Re: AP development for microtonality (was: new perfect pitch paper f

Hi George & Carl,

> > In other words, accordion basses
> > are set up to *decrease the perception of pitch height while
> > reinforcing (through octave doubling) the perception of pitch
> > color!

> What a sweet concept. Somebody should make AP quiz software
> on this principle.

I've got a quiz in Tune Smithy which focuses on trying to develop
good relative pitch for intervals. But I could also generate
Shephard tones, and use it for an AP quiz as well, may
give it a go for the new release.

I've been working on pitch memory btw following that
piano on-line book. What I've been doing is to play a short
phrase and then play it back immediately afterwards using
the very short term memory you have for a musical note
immediately after you hear it - that lasts for maybe a few
seconds after the note (if you don't already have AP anyway).
So - very complete with the exact timbre of the instrument
etc.

So - then the idea is to reverse the process (not my idea, from the
piano book) so you play the sound back in your mind immediately
before you play it for real. That seems to involve AP because
if you do it soon after then you find you play it back in your
mind at the same pitch. I mean if you do it e.g. in the same
session.

Then you may find that if you play it back in your mind even on
the next day before you play it on the instrument, that it is still
at the same pitch, at any rate that's what I find for one of my
Teleman pieces - the first note is a G and if I play it back in my
mind now, then play it on the instrument, then I hear the same
pitch. The book advised one to do it that way first and not attempt
to reproduce the pitch oneself until you get it in your mind
- to avoid limitations that you can only play things in your mind
that you can produce using your own vocal chords or whistle
etc.

So anyway I did that for a bit, now if I play it in my mind,
and whistle the pitch I heard in my mind to test that
it really is my memory of pitch - I always seem to come up with a G.

Also - could start the tune from almost any point in my mind now,
so it seems, not very far from the point where I could
whistle any pitch that occurs in the tune in pitch.
By learning a variety of tunes one could fairly quickly
reach the point where one can whistle any of the twelve
pitches of 12t in pitch. For microtonal music one would
need to learn some microtonal pieces similarly so that
one can play them back in ones mind too.

But I can't yet instantly name a heard note
as a G if it is the same pitch as the first note in the tune,
I'd have to play the tune in my mind first and then see if
it sounds like the same note. Which one would think
would be a very small step but it seems there is
more to it than that. The mind is funny sometimes,
how it works!

BTW though - I have a constant high pitched very quiet tone
in one ear - like an internal tuning fork, I know that's often
a nuisance but this is so quiet I only notice it when it is quiet
around me. But it somewhat muddies the water when it
comes to trying to work out if I am using AP or the
internal "tuning fork" to identify a note, not sure how one
could ever be certain about it.

Anyway back to the quiz, it seems to me that one thing
that's very confusing if you tend to hear notes in terms
of component pitches is that if you hear a G, actually
it is a whole blend of different pitches - even on a pure
sounding instrument like flute or recorder, the note is
coloured quite a bit by the inharmonic pitches in the
attack (otherwise it would sound like a sine wave)
- which confuses the issue, since you are trying to
remember a cluster of pitches rather than an individual
pitch.

So it sounds as though it may help - as an adult anyway - to use
shephard tones made up of pure sine waves
in octaves, to give fewer pitches to remember.

Also the octaves make sense to me as a simplifying
concept since octave equivalence is something that
for me isn't very strong. I can hear it but the same
note in different octaves is notable more for the
difference in the pitch sound than the similarity.

To hear octave equivalence, if playing it
as exactly as I can to get really in tune, and for two
notes played one after another (so not using beats),
I tend to listen out for the first partial in the lower pitched
sound and compare that in my mind with the higher pitched note.
That's easier to hear than the octave equivalence of the
fundamentals for me. Also I hear many more than
twelve distinct pitch sounds to an octave. All that
makes the task of identifying pitches by name much
harder as you have maybe a hundred or so "pitch colours"
per octave, and several octaves.

