Re: octave equivalence

I think if one really wants to investigate the octave
invariance phenomena, then an experiment closer to
the following is more in order

1) melody
2) melody transposed by octave
3) melody transposed by non0-octave
4) melody with individual notes transposed by octave
but melodic contour is maintained
5) melody with individual notes transposed by octave
and no maintenance of melodic contour

All of the above with varying amounts of wrong notes
which do or don't preserve melodic contour.

The question is "is this the same melody" as the control.

Yes or no isn't sufficient, you probably need someone
in there to decode "it is, but it sounds a little higher"
and "it is, but theres something about it that makes me
queezy in the middle".

What would my suppositions be? That melodic contour is
much more important than octave invariance. Actually, I
think this is pretty trivial. Play "Happy Birthday" for
someone in minor, for instance, they'll recognize it.
Transpose it and do the same, they'll still recognize
it. Those of you with kids should try "Spoonful of Sugar"
from Mary Poppins in whole-tone, drive your kids nuts!

So, I probably agree with Monz (and others) that
equivalences at prime 2 are pretty substantial, at
prime 3 present but less substantial, but I don't know
if that psychoacoustic phenomena is necessarily a
strong force in musical perception (as my proposed
experiment would no doubt show).

Bob Valentine

--- In tuning@y..., Robert C Valentine <BVAL@I...> wrote:
>
> I think if one really wants to investigate the octave
> invariance phenomena, then an experiment closer to
> the following is more in order
>
> 1) melody
> 2) melody transposed by octave
> 3) melody transposed by non0-octave
> 4) melody with individual notes transposed by octave
> but melodic contour is maintained
> 5) melody with individual notes transposed by octave
> and no maintenance of melodic contour
>
> All of the above with varying amounts of wrong notes
> which do or don't preserve melodic contour.
>
> The question is "is this the same melody" as the control...
>
> Bob Valentine

Bob Wendell:
Yes, a much better test! But how do you propose to change individual
notes (#4 above) without changing the melodic contour? This is by
definition impossible, since transposing any fewer than ALL the notes
by an octave automatically implies a change of melodic contour, since
at least one interval will be inverted, changing both its melodic
direction and in all but a tritone, also the pitch distance.

> From: "robert_wendell" <BobWendell@technet-inc.com>
> Subject: Re: octave equivalence
>
> --- In tuning@y..., Robert C Valentine <BVAL@I...> wrote:
> >
> > 4) melody with individual notes transposed by octave
> > but melodic contour is maintained
> >
> > The question is "is this the same melody" as the control...
> >
> > Bob Valentine
>
> Bob Wendell:
> Yes, a much better test! But how do you propose to change individual
> notes (#4 above) without changing the melodic contour? This is by
> definition impossible, since transposing any fewer than ALL the notes
> by an octave automatically implies a change of melodic contour, since
> at least one interval will be inverted, changing both its melodic
> direction and in all but a tritone, also the pitch distance.

I'm using melodic contour loosely here meaning instantaneous up'n'down
rather than "its higher than the last note but not as high as the one
before that".

So Happy birthday : G3 G3 A3 G3 C4 B4
#4 : G3 G3 A4 G3 C4 B4

which IS much different than
: G3 G3 A2 G3 C4 B4

Bob Valentine