Pieces sans Basse, by Jean Daniel Braun.
    This is a book of exercises, or etudes, which was engraved and published in Paris in 1740. The title page is here. So the book contains a sonata with basso continuo in four movements at the beginning, which I haven't transcribed here, and then a collection of further pieces. The fact that the book exists at all shows that the flute was very much in fashion in the middle of the 18th century. But the title indicates that the bassoonist can also use these pieces if he imagines that he is dealing with the bass clef, rather than the violin clef, and furthermore, the key should be adjusted so that instead of having one sharp (g major), one should imagine two flats (b-flat major). (Then of course, d major should also be transposed to f major.)
    The original printing is quite clear and done with much style, so many people would say that there is really no excuse for setting it using the computer. (The source file is here.) But still, it gives me something to do when at loose ends, and anyway, I find that when playing something like this, where the eye must rapidly find the right notes in this uneven, ancient printing, things become even more difficult than they otherwise would be. Furthermore, if you have the notes in the computer, then it is easily possible to change the notation as you would like. I have retained the original pagination, since the pieces on the same, and on opposite pages, are often musically related to one another. I have also tried to reproduce exactly the original notation. Nevertheless, at a couple of places - the minuetto on page 5 and the fantasia on page 6 - Braun has just written a middle c, or c sharp, which of course is too low for the baroque flute. I suppose this was done for the convenience of the bassoon player. But rather than thinking about this each time I play it, I have put in the appropriate note an octave higher.
    Of course there is a huge amount of flute music from the baroque period. On the other hand, as far as exercises for the flute are concerned, I only have Frederick the Great's 100 exercises, Quantz's Solfeggi, and his Caprices. These were private books of exercises, perhaps mainly by Quantz, written by hand, which were not openly published back then in the 18th century. However they can now be bought in modern editions at any music shop. Frederick the Great's exercises are just short technical finger exercises. They are sometimes enjoyable to play, but really, there is no musical expression in them at all. The solfeggi contain lots of very technical exercises, and fragments of exercises. They seem to be the things which Quantz had worked out specially for Frederick, so that for an outsider - which is to say, for everyone today - they are not so much interesting as music, but rather as an insight into the lessons Frederick was taking. Finally the caprices are in a manuscript by an unknown copyist, but attributed to Quantz, which is in the Royal Library in Copenhagen. They are of a very uneven nature.  It seems to me that the numbers 1 - 3 and 5 - 23 must be from Quantz. But the numbers from 42 to the end are different, and much more interesting and pleasant as music. Of these, the pieces 44 - 46, 48 -50, and 52 are simply taken directly from Braun's book (but with some changes, perhaps for the worse). Interestingly enough, I think that these are the least interesting pieces of Braun. Taken as a whole, Braun's exercises give you a satisfying variety of moods, and pace. And as the title says, they certainly are good exercises for the embouchure and the fingers.


The Bicinien of Pietro Vinci.
    This is a collection of 29 duets by Pietro Vinci, published in 1586 in Venice. Just this last summer, on a trip to Austria, Rudolf Tutz was kind enough to make a renaissance flute at a=415 for me at very short notice. It has a beautifully soft tone, and so now we are in a phase of playing this renaissance music. In those days, music was published with a separate booklet for each player, containing only the music for one instrument each. For example these "bicinein" are written for two voices: Cantus and Tenor. Of course the music of those days had no bar lines and no beams. Sometimes this renaissance music is published in modern notation, putting bars in for 2/2 time and thus suggesting that the rhythm is given by the half notes, or even the quarter notes. This leads to a ridiculously slow tempo. In reality, the rhythm is given by the full notes, and I suppose one should play at about the tempo of the beating heart, say 70 beats per minute. (Of course the full note was called a "semibrevis" in those days, indicating that in still earlier music, the brevis was the "short" note, giving the tempo at double that speed again.) Therefore this music is a difficult, but very rewarding challenge to play. All of the pieces are quite different from one another. There are strange rhythms, almost avant-garde combinations of notes. It is a totally different musical world than that which we are familiar with. And it only makes sense when both voices are played together.
