Initial contact came from the conductor Thomas Sanderling, a Shostakovich specialist and a young colleague of the composer. He explained that he had the go-ahead from Irina Shostakovich, the composer’s widow, to record the Suite on Verses of Michelangelo Buonarroti with the original Italian texts, and also the Romances in a supposedly ‘newly discovered orchestration’ and in English. This was an exciting prospect; it seemed a wonderful chance to sing some not-often-heard voice-and-orchestra repertoire incorporating my own education involving medieval Italian. Since I had recently worked on and recorded Liszt’s Petrarch Sonnets for piano and voice, this would be a further foray into this luxury era of poetry. Maestro Sanderling sent a copy of the Suite set in an organ transcription, with the Italian text set. As is my habit, I scanned the pieces, found them to be well within my range and discovered moments of great beauty. The Italian seemed to be occasionally out of line with the actual music, but I assumed that they had been set with care.

I knew that the recording would take place over a few days including two live performances of the Romances, and although a substantial programme, there seemed to be no reason to think the works would involve anything other than normal preparation and 'singing them in'. Thomas Sanderling agreed to meet me about six weeks before the concerts and sessions. In preparation, I worked on the Michelangelo sonnets with my collaborator Stephen Higgins and we began to sense that the setting of the Italian text was occasionally arbitrary, and sometimes very unhelpful in terms of stress and musical impact. At our working session with Sanderling, he was tolerant of the suggestions we made, but because there were a number of uncertainties, it was clear that a definitive arrangement was going to be needed. That was clearly my job; unexpected, but a worthy challenge.

Since the music of Shostakovich was the starting point, written in response to the Russian version by Thomas Efros, any suggestion of compromise could only be contemplated with the precise texts of Michelangelo. I ended up with about half a dozen versions! Without overblowing the problem, I would need to make some critical decisions urgently about the basis of the original texts, using the supplied music as a guide, the poems printed in the Shostakovich vocal score, and a number of other reliable sources.

I settled on an excellent combination of source and translation, by James M Saslow, in a 1990 Yale University edition of Michelangelo’s poetry. The translation was particularly helpful in conveying the directness and passion of the Florentine. His commentary was remarkably supportive, by declaring his own frustration in finding the true originals which, over centuries of adaptation and ‘improvement’, have been occasionally rewritten, hence the discrepancies between ‘original’ texts. He also admits that presenting the original Italian was already fraught with difficulty due to the evolution of grammar and language since Michelangelo's time. Even Michelangelo’s nephew in the ‘first’ editions had rewritten sections to heighten the poetic and lyrical nature, as homage to (or perhaps out of embarrassment for?) his uncle.

How did this ‘original’ text begin to fit with the music of Shostakovich? He was obviously writing for his countrymen in the vernacular, and the translation by Efros would serve very well, even if it did not sometimes allow the hard edge of the original script to come forth, due to that translator’s use of rhyme and his syllabic restrictions. Of course, there is a meter and pulse within the convention of the sonnet, and the Russian version respected this. Russian is a very vertical style of language, with definite pulse and stress within phrases, which musically can be heightened, or offset, to evoke meaning beyond just the words themselves. That is what vocal music is all about! Here would be the essential challenge for me: How could I take Russian music set to Russian text, and offer a response in retrospect to the strength and precise framework of the Florentine text?

I sympathise very much with opera translators who, in past years, were the lynchpin between artists and audiences for the performance of material which allowed a more complete experience. In efforts to make da Ponte understood, occasional rhymes, unashamedly crass, seemed to waft across the footlights. Don Giovanni’s Serenade was presented to me once as: ‘Your breath is like honey flowers, … Oh let us spend all hours…’ However, suppose a Mozart opera was only available in a viable translation as a source, performed with success when suddenly we ‘discovered’ the original da Ponte text?

It is an exciting mental adventure to have imaginary conversations with the composer, and ask if he felt that this or that pitch, rhythm, pause, accent, harmony, dynamic and in the end, the meaning of the music could possibly suit the new structure of text. In many cases, there would be firm rewriting. That is certainly not the task of the singer who is to perform the work in less than three weeks… and so the idea of compromise must take hold and become part of the creative process.

I am in the business of communication, and through the years, it has become almost second nature to prepare and nurture an emotional impact of a phrase. As any singer knows, it doesn’t matter what experience one has with languages which are not one’s own; it is essential that a native speaker with a cultural affinity to the text has an opportunity to be involved. My great fortune was to enlist the services of Glyndebourne language coach, Barbara Diana, also a scholar in medieval Italian. Her assistance in comparing texts and verb tenses was invaluable. Luckily we had a few sessions to grind out the text discrepancies, and also to make final adjustments to the ‘performing’ edition I would use. Of course, it was then important to copy the entire text into the full score, adjusted and legible, so that conductor, producer and singer could all be as one during the recording sessions. Job done, or so I thought.