I wonder if that may be what it sounds like to a child?
If so then it may help a lot to have a fixed pitch instrument
if you want to learn the 12-t note names.

It would be interesting to know - do most people
who are able to sing very exactly in pitch, or if using
a sine wave generator to generate a pitch,
when asked to sing a note an octave higher for
an instrument with inharmonic partials like
the piano - do they sing in pitch with the
inharmonic partial, or do they sing an octave
above the true fundamental of the note?

As a child my interests focused on maths and science
much more than music, so at the moment maybe it
is a bit like a child learning to identify pitches.
Maybe lots of adults are in the same position - if so
then trying to test their AP abilities by asking them to
identify notes - or even to sing the same pitch on
successive days - may be not enough to pick up on it
as the mental pitch scene is just too complex for them
and maybe even if they hear sounds in terms of
component pitches, they have given up trying to give names for
anything at all. So some adults may need training for some time
before they can begin to answer the questions.
At any rate I'm sure that right now if tested for
AP I'd come out as one of those who don't have it.
But (whether using the tuning fork in my ear or
through true AP is hard to know) it seems that I
could if I were to work on it, to come to the point
where I could reproduce any pitch I like, and from
that come to be able to identify notes by name.
Whether it would ever be "instantly" like true AP
is hard to tell. But in principle it seems possible that
it could become quite fast anyway if I were to really
work on the mental play - as a biproduct, because
it seems that the mental play is naturally in pitch
- that you don't play it back at a different pitch from
the one you heard it in.

But I wonder - is that true for others - when you play a
tune back in your mind, if you do it before you play the
sound for real - if you have a strong sense of relative
pitch - what happens? Is the tune you hear in your
mind the same pitch as what follows - or does it
drift in pitch so some days it is lower or higher in pitch?
Or does it have an indeterminate pitch? How does it
compare with the short term memory you have of
a melodic phrase immediately after you play
a note - is it the same type of thing or is there
a difference in character?

Robert

--- In tuning@yahoogroups.com, "Robert walker" <robertwalker@...>
wrote:
>
> Hi George & Carl,
>
> > > In other words, accordion basses
> > > are set up to *decrease the perception of pitch height while
> > > reinforcing (through octave doubling) the perception of pitch
> > > color!
>
> > What a sweet concept. Somebody should make AP quiz software
> > on this principle.
>
> I've got a quiz in Tune Smithy which focuses on trying to develop
> good relative pitch for intervals. But I could also generate
> Shephard tones, and use it for an AP quiz as well, may
> give it a go for the new release.

Hi Robert, I'd be very interested in something like this. If either
you or Carl are going to follow through on any activity involving AP
development, then we should continue that discussion off-list.

> I've been working on pitch memory btw following that
> piano on-line book. ...
>
> So anyway I did that for a bit, now if I play it in my mind,
> and whistle the pitch I heard in my mind to test that
> it really is my memory of pitch - I always seem to come up with a
G.

I believe that the act of "playing something back in your mind" is a
crucial part of AP development, so I think that you're on the right
track.

> ...
> But I can't yet instantly name a heard note
> as a G if it is the same pitch as the first note in the tune,
> I'd have to play the tune in my mind first and then see if
> it sounds like the same note. Which one would think
> would be a very small step but it seems there is
> more to it than that. The mind is funny sometimes,
> how it works!

If the note is in a different timbre or different octave (or is
different in some other way from what you heard when you heard the
tune), then you may have difficulty with note-naming, because you
haven't sufficiently generalized your perception of pitch-color. If
you have an electronic keyboard, you could try changing the timbre
and playing the same tune for several days, then change to another
timbre and do the same thing for. Also, I think it would be best to
play the melody in octaves, without any accompaniment.

> BTW though - I have a constant high pitched very quiet tone
> in one ear - like an internal tuning fork, ...

Just do your best to ignore it.

> Anyway back to the quiz, it seems to me that one thing
> that's very confusing if you tend to hear notes in terms
> of component pitches is that if you hear a G, actually
> it is a whole blend of different pitches - even on a pure
> sounding instrument like flute or recorder, the note is
> coloured quite a bit by the inharmonic pitches in the
> attack (otherwise it would sound like a sine wave)
> - which confuses the issue, since you are trying to
> remember a cluster of pitches rather than an individual
> pitch.

If you're being distracted by partials that aren't merging together
into a single perceived pitch, then you need to use a different
timbre.

> So it sounds as though it may help - as an adult anyway - to use
> shephard tones made up of pure sine waves
> in octaves, to give fewer pitches to remember.

I don't quite understand what you mean by this. You shouldn't be
listening to the component partials of a sound, but you should be
hearing a single (merged) sound having a single pitch-color.

> Also the octaves make sense to me as a simplifying
> concept since octave equivalence is something that
> for me isn't very strong. I can hear it but the same
> note in different octaves is notable more for the
> difference in the pitch sound than the similarity.

Different octaves of the same note have different pitch height but
the same pitch-color, so what you're discussing is *not* what one
would want to be listening for to develop AP.

> To hear octave equivalence, if playing it
> as exactly as I can to get really in tune, and for two
> notes played one after another (so not using beats),
> I tend to listen out for the first partial in the lower pitched
> sound and compare that in my mind with the higher pitched note.
> That's easier to hear than the octave equivalence of the
> fundamentals for me. Also I hear many more than
> twelve distinct pitch sounds to an octave. All that
> makes the task of identifying pitches by name much
> harder as you have maybe a hundred or so "pitch colours"
> per octave, and several octaves.
>
> I wonder if that may be what it sounds like to a child?

I don't think children will likely be hearing partials distinct from
the fundamental. When I was young I didn't hear the accordion basses
as multiple pitches in octaves or a melody played in octaves on the
piano as separate pitches. It was only when I was older that I heard
the sounds as separate tones in octaves.

> If so then it may help a lot to have a fixed pitch instrument
> if you want to learn the 12-t note names.

Yes.

> It would be interesting to know - do most people
> who are able to sing very exactly in pitch, or if using
> a sine wave generator to generate a pitch,
> when asked to sing a note an octave higher for
> an instrument with inharmonic partials like
> the piano - do they sing in pitch with the
> inharmonic partial, or do they sing an octave
> above the true fundamental of the note?

This (and the remainder of your questions and speculations) I'm not
going to attempt to answer, because there are too many variables to
take into account.

--George

Hi George,

> I don't think the objective is to completely eliminate differences in
> pitch height, but rather to *emphasize pitch color* (over pitch
> height) by reinforcing it with multiple octaves. It's a lot like the
> difference between trying to perceive green by seeing a green object
> vs. walking into a room filled with green objects.

Oh well that's very easy - in FTS I have "custom voices" which consist
of several midi instruments played simultaneously - so the player
plays a single note on the keyboard say, and they hear as many
instruments as they like playing that note simultaneously.
It can be the same instrument or different instruments, and
various other options - the option of interest here is that they
can play at fixed intervals from the note the player played.

My idea would be just to have a quiz where the program plays
a note doubled in octaves (or several octaves), and asks the
user to guess which note it is - and of course let them use any
tuning for the underlying scale to pick the note from.
Once I add in the Sagittal note names after the release it
would be able to show the note name in Sagittal too.

Probably would be something you usually would do later on
once you have gone through the mental play stage and can already
play tunes in the desired tuning in your mind, and just want
to firm up your grasp of the note names for the tuning.
Anyway would be fun to try it out and see if it works and
whether the octave doubling helps.

At the moment I'm doing mostly debugging, in the lead
up to the release, but do just a little bit of extra coding
from time to time, like a day or so per week say, and since
this is a small change in the already programmed
quiz, I may well do it before the release. I thought I might make
it into a separaete task for 3.0 anyway to make it easier to find it
and use it.

Anyway if I do get the time to program it then we can discuss
it off-list or in a separate group as you suggest.