    The original notation poses a number of problems for the modern eye. For one thing, one is expected to skip back and forth fluently between different kinds of clefs. Then the rhythm often goes against the beat for a time, through different note values, confusing both players. In fact, for anyone who still believes that renaissance music is simple, I challenge you to play the 27th piece in this collection fluently without a great deal of preparation!
    In any case, our difficulties were such that I decided to try putting these pieces of Pietro Vinci into the computer in order to have a more easy score to read. At first this was just a matter of having modern notation. Yet for the purist, that distorts the final result so that you end up playing the music as if it were some kind of baroque monstrosity, or whatever. But then I discovered that Lilypond does offer the possibility of setting music in the mensural notation. Perhaps this is still too "easy" to read. Lilypond does separate the notes slightly into whole note groups so that one does not get lost so easily in syncopated rhythms. Also, having both voices together on a single page results in the player looking at the music of the partner, rather than listening to it. This was certainly not Vinci's intention! In any case it is not quite as bad as modern notation, and it does enable us to play the music in the first place.
    The source file is here. I have corrected two quite gross mistakes in the "La danzulina" piece (although it remains a very unusual piece of music). Also I have put in an accidental which wasn't in the original music in a couple of places. The musician is expected to accommodate this "musica ficta" spontaneously into the playing, especially to avoid the unpleasant tritone. But I have found that in one or two places this comes up so unexpectedly that I needed to have it in the written music. In five of the pieces, the cantus voice is placed in the violin clef. Of course in those days, that meant it was played an octave higher than the modern convention. This is a bit high for my tenor flute, so I have transposed them down either a fourth or a fifth, as has been indicated in the titles. One little "problem" with the Lilypond mensural notation is that when writing an accidental to indicate that a b should be played, rather than b flat, the natural symbol is set, rather than the sharp symbol which was usual in renaissance printing. Therefore, if you play the music in transposition, you will have to make the effort to remember to play a sharp when you see one of these natural accidentals. I could have told Lilypond to write #b, but that seems to be taking this imitation of the authentic somewhat too far!
    Of course, given that you have worked out how to play in transposition, then the question arises as to why I bother to transpose the five pieces here which are in the violin clef in the original printing. For one thing, it was fun to see how nicely Lilypond can do this. But also it is not quite so trivial to transpose down a fourth while playing music which is as complicated as this. (I am still trying to avoid the lower two e flats on the renaissance flute at all costs!)


Telemann's 12 Fantasias.
    There are various editions of these fantasias. For years I have had the Bärenreiter edition which was edited by Günter Haußwald. Every so often I have a phase of thinking of them as being very superior exercises for the flute, and I just play through them without thinking about repeats, just having fun. So my copy of these notes is beginning to look rather seedy, and it is falling apart.
    After concentrating on something else for our Christmas concert this year, I got back to the fantasias; but more and more, I found that playing from the Bärenreiter edition was becoming a frustrating experience. In particular, with all these accidentals, I find with increasing age that my mind wanders too much, and so when arriving at the end of a bar, I must ask myself if a given note was already declared to be an accidental more towards the beginning of the bar, or not? Back in the old days, this rule about an accidental holding for the rest of the bar, then forget about it in the next bar, was not applied. On the contrary, accidentals were marked accidentals if they were that, and not if they were not. How much simpler life was in those days! Also, of course, going back a bit further, people didn't even use bars in music at all, so this complicated modern rule would have been meaningless.
    Thus I decided to try and set the fantasias myself in large print Lilypond type, trying to give an exact reproduction of the original printing which Telemann himself organized back in 1728, or thereabout. For this, I needed a facsimile edition, and I see that Barthold Kuijken published one in 1987, originally by Musica Rara, but now by Breitkopf and Härtel. And for me, his recital of all 12 in one go in the historic town hall at Osnabruck was one of the most impressive concerts I have experienced.
