back to list

More info on Wolfram Tones project

🔗Jon Szanto <jszanto@...>

9/12/2005 2:34:44 PM

I received a nice response back from one of the members of their team,
and would be pleased to see if anyone has suggestions, either to me,
or you may write them (I'll include email at bottom). I've let them
know about the main tuning list, and if it doesn't feel too OT, I can
post this there for a larger readership.

Wolfram Tones wrote:

"Dear Jon Szanto,

Thank you for your enquiry about WolframTones. We are getting a lot
of correspondence, but we hope that the following remarks will serve
as a useful response to your enquiry.

At present we are transmitting all data in MIDI, and then relying on
browsers to render the MIDI. This means that (so far as we can tell)
we can't reliably handle anything other than equal temperament tuning.
In the future, we're hoping to make the site have the option to
render WolframTones compositions into formats like MP3, and in this
case we would hope to be able to support other tuning systems.

The issues with MIDI also prevent us at present from handling anything
other the standard 12 tones per octave system.

However, within this system, we can handle fairly general scales, and
we hope very soon to make a small additional generalization. We label
scales by an integer code whose binary digits correspond to the
semitones that are selected for the scale (see
http://tones.wolfram.com/about/controls-side.html#musicalscale).
Scales between 2048 and 4095 repeat each octave in the standard way.
As of today, these are the only scales supported on the site.

However, the WolframTones server can in fact handle other scales, and
in response to significant interest, we hope in the very near future
to add the user interface on the site. We would be interested to know
what non-octaval scales would be of particular interest, and whether
there are examples that should be listed in our list of named scales.

Note that some scales (e.g. the various Messiaen modes) that repeat in
less than an octave are already available with the current interface.
(See e.g. scale 3822; note that you can type in a scale number
directly where the number is displayed in the Pitch Mapping panel.)

You may be interested to know that extensive investigation of scales
is now possible in Mathematica. Mathematica has supported
waveform-based sound output since 1991. We have now developed
extensive capabilities for MIDI output (as well as MP3 etc.) We will
soon be making these capabilities available to a limited group of
music-oriented test users. Please do let us know if you might be
interested in this program. We believe Mathematica is an outstanding
platform for music development, and are keen to do what we can to let
it achieve its potential in this area.

Thank you for again your interest in WolframTones. We are happy to
receive constructive input."

Jon again: if any of you want to pursue this further, you may write
them directly: tones-info 'at' wolfram 'dot' com

Cheers,
Jon

🔗Carl Lumma <clumma@...>

9/12/2005 10:06:19 PM

Fantastic, Jon!!!

I'd reply that:

1. When they get mp3 functionality going, they should get
a GM-capable softsynth that supports full-scale tunings
(Kontakt2 ?) and ask us to submit some cool tunings. This
is the ideal way to tune a synth, and it would let them
break the 12 tones/octave barrier (though clearly their
algorithm works best with fewer tones/octave than that).
I think harmonic series segments would be a great first
choice. Additionally, some of the scales in Paul's latest
paper.

2. In the meantime, they could implement microtuning for
their monotimbral mode with standard MIDI using 12 channels
and pitch bend, though building a feature that applies only
to that mode might pain them.

-Carl

🔗Jon Szanto <jszanto@...>

9/13/2005 1:13:29 PM

--- In metatuning@yahoogroups.com, "Carl Lumma" <clumma@y...> wrote:
> Fantastic, Jon!!!
>
> I'd reply that:

Some good ideas, Carl, that I might possibly ammend or clarify as
well. I'm not inclined, however, to pursue this any further if there
aren't more than a couple people (like you and me) interested - I
don't really have the energy or see the need to campaign for changes
unless there is a desire on more than an individual level. At that
point I'd just encourage individuals to seek their own ways to do this.

Anyone else see a need here?

Cheers,
Jon

🔗Carl Lumma <clumma@...>

9/13/2005 2:56:51 PM

Kurt's onboard.

-Carl

> > Fantastic, Jon!!!
> >
> > I'd reply that:
>
> Some good ideas, Carl, that I might possibly ammend or clarify as
> well. I'm not inclined, however, to pursue this any further if there
> aren't more than a couple people (like you and me) interested - I
> don't really have the energy or see the need to campaign for changes
> unless there is a desire on more than an individual level. At that
> point I'd just encourage individuals to seek their own ways to do
> this.
>
> Anyone else see a need here?
>
> Cheers,
> Jon

🔗Kraig Grady <kraiggrady@...>

9/13/2005 4:12:30 PM

hell if they don't use microtones , if just leaves more room for us! :)
Anyway microtones are "sacred" who wants them used for such secular useages! :)
they'll catch up, but it is nice to have a head start!

