Multiple Array Read, using Assembler

[HughBanton]

Definitely works Martin. Quite easy to test because I can replace my 'changed' line with int 1 to universally defeat all the hops.

I checked 1-4 notes, 5-8 notes, and 9+ notes, and I get roughly 27% better cpu %ge in all cases with the 'changed' hops reinstated. (yay!) Can't detect any poly issues.

While on this subject, (and irrespective of the above), I've noticed for a long time that if you hold down 1 note, check the cpu%, then press another 4 notes, then release 4 of them again, cpu% only drops to an intermediate figure, not the original. Until you release all notes and start again.

Can anyone replicate that? Or is it just a quirk of mine?