Agemixer (02:05:49)
probably some other better ideas, but that's one "easy" approach to explain.

Agemixer (02:04:51)
lda sample,x , sta SIDsampl, inx, nop*6, lda sample,x , sta SIDsampl , bit $dcod .... or something, double writes during badline. Probably a scanline table lookup to check whether a badline occurs there.

Agemixer (01:58:49)
otherwise at badline it requires kinda little buffer to overcome that little timing issue, or speedcode

Agemixer (01:56:48)
badlines can be avoided though, with a correct timing.

Agemixer (01:56:17)
yes, it removes the badlines, so all the time can be reserved for CPU

Agemixer (01:55:45)
you mentioned blanking screen..

Rapture (01:42:50)
oki

Agemixer (01:40:37)
CPU

Agemixer (01:40:24)
so it's pretty much about the CPau instructions speed...

Agemixer (01:40:02)
Rap: only that much SID can do for the sample. mixing samples require software mixing of course. Or. you can of course engage all 3 voices as samples that way... but it eats almost all the available CPUtime unless some clever buffer is involved,