I'm confused about the source code. Why do you use harmonics 1 through 6 in one place, but then "1:nharmonics*2" (= 1 through 12)?
and you say "(1/uppernoteharmonic) * getroughness(f1s*uppernoteharmonic, f2s*lowernoteharmonic, 1, 1)"
but shouldn't the amplitude be inside the getroughness call? Like the fundamentals should be compared with both at amplitude 1, then the fundamental at 1 with the second harmonic of the other note at 1/2, then the fundamental at 1 with the 3rd harmonic at 1/3, etc?
and you say "(1/uppernoteharmonic) * getroughness(f1s*uppernoteharmonic, f2s*lowernoteharmonic, 1, 1)"
but shouldn't the amplitude be inside the getroughness call? Like the fundamentals should be compared with both at amplitude 1, then the fundamental at 1 with the second harmonic of the other note at 1/2, then the fundamental at 1 with the 3rd harmonic at 1/3, etc?
This was a fun post for me. Do you think you could post that picture again, but add the letters for the notes so we can see it clearly?