I think with this system you will end up with too many large difficult and uncatchy jumps. Plus similar phone numbers will sound similar which is not what you want.
That’s what I did, though with do on 1 and mi′ on 0 (treating it as 10). I’m not sure what similarity metric is most relevant here, but in case of near-collisions, small additions of supporting rhythm or harmony could help nudge some sequences apart; swing time would give added contrast to even vs odd positions, for instance. Anyway, it sounds like this might not generalize well across people…
Aside: this is a time when I’m really appreciating two-axis voting. The vote-signal of “that’s mildly interesting but really doesn’t seem like it’d work” is very useful to see in a compact form, even though the written responses contain similar information in more detail.
I think with this system you will end up with too many large difficult and uncatchy jumps. Plus similar phone numbers will sound similar which is not what you want.
That’s what I did, though with do on 1 and mi′ on 0 (treating it as 10). I’m not sure what similarity metric is most relevant here, but in case of near-collisions, small additions of supporting rhythm or harmony could help nudge some sequences apart; swing time would give added contrast to even vs odd positions, for instance. Anyway, it sounds like this might not generalize well across people…
Aside: this is a time when I’m really appreciating two-axis voting. The vote-signal of “that’s mildly interesting but really doesn’t seem like it’d work” is very useful to see in a compact form, even though the written responses contain similar information in more detail.