Jon Szanto wrote:

>--- In metatuning@yahoogroups.com, "Carl Lumma" <clumma@y...> wrote:
> >
>>Fantastic, Jon!!!
>>
>>I'd reply that:
>> >>
>
>Some good ideas, Carl, that I might possibly ammend or clarify as
>well. I'm not inclined, however, to pursue this any further if there
>aren't more than a couple people (like you and me) interested - I
>don't really have the energy or see the need to campaign for changes
>unless there is a desire on more than an individual level. At that
>point I'd just encourage individuals to seek their own ways to do this.
>
>Anyone else see a need here?
>
>Cheers,
>Jon
>
>
>
>
>
>
>Meta Tuning meta-info:
>
>To unsubscribe, send an email to:
>metatuning-unsubscribe@yahoogroups.com
>
>Web page is http://groups.yahoo.com/groups/metatuning/
>
>To post to the list, send to
>metatuning@yahoogroups.com
>
>You don't have to be a member to post.
>
> >Yahoo! Groups Links
>
>
>
> >
>
>
>
> >

--
Kraig Grady
North American Embassy of Anaphoria Island <http://anaphoria.com/>
The Wandering Medicine Show
KXLU <http://www.kxlu.com/main.html> 88.9 FM Wed 8-9 pm Los Angeles

🔗Carl Lumma <clumma@...>

9/14/2005 4:55:14 PM

> Some good ideas, Carl, that I might possibly ammend or clarify
> as well. I'm not inclined, however, to pursue this any further
> if there aren't more than a couple people (like you and me)
> interested - I don't really have the energy or see the need to
> campaign for changes unless there is a desire on more than an
> individual level. At that point I'd just encourage individuals
> to seek their own ways to do this.

Since you didn't ammend or clarify, what they got was what I was
able to write. Which was:

Hiya,

Your project has been the subject of some discussion on a music
theory mailing list I'm on -- a reply from your team was forwarded
to the list re. microtuning support for Tones.

It stated that since you are using General MIDI, you cannot
deliver microtonal music. That's not entirely true. General
MIDI supports pitch bend, which is (in most implementations) a
very accurate way to tune pitches. Unfortunately, it is a
channel-specific message.

But with monotimbral music, up to 15 tones/octave can be delivered.
Each note class is associated with a MIDI channel. The appropriate
offsets are sent to each channel before the piece begins. Any
expressive pitch bends in the piece can ruin things, but it doesn't
look like Tones scores generate expressive bends yet.

With scores generated in advance (as with Tones), a slightly
trickier approach can deliver multitimbral music with unlimited
tones/octave, with up to 15 monophonic instruments. In this scheme
channels are associated with instruments, and tuning bends are sent
immediately before each note-on. If I am correct, though, there
are some polyphonic instruments in Tones. A piano playing a triad
would use 3 of 15 channels in this scheme.

In fact, pitch bend *can* deliever unlimited tones/octave and
polyphony, through clever channel-hocketing techniques, such as
those used by John deLaubenfels in his "adaptune" project.

However, none of these approaches are completely satisfactory.

Fortunately, there is the MIDI Tuning Standard. Unfortunately,
not many synthesizers support it. However, your reply indicated
that mp3 export functionality is planned for Tones. In this case,
something like Native Instruments Kontakt2 could be used to
deliver music in any tuning at all, without any pitch bend hiccups
or limitations.

As for which scales to use, I think harmonic series segments of
between 5 and 12 notes would sound wonderful with Tones. And on
the cutting edge, subsets of linear temperaments discovered in
the last decade provide pleasant but unusually-sounding
structures analogous to the diatonic scale in meantone
temperaments. See...

http://lumma.org/tuning/erlich

...or contact me for more information.

-Carl

🔗Carl Lumma <clumma@...>

9/14/2005 4:57:12 PM

> hell if they don't use microtones , if just leaves more room
> for us! :)
> Anyway microtones are "sacred" who wants them used for such
> secular useages! :)
> they'll catch up, but it is nice to have a head start!

We have a huge headstart, and... we can use all the help
we can get. If this thing had microtonal scales and became
popular, cell phones would be ringing all over the world
in JI or (insert favorite scale here). That's worth writing
a few e-mails.

-Carl

🔗Kraig Grady <kraiggrady@...>

9/14/2005 5:36:42 PM

why only 15?

Carl Lumma wrote:

>
>
>But with monotimbral music, up to 15 tones/octave can be delivered.
>Each note class is associated with a MIDI channel. The appropriate
>offsets are sent to each channel before the piece begins. Any
>expressive pitch bends in the piece can ruin things, but it doesn't
>look like Tones scores generate expressive bends yet.
>
>
>
>
> >

--
Kraig Grady
North American Embassy of Anaphoria Island <http://anaphoria.com/>
The Wandering Medicine Show
KXLU <http://www.kxlu.com/main.html> 88.9 FM Wed 8-9 pm Los Angeles

🔗Kraig Grady <kraiggrady@...>

9/14/2005 5:39:22 PM

I think we already are surrounded by microtones, but perhaps you are right.
Anywise any type of subversions is good, and if plato is right, a change in music will result in a change in society.