The recording

Arriving in golden autumn Helsinki with the copied text and penciled scores was a tremendous feeling. My last visit was for a recital at the Opera House, where I had performed in Kaija Saariaho’s L’amour de loin nearly 10 years previously. Helsinki reminds me very much of Canada, not just in its autumn display, but in its geology and climate – almost like coming home. For now, though, Shostakovich was on my mind. I was proud of the attention to detail in the language arrangement we had made and I knew that the music felt good in the voice. I knew, too, that Sanderling would grasp the Italian setting with his acute ear, and we would make a ‘new’ piece of music from this. In spite of the language feeling fragile due to its recent birth, it looked quite secure on the page. However, as soon as I heard the orchestra at the first rehearsal and began to feel the strength of the music, it became clear that further amendments would be necessary here and there. Thankfully, and helpfully, the Romances had already been adapted for the original English and Scottish, and they did not pose any of the same challenges linguistically, except to decide how much of a Scottish accent to use in the Burns settings!

Our concert-day dress rehearsal of the Romances was being recorded but since a recording day following the concerts would allow retakes, it was mostly a run-through for sound balance. The first concert went very well, and there was enough to know that a recording would want only a few improvements. In any case, the microphones for the concert seemed to be in the same position as the dress rehearsal, completely in front of me, blocking any relationship to the audience. I insisted that for the second concert they be adjusted so the audience would get more of a 'performance' rendition. The recording team seemed surprised; I discovered that there was no time for retakes the following day! As a compromise, they agreed that the one Romance (Shakespeare Sonnet) for which they required a second complete take, we could perform as an ‘encore’, which I announced at the concert as ‘a technical issue’ and the microphones were duly reset. The audience seemed to accept it graciously. The Romances were ‘in the can’.

Between the rehearsals, recording and performances of the Romances, I was busy verifying the Italian texts as stresses and rhythmic choices began to need adjustment. It was non-stop, since all the scores needed to be marked correctly and recopied and I needed to readjust my own performance energy with each small variation. Tipp-Ex and erasures became part of my score and notelets to myself began to fill my pockets.

The day of the Michelangelo recording dawned a beautiful October morning in Helsinki. The invigorating walk to the wonderful new Musiikkitalo was framed by golden trees and a carpet of diverse yellows and bronzed leaves. It was a busy day, with a video interview for promotion and a photo shoot for the cover of the CD, before the session. Also we needed to confirm that the internet connection to the UK was operational.

As well as aiding my initial preparation, Barbara Diana became, I believe, the first language coach ever to supervise a recording from 1200 miles away in Sussex, England, over Skype. Since this was a completely new rendition of this music, it was already an uncertain enterprise, and so to have native ears continuously assessing the phrasing and meaning during the recording was critical. The IT gurus of Ondine Records and the Helsinki Philharmonic arranged it so that the feed could be direct from singer’s microphone to Skype connection, and with private headphones so as not to interfere with the orchestra microphones and to allow an interface between producer and singer. Whatever they did, it worked like a charm. Occasionally, it seemed the entire orchestra and conductor were waiting for me in silence as I digested language commentary, but I was extremely confident that the sensitive Italianate assessment would bring great rewards when trying to perfect the accents and flavour of the language.

We tackled the larger movements first. It was a strategy that the conductor and producer knew would keep the orchestra fresh and alert, and the intensity of Shostakovich needed attention at all times. Unusually, particularly for a purely vocal work, I had not listened to any playbacks in the session due to time constraints, but I felt that the producer and sound engineers were absolutely on top of the issues. By the end of our session, we had handled the greater challenges of volume and vigour, but we still had the major part of the Suite to finish.

From a vocal point of view, beginning at 9.30 in the morning is not ideal, particularly after all the stresses of editing, performances and initial recording. However, I was ready to go at 9.29 the next morning, when the more lyrical movements would be committed to track. I then realised, of course, that the session for the language coach began at 7.30am UK time, and I had little to complain about! We began, and I was relieved that the voice was in good enough shape to handle the subtleties we had hoped for in the lyrical movements and had enough stamina to sustain through to the end of the session, which we finished a little earlier than expected, but happily.

Smiles and warm appreciation all round, then, and a major work in its exciting new form committed to posterity. As a particular bonus, I was then able to join Einojuhani Rautavaara and family and friends, as they celebrated his 85th birthday, and ironically, on the birthday itself of Kaija Saariaho.

Gerald Finley sings the world premiere of the orchestral version of Rautavaara’s Rubaíyat with the Helsinki Philharmonic and John Storgårds in Helsinki on March 31, 2015. He will sing the world premiere of Kaija Saariaho’s True Fire for baritone and orchestra with the Los Angeles Philharmonic Orchestra and Gustavo Dudamel on May 14-17, 2015, at Disney Hall, Los Angeles

Read the Gramophone review of Gerald Finley's new Shostakovich recording in the July issue (on sale from June 20) or in the Reviews Database here. To find out more about subscribing to Gramophone visit our Subscriptions page