To explain the bit about octave differences, from my own experience,
I've not made a big distinction yet between pitch height and
pitch colour in an octave - and of course as you suggest,
if the aim is to develop conventional AP then one should make
that distinction and aim to perceive pitches using
octave reduced pitch classes,

However, from my own experience, I wonder if the normal octave
doubling type pitch colour of AP may be one
of several options. E.g. I wonder, if one doubled at 3/1, might one
be able to learn to perceive a different set of AP like pitch
classes, for the Bohlen Pierce scale? I wonder what
happens if one is a long term member of a particular
gamelan ensemble playing a particular instrument,
whether one might develop non octave pitch colours.

Also - I'm not sure whether I want to develop AP myself actually
as my weakness is in relative pitch distinctions and if one had
to choose one or the other, though both are useful, of the two,
perception of relative pitch for intervals seems a more useful
ability to have.

But developing stronger mental play seems a good thing too,
and it seems in my case anyway that a form of AP would
result from developing that. I imagine developing both
at once may be optimal.

With AP - I wonder if one can combine it with
a relative pitch type system as well?

You'd imagine that might be ideal - to have the absolute pitch
perception, and a secondary note name system
that you can slide up / down to any root you like.
Another thing I sometimes do is to practice playing the
same tune in all the keys or at least in as wide a variety
of keys as I can easily manage. Or just a phrase,
play a phrase in all the twelve keys - if in 31-et I suppose
one would do all the 31 keys too - go round the circle
of fifths. The idea there is to strengthen my perception of
relative pitch.

The one bane of AP is when you hear someone and they
sound out of tune in the sense that every single note they
play - even a single note - sounds out of tune, e.g.
every single note they play is a tiny bit flat. That's
happened to me once or twice, while listening to
early music, doesn't happen any more, but
something like ten years ago I remember it
happening from time to time, when I had a 'cello
and practiced it a lot - and it was a nuisance
because you couldn't enjoy music sometimes
because though in-tune in relative pitch
terms, it sounded out of tune in this sense.

I imagine that the solution to that is just to make sure
you develop a sufficiently good sense of relative
pitch at the same time as you develop AP
- I'd be interested to know what your experiences
have been in this area yourself.

The pitch in my ear is not a nuisance at all
- normally I only notice it when there is no music playing
- it is very high pitched and extremely quiet, most would
probably describe it as more like a quiet hiss han a musical note
- actually several pitches, but one prominent one.
I'm used to ignoring it - perhaps its one of those things where
the mind is able to edit it out because it doesn't change.

> I don't think children will likely be hearing partials distinct from
the fundamental. When I was young I didn't hear the accordion basses
as multiple pitches in octaves or a melody played in octaves on the
piano as separate pitches. It was only when I was older that I heard
the sounds as separate tones in octaves.

> If you're being distracted by partials that aren't merging together
into a single perceived pitch, then you need to use a different
timbre.

Okay that' makes sense, sounds as if for most listeners
there won't be that much advantage in using sine waves
for the instrument. Though I wonder if it may explain
why players of an instrument may find it easier
to learn to identify pitches on their own instrument
to start with.

Even if they hear it as a merged single pitch, it may
be that the "colour" they perceive for that pitch is
affected by the other constituent partials that make
up the instrument timbre, even though they merge
them together.

I tend to hear most instruments with some element
of unmerged partials. E.g. on recorder, I hear
the 2/1 and a very faint 3/1 on most notes.
But that's probably very listener dependent.
It's probably because I do a lot of listening to
component pitches of notes. I can ignore
them, and let the partials merge into a single
note - but with just about any instrument
I can hear the partials if I pay attention to them,
especially in a long sustained note.

BTW by "colour" here I don't necessarily
mean like a visual colour, as I gather it is listener
dependent. I wonder if AP always leads to some element
of synaesthesia or if you can have AP with no
cross over to other senses at all? One would sort
of expect if one had never heard of it before, that
you could just hear a pitch in the same way as you
see a colour, without any element of any other sense
involved. I'm not sure what it is in my case
- maybe a colour, but it isn't developed enough
to know quite what it is.

