Wednesday, October 5, 2016

Writing music score on the spreadsheet


0


4


8


12



C4 q2 h










h'


A3








h'






G3
















F3 q4



h'










E3
















C3
















B2
















A2
















G2

q2 q3

q2 q3

q2 q3

q3 q
E2 q1


q1


q


q


























































Exactly from my spread sheet, and the score, along with the set up parameters, generates a wave file. The link has the source,maybe not  always up to date.

But it is obvous, we can set a beat 1-6 semi-tones above the ledger tone one the left.  There is no real limit to the boundaries.  The macro will organize according to the repeat count, beats per measure, and generae the proper number of measures and play the score.

The accent mark means staccato, for now, but it may mean something similar but slightly different as I learn about music.  The multi tones, chords are automatically diminished for dynamic range, but that can be disabled with a mode switch.

There will soon be a diminish symbol, in the tyop row.  That row will also contain the repeat count and ledger change, it allows scale changes within  the score.

The score can also changr musical instruments, actually, mainly because the 'pattern', the instruments typical multi-spectral response, in the time domain, is stored and reused.

What about the stereo? Dunno yet.  The possibilities range up to an exact simulation of some concert hall.  Otherwise, dunno what the musicians think, and haven;t looked.  My plan is to just leave it in as low priority.  It is more important to get different instrumentation patterns, as my phase adjustment trick removes clicks when  we change the center frequency of any instrument pattern.  I need the characteristic 'pattern' for the trumpet, for example,   The pattern has to have an identifiable start and end that the note is open and closed properly.  The end points are zero.

Multi-insrumentals? I have macros that can read and acquire he pcm codes.  Hence, the capability of combining wave files is perfectly reasonable.  It is all integer accuracy (when phase aligned) , so add and subtract of wave files would work. That is, no information loss because no compression, its all PCM.  The exception is dynamic range management because we are still finite..

 In both cases dynamic range has to be managed, so the same combinations can be done from   multiple scores at run time.  Choices on how to do all this remain. One choice is to use wave files with 32 bit or 64 bit precision, huge files.  But only use these for doing algebra, you can go through multiple passes without running,ma foul of the bounds.  Then solve he dynamic range problem on the conversion to 16 bit.

No comments: