Sheet music

Found in Community

Groups

Piano

371 discussions • 8.1K scores • 52.9K members

Piano Compositions

545 discussions • 3.6K scores • 24.5K members

Short pieces for piano

249 discussions • 2.8K scores • 21.8K members

Discussions

There are lots of democratic candidates running against Trump 2020. Who do you think will win of all of them... not who you want to win. And will they stand a chance against Trump? I'm interested to hear what ya'll think about this
I can reach a ninth (one note more than an octave) and with EXTREME stretching a tenth. What can you reachBonus: Did you know Rachmaninov could reach a 13th? That’s over a foot, from C to the A on the next octave. But hey, he was 6’ 6”
334 years ago on this calendar (that is an issue) day, Johann Sebastian Bach was born in Eisenach, Thuringia (there was no "Germany" yet), arguably the greatest musician-composer to date born on this planet, and without doubt, the man whose proficiency at and practice of Counterpoint, the subject of this group, are now accepted worldwide as the gold standard, goal, and almost definition of this discipline and art.Today, Google offers "as doodle" (go to google.com front page) an automatic, machine-learned chorale harmonizer -- you specify two measures of quarters and eighths (eighths by moving the mouse to the center of a beat), and it provides four-voice harmony which it claims to be in the "style of Bach".  This is hardly the first attempt at machine-learned composition, or chorales in specific - the names Kemal Ebcioglu and David Cope come immediately to mind for me, and I have seen many others more recent (Search in Google for "automatic chorale composition"). I think that this display will help many people be more aware of what machine-learning can accomplish, as well as its limits.  I have seen, in one output from this computer contrapuntist, parallel fifths and octaves, minor seventh chords neither prepared nor resolved, and wholly unexplained dissonances and even sixteenth-note leaps.  All of which are fine in jazz, but the claim is "Bach chorale style".  This particular machine-learned model has a long way to go.But it also demonstrates that this subject can be learned, even if the machine is a poor student.  And, as with Google language-translation, it is a poor teacher for human learners, and denies us the insights, models, and techniques by which a human teacher, living or from the pages of a book, can perpetuate knowledge of the domain.  Explicit computer code and "rules" can be learned from, or benefit from learning, but statistically-trained models cannot help us learn any more than a bird can explain flight in its song!Happy Birthday, Sebastian (as he was called)!