 parts each, 20 overall for quite a stretch. Sorabji employs a variety of playing techniques in this section. At the start the 1st violins are playing very high rapid chromatic lines in block harmony (major 7th chords), the 2nd violins play glissando figures in block major/minor chords, the violas play melodically in block 7th chords, the cellos provide steady pizzicato in major/minor block chords, the basses have sustained notes in this case major and minor 2nds doubled at the octave. Each group will swap what they’re doing and there are also tremolo passages, artificial harmonics, scalic rushing, arpeggio figures and trills all constantly passed around. This is a fairly typical passage. Over the top of this string melee there are snatches of melody in the wind and chorus as well as accompanying figures. At 36′ there is another common gesture when a steady staccato semiquaver pointing starts in the clarinets then moves to the other wind then to the brass.

At 2’08” the strings reduce in texture to an ascending running scale from basses to violins punctuated by short tremolo jabs whilst the wind descends from on high with a rapid staccato tonguing passage. Most of the orchestra then joins building to a semi climax with stamping bass and timpani which dissolves suddenly into an entirely different mood and tempo at 3’09”.

Talking of tempi, this next passage is interesting as it has several changes, gearing up gradually from pulse = 50 BPM to pulse = 180 BPM. Now, you must remember, there are absolutely zero tempo indications anywhere in the score. Everything is at the discretion of the performer and in this case entirely made up by me. The initial 50 BPM section is another typical Mediterranean style passage, long languid melodies in middle strings and wind, steady rhythmic pulse in lower strings and percussion, skittering around in upper strings and wind. This gradually increases in texture until we move into 2nd gear which is a doubling of the speed at 3’57” and a 3/4 pulse marked out by the cymbals and tam-tam. There’s some heavy lower brass interjections and a lot more movement through the orchestra. This doesn’t last long before we move into 3rd gear at 4’08” and a tempo of pulse = 116 this time in a 4/4 feel marked by the gong repeating a leap of a 5th. Again the action increases with more activity in all sections. At 4’20” we arrive at 4th Gear with the pulse at 180 BPM with huge rapid running passages in wind and strings and plenty of energy in the brass and timps and noise from the choir. A pretty thrilling passage.

This passage peters out into calmer quieter fair at 5’20” where we revert to ideas similar to the passage at the beginning of this extract with the strings again divided in 20 parts. I’ll leave you to your own imagination for the rest of the extract but listen out for a nice heart beat thrubbing in the bass drum and lower strings at 7’04” accompanying melody in the chorus and gong. There’s a fuller “Glory Glory” section at 8’48” (without organ).

I have completed my first effort to retrofit the performance of the first 72 minutes of the 3rd movement made in 2013 onto the 2019 performance of the remainder of the movement. It’s been tricky making sure various settings, particularly volume settings match and the performances sound seemless. Much work was necessary for the keyboards, tuned and untuned percussion. It’s taken a couple of weeks and hopefully it’s worked OK.

This excerpt is the last 6 minutes leading up to the excerpt posted on 22 June 2019 commencing MS Page 591 so you play one after the other and judge for yourself whether the performances sound seemless.

The notable thing about this section is it contains the lenghty passage when the 1st and 2nd violins divide into eight parts each. This is the 3rd time I have tried to render this passage and is hopefully the best. However the texture around the violins is itself extremely dense so picking out detail is difficult.

The climax of the third movement.

I’ve said before and I’ll say again this must be one of the most climactic moments in all western orchestral music. It has been extremely challenging although my hardware and software have coped remarkably well. My musical faculties have been sorely stretched to cope with the very end. When you have 4 timpani, bass drum, side drum, field drum, 2 triangles, 2 cymbals and tam-tam (not to mention the castanets and tambourine trying desperately to join in) all rolling at treble forte, the organ at full blast every singer and instrumentalist singing/blowing with all their might, nigh on 500+ musicians busting their guts, you gotta expect a lot of noise. The problem has been to try and make not just noise and if you listen closely you should be able to pick out the voices and maybe the brass and certainly the squealing Eb clarinet as well as the bass instruments vomiting up some sort of C# major/minor arpeggio type thing. The final chord, which is preceded by a long pedal G, is rooted on a C#/G# fifth in the bass with the male chorus, lower wind and brass on a Bb minor chord, the female chorus and upper orchestra on an A minor chord and the upper woodwind wailing out a tremolo Bb to C.

The lead up to the final cataclysm is full of Sorabji’s usual extraordinary invention. There’s a lot of important moments for the tuned gongs and I’ve been imagining them as the voice of Apollo signalling the end is nigh.

Decent audio equipment will improve the listening experience. Turn up the volume and let the room shake.

