Dirk Laurie wrote

>Comment: MIDI transposition should be independent of the printout.

There are three different pitch levels to worry about (relative to "concert 
pitch"): 

1. PMX input
2. Printed output
3. MIDI output

All could in principle be defined differently and independently, and could be 
different for each instrument. And the printed output could be different for 
score and parts. So you could think of each one as characterized by a 
deltap(type,iinst,cat), where type is 1, 2, or 3 above; cat is either score or 
part (but doesn't really apply to PMX input, type=1); and deltap is the 
difference between the item and concert pitch. There are a few constraints, 
e.g., you'd surely want deltap(3,iinst,"score") to be the same for all iinst, 
although not necessarily zero. Throw in issues about key signatures, and maybe 
changing the amount of transposition in midstream such as if you wanted to 
switch from B-flat to A clarinet, and the number of possibilities gets 
absolutely mind-boggling.

So far, I've assumed in effect that deltap(1,iinst,"score")=0 always, and 
allowed deltap(2,iinst,"score") to be whatever you want for each iinst, and 
I've allowed for it to change in midstream along with iinst's key signature. I 
haven't squarely faced up to any of the other possible combinations. So if all 
you ever wanted to do was to print scores and to enter everything in concert 
pitch, I'd be done. And if you were willing to follow the (somewhat klunky) 
suggestion I gave, you could also control the printed output of the parts (but 
then you could have done that before by using the old full-score 
transposition/key change command in a part-only comment, at least for a 
transposition that didn't change in midstream).

>Example 1:
>
>Composer writes a piece in C for singer, solo instrument and piano,
>notated as it should sound.  A performing version is made in which
>the solo is played on a B♭ clarinet.  The instrumental part is
>transposed up by two semitones to a key with two sharps, but the MIDI
>should stay the same.

I had hoped that this might work OK even though I didn't think about it. But I 
just check and it doesn't work. However, it should be easy, since all it means 
is the MIDI has to ignore the transposition. I anticipate having it implemented 
very soon.

>Example 2:
>
>Composer writes the same piece but notates it right from the start
>with the clarinet part already transposed.  The instrumental part
>is transposed by no semitones to a key with two sharps, but the MIDI
>should sound two semitones lower.

I've been avoiding facing up to the possibility of entering info in the 
transposed key, let alone working out how to get MIDI to deal with it. I have 
vague notions of somehow internally doing a reverse 
transposition. This will take longer to work out.

--Don



-------------------------------
TeX-music@tug.org mailing list
If you want to unsubscribe or look at the archives, go to 
http://tug.org/mailman/listinfo/tex-music

Reply via email to