Kimiko Ishizaka and MuseScore Team Release Open Well-Tempered Clavier
libregraphicsworld.org
libregraphicsworld.org
Much of what they have there are 19th-century editions, which are now of course in the public domain. You might think that 150+ year-old editions aren't super-valuable, but many of them are from the Breitkopf & Härtel collected works editions. These editions are some of the most important works of 19th-century music scholarship, and were collected, edited, and subscribed to by important figures. Robert Schumann's works, for example, were all edited by Clara Schumann (an important composer in her own right), and Johannes Brahms is listed among the subscribers to the Bach complete works edition (among many others).
What IMSLP doesn't have a lot of is new, studio recordings, with the highest standard of playing and recording engineering. That's what the Open Well-Tempered Clavier provides!
Also, congratulations to the MuseScore team and to the pianist Kimiko Ishizaka for completing such a beautiful project!
http://www.gardnermuseum.org/music/listen/music_library?filt...
Unfortunately, most people aren't aware of this. It would be a shame to lose some of the nuances of the composer's original intent because of, essentially, propagation error.
I wonder how this edition fares? (I don't have my physical score to compare, sadly.)
Just one example stares at you from the Wohltemperiertes Klavier itself: physical evolution of the capabilities of musical instruments. This work itself was in part a demonstration of a new tuning theory. It would have been impossible to play in all these keys just a few years earlier.
So the question can also be framed as a kind of exegesis: Bach might have wanted the Wohltemperiertes Klavier to sound in an entirely different way given a different context.
And while propagation error is certainly a risk with lesser known composers, in the case of Bach, don't be too afraid of it. At all.
I find the choice of MuseScore and it's somewhat proprietary XML-based format an odd choice for "open" though. For example, I tried to open these scores using the MuseScore binary that comes with the latest ubuntu, and it was unable to open the files due to the fact that they're written using the pre-release 2.0 version of MuseScore.
Personally, I find LilyPond (http://lilypond.org/) a much better choice for these sorts of "open score" type projects. It's editable without a GUI, and produces best-of-class output. Think of MuseScore as "a drawing program for musical scores" and LilyPond as "a language for expressing scores in a human-readable, generic form"
The Mutopia Project (http://www.mutopiaproject.org/) has been hosting and re-engraving public domain scores for a very long time. All the source code including scores is on github via https://github.com/MutopiaProject/MutopiaProject Although MuseScore has a prettier UI, Mutopia and IMSLP have much larger and comprehensive collections.
Anyway, Kudos certainly to the recording and performers, they're great, and I look forward to score engravings that don't rely on the MuseScore software.
All I could find was http://musescore.org/en/handbook/file-format
Without a full specification (ideally with parser implementation, example source code for reading & writing, etc.) the XML is essentially proprietary, although still human readable.
This specific project was engraved using MuseScore 2.0, and their files cannot be opened using MuseScore 1.3.
For me, this is contradictory to the notion of "Open".
I'm not sure if we can call both MuseScore and Lilypond native format open... It's a matter of definition. MuseScore and Lilypond are both GNU GPL and so libre software. The fact that the project uses "Open" Well Tempered Clavier is indeed probably misleading but for most people they get it: The goods are released directly in the public domain and in formats that can be read by software with libre license, and than can be converted in many more formats for different usage.
Lilypond is an amazing piece of software. I still use it from time to time. It's great, no dicussion about it. But...
1/ Where is the specification, parser implementation and source code for reading writing lilypond format? MuseScore had once a Lilypond exporter and importer but nobody could maintain it any more because of the lack of such a spec. And this is perfectly understandable. Lilypond is powerful and need a powerful language and it's extremely hard to document in a formal way. As far as I know, only Lilypond can read Lilypond files. MuseScore format is not documented either, except in the code source. For both software, Also it makes little sense to document the file format since a lot of details in the score are not in the file format. For example, the position of the accidentals in a chord is not encoded in MuseScore or Lilypond files. However, they can provide export filters in documented formats MIDI and MusicXML.
2/ Ever tried to render a Lilypond 2.18 file with Lilypond 2.10? it will fail the same way that MuseScore 1.3 (probably the version in the repositories) will fail at reading MuseScore 2.0 files. And it's probably true for a lot software still in development. I guess that Blender or GIMP native file formats are not forward compatible either. Does that make this format not open? Maybe, so what? In any case it's perfectly understandable, format needs to be upgraded to support new features.
3/ The standard "documented" format for music notation interchange is MusicXML. It's proprietary but under a royalty free licence. There are ongoing discussions to make it a W3C standard. MuseScore import and export MusicXML files. Anyone can get the MusicXML file for the OpenWTC scores. This file can be imported in Lilypond. Unfortunately, Lilypond cannot export MusicXML though. It's one of the oldest feature request in the Lilypond bug tracker...