Some of the speculations and questions are meant here to be
open ended, so if some of them are interesting maybe
they may start up other discussions.

BTW I've been taking a few days off from coding.
Back to coding again tomorrow, so if I don't join
in the debate much more, it's probably just that
I'm pre-occupied with bug fixing or something.

Robert

--- In tuning@yahoogroups.com, "Robert walker" <robertwalker@...>
wrote:
>
> Hi George,
>
> > I don't think the objective is to completely eliminate
differences in
> > pitch height, but rather to *emphasize pitch color* (over pitch
> > height) by reinforcing it with multiple octaves. It's a lot like
the
> > difference between trying to perceive green by seeing a green
object
> > vs. walking into a room filled with green objects.
>
> Oh well that's very easy - in FTS I have "custom voices" which
consist
> of several midi instruments played simultaneously - so the player
> plays a single note on the keyboard say, and they hear as many
> instruments as they like playing that note simultaneously.
> It can be the same instrument or different instruments, and
> various other options - the option of interest here is that they
> can play at fixed intervals from the note the player played.

Sounds good.

> My idea would be just to have a quiz where the program plays
> a note doubled in octaves (or several octaves), and asks the
> user to guess which note it is - and of course let them use any
> tuning for the underlying scale to pick the note from.

You're a good man, Robert! I'd say, "Go for it!"

> Once I add in the Sagittal note names after the release it
> would be able to show the note name in Sagittal too.

That would be a written name you're referring to. As I said
elsewhere, the presently proposed Sagittal spoken names are IMO too
cumbersome. Dave Keenan & I hope to get back to you soon about
Sagittal implementation in FTS. (Drat! I still haven't gotten my
reply off to Dave. He's probably gotten pretty sore at me by now for
putting him off so long.)

> Probably would be something you usually would do later on
> once you have gone through the mental play stage and can already
> play tunes in the desired tuning in your mind, and just want
> to firm up your grasp of the note names for the tuning.

Yes, that's what I figured. I think it would be good if the user
could have the option of being able to choose a subset of an ET
(e.g., only the tones of 31-ET in a chain of 5ths from Gb to A#).
Perhaps that could be done by loading a .scl file containing only the
tones to be contained in the test, so you could then test for any set
of tones in any tuning, such as JI. Of course, you'd need to allow
the user to specify a pitch standard.

> Anyway would be fun to try it out and see if it works and
> whether the octave doubling helps.

Yes, and I'd look forward to trying it out with 31-ET.

> ...
> Anyway if I do get the time to program it then we can discuss
> it off-list or in a separate group as you suggest.

Off-list would be fine.

> To explain the bit about octave differences, from my own experience,
> I've not made a big distinction yet between pitch height and
> pitch colour in an octave - and of course as you suggest,
> if the aim is to develop conventional AP then one should make
> that distinction and aim to perceive pitches using
> octave reduced pitch classes,
>
> However, from my own experience, I wonder if the normal octave
> doubling type pitch colour of AP may be one
> of several options. E.g. I wonder, if one doubled at 3/1, might one
> be able to learn to perceive a different set of AP like pitch
> classes, for the Bohlen Pierce scale?

I think not. Tones having the same pitch color must have their
frequencies in a ratio that can be expressed as a power of 2 (an
octave or multiple octaves) or something within a few cents of that
(i.e., tempered octaves). I don't subscribe to the idea that
anything significantly different than an octave (or multiple
thereof), such as 1:3 or 1/2-octave can be an "interval of
equivalence", because our brains aren't wired to hear the same pitch-
color for tones separated by those intervals. (I recall that Paul
Erlich said something to the same effect. Also, Monz, in msg.
#68709, refers to the "affect" of prime number 2 as "similitude".)
If we're describing tonal systems with multiple chains of generators
separated by an interval other than an octave (such as pajara or
ennealimmal), then we should use the term "period" for that purpose.
(Hmm, I'm wondering how many angry cards and letters I'm going to get
for saying that.)

