Jeff Senn sez:
| On Jan 21, 2004, at 6:16 PM, Phil Taylor wrote:
| > Not that I think that a particularly useful approach, since as
| > Jack says, musicians who use these pitches don't think of them
| > as fractional; we really should not adopt something which is
| > totally at odds with existing musical practice.
|
| I'm not sure how it is at odds with practice -- since no one
| traditionally practices playing music by looking at the ABC notation
| :-)
...
| D-Bayati (has a slightly lower E-half-flat and a B-flat)
| K:C _6/10E _B
|
| True. Not pretty, but someone must define the tone if it will ever
| be rendered to MIDI.
True. And it's probably worthwhile to point out again that
there is a big difference here between music formatters
that only need to produce readable music on the screen or
paper, and music players that need to produce MIDI or WAV
or MP3 or some other sound file for listening. Formatters
don't need to understand intonation; players do.
It is interesting that Middle-Eastern musicians, with their
tonally complex music, have been able to adopt Western
music notation fairly successfully. But this is because
information about the scale can be encoded in a word at the
top ("bayati", "hijaz", etc.) that tells the musicians the
scale being used.
As Jack hinted, this is partly because the scales have at
most 7 distinct notes in an octave, with the qualification
that there are a lot of scales like the Western minor that
have ascending and descending forms. So you can state a key
signature at the start, name a scale to get fine tuning of
the notes, and use accidentals to get the other variant of
the scale if there is one. Western notation is even a bit
overly precise, since it has three accidentals, and any one
note rarely needs more than two in a given piece of music.
But this only works for reading the music notation, and has
the requirement that the reader understand the fine points
of the scales. Aside from naming the scale, the fine tuning
isn't represented at all.
This makes life very easy if you're writing readable music
notation. If you're trying to produce a sound file, life
isn't easy at all. What you really need is for the software
to recognize that scale name at the top, and fine-tune the
notes appropriately. This is going to take some work.
Probably the most practical approach here is to talk about
using the scale name the same way that a musician would.
That "bayati" could be looked up in a library of scales
that gives the intonation of the notes relative to the
tonic. The key-sig stuff would then be only for drawing the
music on screen or paper; an abc player would use the name
instead.
One problem that I see with the above example is that the K
line gives C as the tonic. It really should be K:D... The
notation has been suggested:
K:D_E_B "bayati"
This would mean a tonic of D, a key signature of _E_B, and
the "bayati" intonation or scale. A music formatter would
draw the _E_B signature and ignore the scale name, while a
music player would ignore the _E_B keysig stuff and look up
the scale name to tune each note. This is similar to how,
if you tell a musician "D major", you don't list the
accidentals, because that will be obvious.
How to deal with other accidentals within such a scale, I
don't know. Again, a formatter only has to draw them. But a
music player has to somehow figure out the intonation.
To subscribe/unsubscribe, point your browser to: http://www.tullochgorm.com/lists.html