This extract opens with a characteristic Sorabjian mediterranean dance with castanets, tambourines and harps. The static accompaniment in the strings (quasi strummed guitar?) building from 6-part bassi and celli adding each section until we have 17 parts and extremely high violin harmonics. Over this static accompaniment the violins start a sultry/languid tune and the wind and chorus take over. This reaches a peak (2’02”) which dissolves into quiet accapella singing and gong strokes which soon leads to another full bodied multi singing polyphonic section without the organ. During this passage a meandering quaver bass line starts (3’07”) which will dominate the next two sections. As the choir and upper strings fade away the persistent meandering quaver line emerges in the basses alone (3’48”) and then wanders it way up through the cellos, violas and 2nd violins. When it reaches the 1st violins it is then joined successively at octave intervals back down through the strings plus woodwind until we have four unison octaves which continue to descend and build into a huge polyphonic slightly disturbing Glory Glory section (5’36”). The organ re-enters, the meandering quavers continue in parallel harmony in the upper woodwind and violins, there are massive punctuating chords in the brass, the chorus sing the glory and there is a quasi-chorale like accompaniment in the lower wind and strings. What a marvellous section. The volume is pared down after a while and the chorale breaks up into more complex polyphony. Stamping percussion (7’05”) signal the climax of this section which abruptly breaks into rapid semiquavers in wind and violins with rhythmic marking of the beat in brass and percussion. The semiquavers get faster and faster building up through the whole orchestra before suddenly breaking off into….. (watch this space).

How one man could imagine all this music is unfathomable.

This excerpt features an extended choral passage where they divide from their usual 8 parts into 16 parts and then into 32 parts.

A quiet mysterious opening reaches a first climax at 1’32” at which point the lower strings start a rocking figure in parallel fourths which emerges from under the receding choir. Additional layers join up thorugh the strings until we have 10 stacked fourths and a nasty shout from percussion, piano & strings at 1’44”. A return to pleasenter stuff leads to the start of the choral passage at 2’12” at first in 8 parts and then 16 parts at 2’38”. I’ve rendered this passage, and much of this section, as quietly as possible. There’s a five part male chorus moment at 3’13” marked “[i]ferme bouche[/i]” . The VSL chorus samples don’t have ferme bouche (humming) samples so I’ve used “ooohs” instead of “aaahs” which I think works as a nice contrast. After another climax at 4’08” the 32 part chorus kicks in. As 32 staves of Sorabji’s manuscript paper were taken up with the choir the orchestral writing is very thin and for much of the time the chorus is acapella. The density of the writing is extreme and it’s hard to make sense of it. The best tactic (as is so often the case in this symphony) is to let the glorious sound wash over you. At 6’45” the chorus returns to 16 parts and the orchestra thickens out. A gradual build up leads to a climax at 7’40” and another return of the mighty organ and a signature passage which occurs several times throught the whole symphony, a “hallelujah” grand joyous noise for everyone.

This extract starts where the last one from 2013 ended. I hope the quality of the performance is better. It sounds better to me. One of the significant changes is to add a second microphone on the virtual stage. One is in the 7th row and the 2nd is where the conductor stands. I’ve also overhauled the settings in the virtual mixer which allows me to increase the presence of individual instruments and sections far more than I could before. Now I can have the choir dominate if I wish. The sound of the performance is from the perspective of a live performance and not a studio recording.

Of note is the return of the big chorus, organ and full orchestral ecstatic singing at 2’55” so characteristic of the symphony as a whole, and when this returns again at 4’24” it is proceeded by nothing less than a complete descending F major scale. This section peaks at 5’35” on a clear F major chord and is followed by a lovely rippling section in wind and strings. Some quiet clusters in the chorus, harmonics in the strings and pointilistic hocketing in the wind at 7’48” leads to a step up in the tempo at 9’15” and the final part of this extract, a vigorous section for rapid winds and loud brass with several aggressive fortepiano crescendo chords in the brass (really hard to pull off in Sibelius/VSL).

I could witter on for ever about how wonderful this music is. It is wonderful. It is also very different to the keyboard music. I believe it is much more accessible. I have read concern about Sorabji’s vertical harmony and melodic invention. Well this symphony has melody in abundance and clear and obvious vertical harmonic progression to satisfy the nay sayers. This extract has both to hear.

In case there are any computer freaks listening in I thought I’d post the specs of my new PC. The components which make it possible to cope with the huge task of creating this virtual performance are 64GBs of top quality RAM, the high end processor and the super fast Solid State Drives for the operating System and the Sample drives.

The software I use hasn’t changed other than the several updates over the years. Sibelius 8.2 and the Vienna Symphonic Library. I use the VSL Symphonic Cube sample collection. I have not moved on to their newer “Dimension” or “SYNCHRON-ized” sample sets, just too expensive. The main VSL software is the Vienna Ensemble Pro 6 (the virtual mixing console) and MIR Pro (the virtual concert hall where the virtual performers gather).

I’m back on it. Following xmas 2018 I managed to get together enough finance to build me a monster PC. Once assembled I went straight into testing it on the remainder of the 3rd movement from where I had left off nearly six years ago. It quickly became apparent I was good to go and the new machine could cope with whatever I wanted to throw at it without compromise. Needless to say I was too excited creating the performance to worry about updating this blog and in any event I had forgotten usernames and passwords. Obviously I’ve now resolved that and can now regularly update the legion of followers of this blog and Sorabji fans worldwide (insert emotican for tongue in cheek here).

I know there are lots of broken links to earlier posts and I will get round to trying to sort that in due course but first I will post all the exciting stuff I’ve been doing in the last few months. Regular visitors to the Sorabji Archive forum will spot much duplication of posts there.