> I wonder what
> happens if one is a long term member of a particular
> gamelan ensemble playing a particular instrument,
> whether one might develop non octave pitch colours.

With non-harmonic timbres one might hear octaves "tempered" by 10 or
20 cents as having the same pitch-color, but if it's more than a
comma or so, it'll probably be too much of a stretch.

> Also - I'm not sure whether I want to develop AP myself actually
> as my weakness is in relative pitch distinctions and if one had
> to choose one or the other, though both are useful, of the two,
> perception of relative pitch for intervals seems a more useful
> ability to have.

Yes, I agree.

> But developing stronger mental play seems a good thing too,
> and it seems in my case anyway that a form of AP would
> result from developing that. I imagine developing both
> at once may be optimal.
>
> With AP - I wonder if one can combine it with
> a relative pitch type system as well?

If by "relative pitch type system" you mean the ability to perceive
and think using AP and RP simultaneously, then I'd have to say yes,
because I do that myself. In fact, I can more readily identify
*intervals* in 31-ET than the specific pitches involved, because the
psychololgical differences are so obvious. If you can tell the
difference between a major, minor, and neutral 3rd (or 6th, or 2nd,
or 7th), then you'll understand what I'm talking about.

> You'd imagine that might be ideal - to have the absolute pitch
> perception, and a secondary note name system
> that you can slide up / down to any root you like.
> Another thing I sometimes do is to practice playing the
> same tune in all the keys or at least in as wide a variety
> of keys as I can easily manage. Or just a phrase,
> play a phrase in all the twelve keys - if in 31-et I suppose
> one would do all the 31 keys too - go round the circle
> of fifths. The idea there is to strengthen my perception of
> relative pitch.

With a generalized keyboard you would hardly have to think about
which keys to press: just play by interval-vectors and listen.

> The one bane of AP is when you hear someone and they
> sound out of tune in the sense that every single note they
> play - even a single note - sounds out of tune, e.g.
> every single note they play is a tiny bit flat.
>
> That's
> happened to me once or twice, while listening to
> early music, doesn't happen any more, but
> something like ten years ago I remember it
> happening from time to time, when I had a 'cello
> and practiced it a lot - and it was a nuisance
> because you couldn't enjoy music sometimes
> because though in-tune in relative pitch
> terms, it sounded out of tune in this sense.

I've never heard things in different pitch standards that way. I got
an advertisement for an AP training course way back in the 1980's,
and the author (it might have been David Burge, but I'm not sure)
said something to the effect that off-pitch music, with the intervals
of the proper sizes, didn't really sound out of tune to him either,
just color-shifted. Sometimes on a classical radio station I'll hear
early music played at lower-than-A440 pitch (possibly in something
close to meantone temperament), which reminds me of the general
effect of the pitch-colors I've heard on my Scalatron in 31-ET in the
flat keys (which might be described as a color-shift).

> I imagine that the solution to that is just to make sure
> you develop a sufficiently good sense of relative
> pitch at the same time as you develop AP
> - I'd be interested to know what your experiences
> have been in this area yourself.

If you don't already have AP, then I don't know whether you can work
on AP at the same time as RP, because developing AP involves
listening for pitch-colors (which are irrelevant to RP). But if you
already have both AP and RP for 12-ET, then I think that microtonal
AP and RP would naturally be developed simultaneously (although I'd
say that microtonal RP comes more easily).

> ...
> BTW by "colour" here I don't necessarily
> mean like a visual colour, as I gather it is listener
> dependent. I wonder if AP always leads to some element
> of synaesthesia or if you can have AP with no
> cross over to other senses at all? One would sort
> of expect if one had never heard of it before, that
> you could just hear a pitch in the same way as you
> see a colour, without any element of any other sense
> involved. I'm not sure what it is in my case
> - maybe a colour, but it isn't developed enough
> to know quite what it is.