    His edition also includes a transcription in modern notation. This is necessary, even though he always recommends that one play from facsimiles of the original editions where possible, since the only copy - which is in the Brussels Conservatory Library - is in poor condition and in places it is hardly readable. Certainly this modern transcription of Kuijken is much superior to that of Haußwald. Yet, of course, it retains the modern problem of accidentals in bars. Also it is not perfect. For example the sixth note in bar 20 of the first allegro in the 5th fantasia is clearly an "a", and not a "g". In bar 34 of fantasia 6, the 4th note is elongated, something which occurs many times in Telemann's printing. In such cases, he had corrected a mistake in the original engraving. Barthold Kuijken says that in this instance, the correction is from "f" to "g", but to my ear, it sounds better if we think the correction was in the other direction.
    Anyway, I have tried to reproduce the notation of Telemann's original printing as exactly as possible - including the methods of notating repeats and alternative endings which were used in those days. Also it seems to me that these big, crowded notes give more the feeling of the original. (Telemann crowded each fantasia onto a single page!) Undoubtedly I have also made mistakes, but the intentional changes from the original notation are in parenthesis. In particular in fantasia 8, bar 5 of the largo, Telemann's "g" for the 9th note seems just too strange to me, and so I would prefer to play a "g-sharp". Also the 2nd note in bar 13 of the allegro is a "g". I have left it there, but it seems to me to be very strange, and I would agree with Haußwald's changing it to an "f-sharp". On the other hand, given the obvious fact that Telemann himself carefully corrected the original printing, it seems that Haußwald might have been rather too free in his (often very sensible) editorial corrections.
    In fantasia 10, bar 49, the fifth note in the facsimile seems to me to be quite clearly a "b". Yet in all printed editions, it is rendered as a "c-sharp". After thinking about it, and also looking at the fourth note in bar 50, which is placed a fraction higher on the stave, yet still rather low for a "c-sharp", I've decided to change it. Also in bar 51, the fifth note in the facsimile does not have a natural sign before it, yet playing it as an "a-sharp" really doesn't sound particularly good. In any case, the Lilypond file is here, so you can also change and correct things as you might see fit. Following Telemann, I have written the various movements of each fantasia all together, and so Lilypond has numbered the bars without re-starting at each new movement. (Of course Telemann did not bother to number his bars in the engraving!)
    As a final note, and to realize the importance of these fantasias, it is interesting to know that after Quantz published his famous Versuch, an earlier student of his, the Danish aristocrat, Joachim von Moldenit, published his own discourse, in which he claimed that he had found much better methods of playing the flute. Thus Quantz ideas had become outdated. Naturally, the great Johann Joachim Quantz did not agree to this, and so he issued a challenge to his upstart student, proposing a musical competition in which both would play Telemann's famous fantasias, the winner to be decided by an impartial jury. Unfortunately though, von Moldenit got cold feet and failed to appear!

    (A further note: I have just discovered that there was a mistake in bar 3 of the 6th fantasia, and this despite the fact that I have often played from these notes myself. Somehow, one anticipates how the melody should go, plays it, and the eye simply glides over a wrong note without noticing the difference. Of course, as always with Lilypond, if you compile a score with a new version, then you must fiddle with things so that it again works. The syntax is always changing! That is the unpleasant side of Lilypond. But given this, I then decided to reduce the size of the notes to 22 point. The music seems to me to be somewhat more readable that way. I'm using Lilypond version 2.8.8 here, despite the fact that the current "stable" version which comes with Ubuntu 7.10 is version 2.10.xx., or whatever. But in 2.10, they made changes to the page-breaking algorithm which results in a seemingly endless recursion in longer files (and that, despite the fact that this score has the page "breakbefore" instruction at every second page). This problem has been corrected in the "unstable" version 2.11.xx, but for now I am just staying with 2.8.8., which works nicely.)


Six Sonatas, by François Devienne (Op. 68 Livre 4).
    I don't know if there is a modern edition of these sonatas. The original edition was published in Paris in 1799. It is for flute with bass, and the bass line is very simple, no chords or bass figures. The printing is clear, and certainly adequate for the bass player. I suppose it should also be clear enough for the flute player. But in reality, my problems with the facsimile were such that I have decided to typeset it myself.
