Hey JdL

Hi John,

I have no idea what list this thread should
go to.

You have had some recent success with tuning other
composers (living composers) works (the Monz and
Christopher Baily).

I haven't heard how you intend to use your tool
yourself or what features you think it should/could
have next.

I haven't used it (and don't know if I ever will)
but it seems as a composition tool, it might be
nice to submit a "map" file in parallel with the
midi file. This would have parameter changes over
time so that a composer could actually influence
which portions of a piece were "5 limit JI let
the melody be damned", which had negative melodic
springs, etc, etc...

I guess my thoughts are to give the program some
of the intent that is missing in the midi file that
a composer might try to get, in some cases via
enharmonic spellings (diminished and augmented
intervals) or by "molto expressivio" *(where I'll
bet most interpreters start to err sharp)...

I dunno, pick a group and tell us whats next for
your work.

Bob Valentine

[Bob Valentine wrote:]
>I have no idea what list this thread should
>go to.

Here or crazy_music; here is fine.

>You have had some recent success with tuning other
>composers (living composers) works (the Monz and
>Christopher Baily).

The reaction has been most gratifying!

>I haven't heard how you intend to use your tool
>yourself or what features you think it should/could
>have next.

I sort of make it up as I go along. But a plan wouldn't be
a bad thing!

>I haven't used it (and don't know if I ever will)
>but it seems as a composition tool, it might be
>nice to submit a "map" file in parallel with the
>midi file. This would have parameter changes over
>time so that a composer could actually influence
>which portions of a piece were "5 limit JI let
>the melody be damned", which had negative melodic
>springs, etc, etc...

Funny you should mention that; the Monz and I have been exchanging
speculations off-list about this exact thing! It would not be
difficult for me to include recognition within the program of special
sysex's (and/or meta-events in a .mid file) which instruct the program,
on the fly, to behave in different ways. Other commands would release
it to make decisions on its own, as it now does, again on the fly.
In the short-term, existing sequencing programs could be used to apply
these sysex's (and/or...); in the long term, a nicer integration with
other programs would be a lot more fun to use.

>I guess my thoughts are to give the program some
>of the intent that is missing in the midi file that
>a composer might try to get, in some cases via
>enharmonic spellings (diminished and augmented
>intervals) or by "molto expressivio" *(where I'll
>bet most interpreters start to err sharp)...

Hmmm, interesting; this goes beyond what I had considered, but all your
suggestions make sense.

>I dunno, pick a group and tell us whats next for
>your work.

The thing I'm right on the verge of doing is allowing more than 12
internal pitch classes (closely connected, of course, with recognition
of enharmonic spellings). But I'd like to introduce a set of commands
which composers could use to communicate to me: picking the tuning file
(5-lim, 7-lim, 11-lim, ...) and the key on the fly is my first priority
there. In fact, that could probably be moved ahead of the pitch class
enhancement.

Before too long, I need to address the question of trying to attach to
other programs. The Monz's own JustMusic, under development, is tops
on my list, except that it's still in a highly unfinished state.
Whether I'd have any leverage with, say, the Cakewalk people is an open
question, but the request might best come from users of the program who
would like to attach to my program (interesting typo: I wrote "attack"
instead of "attach" - shades of the other list??).

If you are interested in plugging in, then I'm interested in knowing
what features are most important to you.

JdL

On Mon, 16 Jul 2001 10:44:47 -0600, "John A. deLaubenfels"
<jdl@adaptune.com> wrote:

>Funny you should mention that; the Monz and I have been exchanging
>speculations off-list about this exact thing! It would not be
>difficult for me to include recognition within the program of special
>sysex's (and/or meta-events in a .mid file) which instruct the program,
>on the fly, to behave in different ways. Other commands would release
>it to make decisions on its own, as it now does, again on the fly.
>In the short-term, existing sequencing programs could be used to apply
>these sysex's (and/or...); in the long term, a nicer integration with
>other programs would be a lot more fun to use.

Another thing that might be helpful would be to use the key signature in
the MIDI file to distinguish between sharps and flats. You could also use
some of the unused MIDI controllers to set options. Either of these would
be a lot easier than adding sysex messages.

> The thing I'm right on the verge of doing is allowing more than 12
> internal pitch classes (closely connected, of course, with recognition
> of enharmonic spellings). But I'd like to introduce a set of commands
> which composers could use to communicate to me: picking the tuning file
> (5-lim, 7-lim, 11-lim, ...) and the key on the fly is my first priority
> there. In fact, that could probably be moved ahead of the pitch class
> enhancement.
>

I think the thing to come up with is how you want to communicate. I
suggested a text file that could be synced internally with the midi
file. You just need a way of specifying the parameter and specifying
time. As has been pointed out, key signatures (to a certain extent)
are communicable through midi. You've suggested using sysex messages
which your code would snarf. The problem I see with that is a friendly
way to insert the sysex messages.