In the end, there are two more works in the public domain in a variety of formats that can be used for many different use cases. It's open the good word for it. I don't know. But I want more of work like this!
As for whether the MSCX format is documented or not, of course it is. MuseScore is fully open source. Every line of code responsible for reading and writing this format is freely available to you to examine, compile, etc. It doens't get more open than that.
The MuseScore XML files do not contain any link to a DTD, XSD or other XML DOCTYPE metadata, and thus, they are not self-documenting.
By comparison, the MusicXML format is open, because the following documentation is available: http://www.musicxml.com/for-developers/ Those zip files contain a complete and full specification of how to read & write MusicXML files, as I'm sure the MuseScore developers are familiar with.
I wouldn't consider the MuseScore 1.3 or 2.0 format "open" until there is available full specification of the XML file formats, and the ZIP-based container they're usually wrapped in.
Even the MuseScore core developers admit there is no save file documentation and warn against attempting to parse MuseScore's save file format: http://musescore.org/node/13837
This is not "open", and I think it's wrong to sell these engravings as "open scores" in this context.
Also, I think our views on what counts for open diverge by a few parsecs, because a) MuseScore 2.0 beta is available and will open the files, and b) the score is also available in MusicXML which you can open with just about any contemporary score editor, including MuseScore 1.3 that you have installed.
Sounds very interesting. I wonder if there will be any attempt to reflect the performance styles current when Chopin was writing. It would be fun to try to match the acoustic of the first performances as well perhaps.
One way of recovering some flavour of performance style is through students of the composer who recorded...
(Mind you, Cortot and Schnabel where not what one might call purists).
https://musopen.org/music/1254/johann-sebastian-bach/the-wel...
Shouldn't it say CC0 per the Kickstarter which says "We are creating a new digital score and studio recording of J.S. Bach's Well-Tempered Clavier (Book 1), and we're placing them in the public domain for everyone to download, own, share, and use, without any limitations."
Could someone point me to an explanation for this? How does a newbie such as myself discover the compositional techniques hidden in these?
In addition, Bach used this work to teach his students musical composition, and many of the pieces exemplify certain compositional techniques. For example, take a look at prelude #1. The entire piece is based on moving between chords by changing a single note, effectively demonstrating voice leading and harmony.
Unfortunately, the WTC was not particularly well known during Bach's time or in the century after. If I'm remembering my music history correctly, it was barely even published during his lifetime. Eventually, it was rediscovered by the Romantic composers and has been famous ever since.
I do not believe Bach actually had any hand in popularizing equal (or well) temperament. This might just be one of those legends attributed to the great composer. (We covered this in music history class when I was at university, but sadly, I don't remember the details.)
This statement ceases to be correct for very first chord change of the piece, which goes from C to Dm7. In fact, it's an incorrect description of the piece in general.
Bach was exceptionally good at writing parallel lines of music while keeping them separate and individually interesting.
The technical name for this is counterpoint. While counterpoint never quite died out - it's still used in pop and film scores - no one else has ever written counterpoint quite as dense and inventive.
And Bach didn't just write it, he improvised it, with six or even eight lines at a time.
WTC is an example of writing in different keys - identical note scales that start on different root notes.
Before modern tunings keyboards were tuned to a specific key. The other possible keys were always more or less out of tune.
With modern tunings the sourness was smoothed out. None of the keys are exactly in tune, but they're all similarly good-enough.
Bach's tuning probably wasn't quite the same as a modern tuning. But it was close enough to allow him to write a book of pieces that worked through all the keys on a single instrument without retuning - which was a novelty for the time.
WTC is also a kind of pattern book of contrapuntal and compositional techniques. So it made a huge impression - and still does.
If you're not a musician I wouldn't get too distracted by the theory. It helps to know what a fugue is and maybe a little about form, but the music is very good at speaking for itself as a self-contained experience without insisting that you understand how it was put together.
When I looked for it now, they seem to have something entirely different in its place [1]. Incidentally, they also have a talk on Bach's Preludes and Fugues [2]. I don't know if these are as good as the one I heard years ago. I'll find out tonight.
I have read a very little bit from music theory textbooks. I don't play an instrument, so the books quickly became tedious for me.
I thought it should also be noted: species counterpoint was being formalized just around Bach's time by Johann Fux[4], and modern clarifications and extensions to Baroque formal species counterpoint are largely derived from reverse-inferring rules out of Bach's masterful use of counterpoint and his inventive ways of twisting the rules (often to make his own consistent contrapuntal language). In this sense, WTC is an absolute masterclass of the rules that formed the core of the defining era of European music, which is why it's so famous.
[1] http://tobyrush.com/theorypages/index.html
[2] http://www.ars-nova.com/CounterpointStudy/intro.html
[3] http://www.listeningarts.com/music/general_theory/species/me...