I'd say that crossover between aural and visual color perception is
largely, if not completely, subjective, a matter of chance
association. The only possible objective association between them I
can think of is that the frequency of middle C multipled by 2^41 (41
octaves up) corresponds to a light frequency that we perceive as
green, but I don't hear "C" as "green" (or any other visual color,
for that matter). The only sound/color association I can recall is
that the 2nd B-flat below middle C on the piano at one time reminded
me of "brown", which, coincidentally, has a chroma value (orange-
like) that corresponds rather closely to a forty-something octave
downward transposition of the frequency of orange light (which works
best, BTW, with a pitch standard of C=256 hz.)

> Some of the speculations and questions are meant here to be
> open ended, so if some of them are interesting maybe
> they may start up other discussions.

Some of these things tend to get OT, however. :-(

> BTW I've been taking a few days off from coding.
> Back to coding again tomorrow, so if I don't join
> in the debate much more, it's probably just that
> I'm pre-occupied with bug fixing or something.

I didn't know we were having a debate (but I don't care to debate
that).

Anyway, if your bugs are broken, you'd better get them fixed or
something. ;-)

--George

George D. Secor wrote:

> I'd say that crossover between aural and visual color perception is > largely, if not completely, subjective, a matter of chance > association. The only possible objective association between them I > can think of is that the frequency of middle C multipled by 2^41 (41 > octaves up) corresponds to a light frequency that we perceive as > green, but I don't hear "C" as "green" (or any other visual color, > for that matter). The only sound/color association I can recall is > that the 2nd B-flat below middle C on the piano at one time reminded > me of "brown", which, coincidentally, has a chroma value (orange-
> like) that corresponds rather closely to a forty-something octave > downward transposition of the frequency of orange light (which works > best, BTW, with a pitch standard of C=256 hz.)

I think there's a fundamental difference between pitch perception and color perception. (Besides the fact that two colors together just looks like a different color). You can transpose music to a different key and it still sounds more or less the same (unless it goes into a different register of the instrument). Small changes in hue might go unnoticed, but if you shift the hue of a picture significantly, it looks unnatural. Red is fundamentally a different color from green, as yellow is different from blue. The kind of relation between tones that allows for transposition doesn't have any equivalent with color. Does the interval from red to yellow look like the interval from green to blue? Is it smaller or larger? There's not really a perception of exact interval size with colors.

Still, it's hard to think of a better equivalent. Compass directions? But even then, "north" and "south" have a special position on the scale. It's only arbitrary that A=440 is a standard reference pitch (marching bands, for instance, tune to Bb).

Hi George,

> If you don't already have AP, then I don't know whether you can work
on AP at the same time as RP, because developing AP involves
listening for pitch-colors (which are irrelevant to RP). But if you
already have both AP and RP for 12-ET, then I think that microtonal
AP and RP would naturally be developed simultaneously (although I'd
say that microtonal RP comes more easily).

Since both would be a long term process, my plan would be to work a
bit on one then a bit on the other - and have some pieces that I always
play in the same key for now, to practice mental play and AP, and
at the same time while practicing, to play intervals around the circle
of fifths etc in order to strengthen relative pitch.

BTW I'm trying playing back my Teleman pieces using recorders
in parallel in four octaves, and I think it helps.

I'm glad to hear that AP needn't mean that one hears music that's
out of tune as "flat". I think what happened there is probably that
I hadn't got strong enough relative pitch at the time. That also
would be a good reason to develop both together as much as
one can. Maybe that's not a problem if one already has a strong
sense of relative pitch.

Yes, I hear differences in the different types of chords - but the way
in which my relative pitch is weak is that I don't hear the differences
at all easily if it is just an isolated interval. But add a couple of
extra notes and it becomes obvious.

When it is a single isolated diad suddenly out of context
then - well of course I could sing the notes and work out
the interval - or could listen to the partials if the notes
are long sustained notes.