> Before too long, I need to address the question of trying to attach to
> other programs. The Monz's own JustMusic, under development, is tops
> on my list, except that it's still in a highly unfinished state.
> Whether I'd have any leverage with, say, the Cakewalk people is an open
> question, but the request might best come from users of the program who
> would like to attach to my program (interesting typo: I wrote "attack"
> instead of "attach" - shades of the other list??).
>

Power Tracks Pro, a sequencer from the same people who do Band in a Box
is not only dirt cheap but it has the capability to pass its data
structures to external programs (either a dll or an exe). I did some
experiments with this a few years ago and it worked fine.

> If you are interested in plugging in, then I'm interested in knowing
> what features are most important to you.
>

I'm really not sure. Maybe I'll have a better idea if I get some 12tet
retuned.

Bob Valentine

> JdL

(Bob, my reply to you is below...)

[Herman Miller wrote:]
>On Mon, 16 Jul 2001 10:44:47 -0600, "John A. deLaubenfels"
><jdl@a...> wrote:

[JdL:]
>>Funny you should mention that; the Monz and I have been exchanging
>>speculations off-list about this exact thing! It would not be
>>difficult for me to include recognition within the program of special
>>sysex's (and/or meta-events in a .mid file) which instruct the
>>program, on the fly, to behave in different ways. Other commands
>>would release it to make decisions on its own, as it now does, again
>>on the fly. In the short-term, existing sequencing programs could be
>>used to apply these sysex's (and/or...); in the long term, a nicer
>>integration with other programs would be a lot more fun to use.

[Herman:]
>Another thing that might be helpful would be to use the key signature
>in the MIDI file to distinguish between sharps and flats. You could
>also use some of the unused MIDI controllers to set options. Either of
>these would be a lot easier than adding sysex messages.

Funny, I'd almost forgotten that the midi file has a place where the key
signature can be specified. Lessee, is that once at the beginning of
the file, or any time via meta-events... [cue sound of books being
opened]... AHA, it's a meta-event that can be inserted at any time! And
for pieces with more than 12 pitch classes, changing the key signature
could clue the program which enharmonic was wanted. Good idea!

Unused controllers, huh... Are you referring to the RPN and NRPN guys,
or to plain ol' Bx xx xx messages? Are holes in the latter guaranteed
to remain throughout time, I wonder? Are they often overloaded by
others?

I don't use commercial sequencing programs, so it's good to know what's
easy and what's not. Is it _quite_ hard to insert a sysex? Or
meta-event?

[Bob Valentine responded:]

[JdL:]
>>The thing I'm right on the verge of doing is allowing more than 12
>>internal pitch classes (closely connected, of course, with recognition
>>of enharmonic spellings). But I'd like to introduce a set of commands
>>which composers could use to communicate to me: picking the tuning
>>file (5-lim, 7-lim, 11-lim, ...) and the key on the fly is my first
>>priority there. In fact, that could probably be moved ahead of the
>>pitch class enhancement.

>I think the thing to come up with is how you want to communicate. I
>suggested a text file that could be synced internally with the midi
>file. You just need a way of specifying the parameter and specifying
>time. As has been pointed out, key signatures (to a certain extent)
>are communicable through midi. You've suggested using sysex messages
>which your code would snarf. The problem I see with that is a friendly
>way to insert the sysex messages.

Parallel text file - wow, my first reaction is that getting it in sync
and keeping it that way might be difficult. But then again, sequencing
programs work with measure counts, right? So maybe that'd work well.

You both hint that it's hard to insert sysex's via a sequencing program.
That's too bad; don't really understand why, unless it's rarely done by
most sequencers.

[JdL:]
>>Before too long, I need to address the question of trying to attach to
>>other programs. The Monz's own JustMusic, under development, is tops
>>on my list, except that it's still in a highly unfinished state.
>>Whether I'd have any leverage with, say, the Cakewalk people is an
>>open question, but the request might best come from users of the
>>program who would like to attach to my program.

[Bob:]
>Power Tracks Pro, a sequencer from the same people who do Band in a Box
>is not only dirt cheap but it has the capability to pass its data
>structures to external programs (either a dll or an exe). I did some
>experiments with this a few years ago and it worked fine.

Kyool!! Thanks for the tip! Might go check'm out. I actually really
_need_ a real sequencing program; my tools are primitive in the extreme
(you might call them assembly language sequencing).

>>If you are interested in plugging in, then I'm interested in knowing
>>what features are most important to you.

>I'm really not sure. Maybe I'll have a better idea if I get some 12tet
>retuned.

Meaning some of your own works, I hope! Send anything across to me that
you'd like to hear, to jdl"at"adaptune.com .

Thanks, Herman and Bob, for all your suggestions!

JdL

On Tue, 17 Jul 2001 10:41:43 -0600, "John A. deLaubenfels"
<jdl@adaptune.com> wrote:

>I don't use commercial sequencing programs, so it's good to know what's
>easy and what's not. Is it _quite_ hard to insert a sysex? Or
>meta-event?

Cakewalk Home Studio 3.01, which is what I use (and by now it must be
*really* obsolete) can insert sysexes from a file, but can't record them.
It'd probably be easy enough if there are only a small number of
pre-existing sysex messages in individual files that need to be inserted.

I probably ought to look for something better, but I haven't had much need
for sysex beyond what Scala provides.