    This music of Devienne is so wonderfully melodic! It is a joy to play, and Devienne (who was the first professor of flute at the newly established Paris conservatory) saw no need for more than the single key on the baroque flute. Thus the music just flows naturally out of the single-keyed flute. I suppose one really needs a flute modeled on the patterns of the late 18th century: a Kirst, or Grenser. Still, it also sounds good on my Rottenburgh, although the occasional high f is a bit strained.
    My problem with the facsimile is that the turn (or gruppetto) symbol comes up often, but also the gruppetto figure is sometimes written out at other places. What is meant? After all, we have one way of playing it if it occurs between two notes, and another way of playing it if it is located above a single note. Often in the facsimile, the symbol floats tantalizingly halfway between the one possibility and the other. And then there is the question of how we should break up the rhythm in these figures? There are various possibilities. Thus, in playing from the facsimile, I find that after I have worked out one solution for a given part of the music, then coming back to it later, I wonder if that was the best, or perhaps I forget what had been decided upon the last time around. Such confusion! So that is the reason I've decided to do this typesetting. It was in order to decide in each case on a definite figure, and to always write it out. But rest assured, I have certainly not tried to fantasize anything new into the notes which isn't in the original, in the style of those editors of the 1950s! And I have not changed any of the figures which were specifically written out in the facsimile.
    I also have another collection of six sonatas by Devienne, published in modern notation by the Swiss publisher "Edition Kunzelmann". There is no foreword, explaining what those sonatas are, but in any case they are not Opus 68, Livre 4. It took me a while to get into playing them, owing to the fact that the spacing from the top line of the staff to the first ledger line is much larger than the spacing between the staff lines, and between the ledger lines further up. Since much of this music of Devienne takes place well above the staff, this additional extra little jump was a source of agitation not really welcomed in the music! I find it astonishing that such expensive sheet music contains such elementary mistakes. Oh well. At least I do have some sensible sheet music by Schott, which was beautifully typeset in 1964 (apparently by "Editio Musica", Budapest), giving Devienne's IV. Concerto in G major. And it is not as difficult to play as either these sonatas, or the ones in the Kunzelmann edition.
    Anyway, to return to this Opus 68, Livre 4, the facsimile has a number of mistakes. For example the number of notes in a bar is sometimes wrong. (Of course with computer typesetting this mistake never occurs, since if there is one thing the computer is good at, it is counting numbers!) Also the occasional accidental is not written out. And so forth. So I have felt free to correct such things in a way that sounds right to me. The staccatos are alternatively dotted or dashed. I can't see the logic of the difference in the facsimile, but still I have tried to maintain the original notation.
    As far as Devienne himself is concerned, the story is that he only lived to be 44 years old, dying in a sanatorium near Paris in 1803 after suffering a mental breakdown. Perhaps the strain of all those revolutionary ideas was too much to endure. Everything became decimalized: the week suddenly had 10 days, and so forth. Perhaps Devienne spent lots of emotional energy arguing against the introduction of the 10 tone scale in music! I thought that perhaps the music publishing people in those days also got distracted by their revolutionary ideas, thus leading to the mistakes which we see in the facsimile, and perhaps that also drove Devienne mad. On the other hand, having gotten started with typesetting this music myself, I see that it is much more difficult than anything that I have attempted before. So I have a great deal of sympathy with Devienne's publishers!
    I have been using the NoteEdit program, which is full of bugs, and it seizes up regularly in this music. Thus one can say that Devienne tests the limits not only of the one-keyed flute, but also the limits of contemporary software! The people who have taken over NoteEdit say that they are abandoning it, and instead are developing a new and better system, called Canorus. I'll wait till Ubuntu includes it in their official depositories. The alternative is Rosegarden. A few years ago I used to use the old Rosegarden, but the new version is too complicated for me. And I'm not interested in all this midi doodling which Rosegarden seems to have mainly become concerned with!