But - if played fairly quickly in an arbitrary key and at any pitch transposition,
and with the notes played one after another with no overlap,
my relative pitch gets completely confused - to the
extent that I can even get confused between a 9/8
and a 3/2. I clearly have relative pitch discrimination
but it only seems to key in strongly with longer phrases
of maybe five or six notes. It is something I can
train in.

While on the other hand notes that follow each other that
differ by a small fraction of a semitone, even with quite a long time
gap between them, are easy to distinguish.
So that's the sense in which my pitch discrimination
seems to be much finer in an AP rather than a
RP sense, even though I don't have AP as ordinarily
understood.

On the 3/1 equivalence - well I'm just not sure,
as someone not trained from an early age, the
2/1 octave similarity doesn't seem to me to be
quite so pre-eminent as all that. It's clearly stronger
than the 3/1 similarity - but not by so much that
it seems to me to be the only possible way
to make equivalence classes of the pitch
"colours".

Yes FTS is all set up to read individual Scala scales, also you can
make a droplist of all the scales in the Scala archive
(or indeed all your scales in any folder).

Yes - will certainly let one select a subset of the current scale.

With the Teleman piece - with the first four notes I have
C, G and A. So one might think that a good starting point might
be to have a quiz on just those notes. So similarly
while developing AP then

I had a look at an on-line page about AP by Diana Deutch
and she wondered if her development of AP may have been
helped by a coloured Xylophone she had as a child.
I wonder if colour synaesthesia may be developed in that
sort of a semi-accidental fashion when very young.

Yes there seem to be many schemes for colour associations
around, - I have several options in FTS which users can
use to colour the note names. A green note could be
anything.

Yes it's an ineresting aspect of colour that you can't
break up an individual colour into its constituent
frequencies in the way you can with sound, e.g.
WHEN MIXING LIGHTS monochromatic yellow
light looks the same to us as a mixture of equal
parts of red and green light, in fact computer
screens use a mixture of red and green
for yellow, and no-one can tell the difference
except that mixed red / green light can never
make quite such a saturated colour
as monochromatic yellow light, always is seen
as having a tiny admixture of white light.

But musicians with AP with synaesthesia for colour
seem to be able to hear the component "colours" of a chord of
many notes. I wonder how it is done? One idea is maybe
they might have a separate spatial "pitch height"
- or keyboard layout - and so see the colours
separated in position as well as colour.

Associating the pitch with position in space does
preserve the relative interval sizes. So a combination
of colour + position would work in that respect,
the colour just as a cue to the pitch like a colour coded
xylophone.

Yes intriguing, never thought of that, "bug fixing"
Actually I do sometimes fix a bug in the sense
that the program does something unexpected
which I find is actually useful, so fix it and make the
bug into a feature.

Robert

Robert walker wrote:

> On the 3/1 equivalence - well I'm just not sure,
> as someone not trained from an early age, the
> 2/1 octave similarity doesn't seem to me to be
> quite so pre-eminent as all that. It's clearly stronger
> than the 3/1 similarity - but not by so much that
> it seems to me to be the only possible way
> to make equivalence classes of the pitch
> "colours". The way I perceive things, the 3/1 similarity is of an entirely different kind from the 2/1 similarity. Notes a 3/1 interval apart share many of the same partials, if they are harmonic sounds, so if the 3rd harmonic is prominent in the lower note, it's possible that some general similarity may be perceived. But the difference between notes an octave apart is almost more a difference of timbre than pitch. Except in the extreme ranges (where pitch sensation starts getting "muddy" in the lower octaves and vague or indistinct in the higher octaves), an A sounds like an A, a C# sounds like a C#, and the notes don't even need to share partials to sound similar in that way. I've just always assumed that everyone perceived sounds this way (even if they couldn't identify a pitch by name).

--- In tuning@yahoogroups.com, Herman Miller <hmiller@...> wrote:

> The way I perceive things, the 3/1 similarity is of an entirely
> different kind from the 2/1 similarity.

Same here, and the 4/1 similarity is of the same kind as the 2/1
similarity.