> > Would the following data be a reasonable specification of a chord:
> > 1) A tonic
> > 2) A collection of intervals above the tonic, specified in semitones
> > (this may or may not include 0).
> > 3) A bass note.
>
> This is how it's implemented (more or less) now.
Ah.... yes, except that (looking at chord.hh) it seems that the
intervals are not specified in semitones, but in musical_pitches.
This is significant, I think, because it ties the implementation of
chords to major scale harmony.
I had some problems with this when changing chord-names.scm (I wasn't
using \chords mode). A single chord could be written in more than one
way, eg: <c ees ges a> and <c dis fis a>. This meant having two
entries in chord-names.scm for this chord. This wouldn't have happened
if I could specify the chords in terms of semitone intervals above the
root. I'm not sure though - does the same thing happen in \chords mode?
What if one tries, say, ees:m7 and dis:m7?
gav.
>