I took away a lifelong lesson from this 1 hour listening session, and all it required was the score and my attention. Counterpoint is a rhetorical tool, and I think it can be fundamentally understood and appreciated even without performing the full dissection.
I also recommend Rachmaninoff's Etude-Tableau Op. 39 No. 4 in B Minor for counterpoint appreciation. That piece never fails to blow my mind. https://www.youtube.com/watch?v=hWLotW5AKjg
All the noises captured were the noises Kimiko made while creating the work.
I'm sure Archive.org and others are going to pick it up quickly.
This statement is wrong on many counts:
- The meaning of "out of tune" is culture- and epoque-dependent. For example, arabic music employs scales with 3/4 tone intervals, which to western ears might sound "out of tune", but to arab listeners would sound perfectly normal. In a similar fashion, an equal tempered major triad which would sound "in tune" to our ears (trained as they are on the now universal equal temperament) would sound "out of tune" to musicians and listeners in the 16th and 17th centuries, because they would expect a major third to be pure, i.e. much narrower than in equal temperament. In short, the meaning of "in tune" and "out of tune" is a matter of environmental conditioning, not an absolute truth.
- Since the middle ages, musicians, instrument makers and theoreticians developed tuning systems that actually allowed transposition. This is true for the hexachord system, based upon pythagorean tuning (perfect fifths of 3/2 ratio), employed in the late medieval era. This is also true for meantone tuning, a system of tuning which tempers the all fifths by a 1/4 of a syntonic comma in order to arrive at pure thirds (a 5/4 ratio interval). Both pythagorean tuning and meantone tuning are acyclical systems (that is, a sequence of 12 fifths does not make 7 octaves), but actually can be considered forms of equal tuning because all scales would sound the same (at least theoretically).
- The problem of transposition becomes an issue mainly on keyboard instruments, because they split the octave into a fixed number of notes. In both pythagorean tuning and meantone tuning, there is no enharmonic identity - C sharp does not equal D flat, etc. This would impose a practical limit on the possible tonalities that can be used. In the 16th and 17th centuries, a harpsichord or organ with 12 keys to the octave would typically be tuned meantone with C sharp, E flat, F sharp, G sharp and B flat. This would make scales such D flat major, F minor or A flat major quite unpleasant to the ear. This would also cause one fifth to be larger than pure and therefore unusable (the "wolf"). One solution would be the addition of more keys, and indeed in the 16th century many keyboard instruments were produced that included split sharps, and versions of harpsichords were developed with up to 36 keys to the octave, allowing the player to reach very remote tonalities.
- In the late 17th century people are beginning to explore less tempered versions of meantone, and Werckmeister is indeed the first to publish several methods of tuning which are "well tempered", i.e. they are cyclical (a cycle of 12 fifths) and all fifths are "good" (pure or smaller). Werckmeister was soon followed by other theoreticians, and the issue of well tempered tuning versus meantone tuning, which was practically universal in Europe back then, was a matter of hot debate at the time, even as late as the French revolution. Bach was indeed a proponent of well tempered tuning, but his contemporary the great organ maker Gottfried Silbermann insisted on employing a form of meantone tuning in his organs. Mozart, working in the late 18th century, 50 years after Bach, still used 1/6th comma meantone. The French have continued to employ meantone and meantone-derived tuning well into the 19th century.
- Well tempered tuning is not a perfect solution to the problem of transposition. In well tempered tuning each tonality would sound slightly different. Some tonalities, usually the ones closer to the C in the cycle of fifths (i.e. C major, G major, F major) would sound purer, and the more remote ones would usually sound more "spicy", "bitter" or "quirky". In some contexts, such as the performance of Bach's cantatas, the keyboard (organ) player would be required to play a transposed part, which would lead him to play in more remote, and therefore more discordant, tonalities.
- Equal temperament as a universal tuning is a relatively recent phenomenon, but one that has completely tainted our ability to listen with open ears to other ways of tuning keyboard instruments. In fact, singers and players of instruments with flexible tuning (e.g. violins) are so conditioned by equal temperament, that they play discordant, equal-tempered thirds instead of pure thirds, even when unaccompanied. By using equal temperament, we impose a uniformity of intervals and tonalities which was certainly not there in the 18th century.
- Bach's "well tempered clavier" != equal temperament. In fact, the beauty of realising the 48 on a well tempered harpsichord, unlike the equal tempered modern piano, is that each tonality would have its own character. That is the prime reason why playing Bach on the piano is wrong.
- The frontispiece of the "well tempered clavier" has led some to speculate that in it Bach has encoded his method of tuning. The first was Bradley Lehman in 2005 [0]. A more recent attempt was made by French harpsichord maker Emile Jobin [1].
[0] http://www-personal.umich.edu/~bpl/larips/outline.html [1] http://www.clavecin-en-france.org/spip.php?article52 (in French)