Carl Lumma wrote:

>>hell if they don't use microtones , if just leaves more room
>>for us! :)
>>Anyway microtones are "sacred" who wants them used for such
>>secular useages! :)
>> they'll catch up, but it is nice to have a head start!
>> >>
>
>We have a huge headstart, and... we can use all the help
>we can get. If this thing had microtonal scales and became
>popular, cell phones would be ringing all over the world
>in JI or (insert favorite scale here). That's worth writing
>a few e-mails.
>
>-Carl
>
>
>
>
>
>Meta Tuning meta-info:
>
>To unsubscribe, send an email to:
>metatuning-unsubscribe@yahoogroups.com
>
>Web page is http://groups.yahoo.com/groups/metatuning/
>
>To post to the list, send to
>metatuning@yahoogroups.com
>
>You don't have to be a member to post.
>
> >Yahoo! Groups Links
>
>
>
> >
>
>
>
> >

--
Kraig Grady
North American Embassy of Anaphoria Island <http://anaphoria.com/>
The Wandering Medicine Show
KXLU <http://www.kxlu.com/main.html> 88.9 FM Wed 8-9 pm Los Angeles

🔗Carl Lumma <clumma@...>

9/14/2005 6:35:57 PM

> I think we already are surrounded by microtones, but perhaps
> you are right.

There are microtones and then there are microtones.

> Anywise any type of subversions is good, and if plato is right,
> a change in music will result in a change in society.

I believe it.

-Carl

🔗Carl Lumma <clumma@...>

9/14/2005 6:35:13 PM

16 MIDI channels, 1 for drums. -C.

> why only 15?
>
> Carl Lumma wrote:
>
> >
> >
> >But with monotimbral music, up to 15 tones/octave can be delivered.
> >Each note class is associated with a MIDI channel. The appropriate
> >offsets are sent to each channel before the piece begins. Any
> >expressive pitch bends in the piece can ruin things, but it doesn't
> >look like Tones scores generate expressive bends yet.
>
>
> --
> Kraig Grady
> North American Embassy of Anaphoria Island <http://anaphoria.com/>
> The Wandering Medicine Show
> KXLU <http://www.kxlu.com/main.html> 88.9 FM Wed 8-9 pm Los Angeles

🔗Jon Szanto <jszanto@...>

9/14/2005 7:23:36 PM

--- In metatuning@yahoogroups.com, "Carl Lumma" <clumma@y...> wrote:
> Since you didn't ammend or clarify, what they got was what I was
> able to write. Which was:

Thanks, I had a bunch of things on my plate that pushed aside the
hobbies. What you wrote looks pretty good, but I have a couple questions:

1. The mp3 thing - just who implements the render instrument (you've
mentioned Kontakt 2): the end user or them and they just shoot out an
mp3 file?

2. Why a sampler? That would mean buying a sample library as well.

But in general I think it is good that you gave them a substantial
piece of info to chew on. We'll see what happens, and I sure wish they
could find a way to code it as an applet, maybe a Java thing you could
download or something.

Cheers,
Jon

🔗Carl Lumma <clumma@...>

9/14/2005 10:45:18 PM

> 1. The mp3 thing - just who implements the render instrument (you've
> mentioned Kontakt 2): the end user or them and they just shoot out
> an mp3 file?

Them.

> 2. Why a sampler? That would mean buying a sample library as well.

Because they're using a wide variety of instruments, and GM
in particular. Name one other GM synth that can handle full-
scale retunings (besides Timidity).

> But in general I think it is good that you gave them a
> substantial piece of info to chew on. We'll see what happens,
> and I sure wish they could find a way to code it as an applet,
> maybe a Java thing you could download or something.

Why would you want to download it?

-Carl

🔗Kraig Grady <kraiggrady@...>

9/14/2005 10:59:10 PM

oh i see , you do a whole pitch bend for the whole channel, is that right?
i have never had to deal with this, so pardon my Pehrsonesque stance

Carl Lumma wrote:

>16 MIDI channels, 1 for drums. -C.
>
> >
>>why only 15?
>>
>>Carl Lumma wrote:
>>
>> >>
>>>But with monotimbral music, up to 15 tones/octave can be delivered.
>>>Each note class is associated with a MIDI channel. The appropriate
>>>offsets are sent to each channel before the piece begins. Any
>>>expressive pitch bends in the piece can ruin things, but it doesn't
>>>look like Tones scores generate expressive bends yet.
>>> >>>
>>-- >>Kraig Grady
>>North American Embassy of Anaphoria Island <http://anaphoria.com/>
>>The Wandering Medicine Show
>>KXLU <http://www.kxlu.com/main.html> 88.9 FM Wed 8-9 pm Los Angeles
>> >>
>
>
>
>
>
>Meta Tuning meta-info:
>
>To unsubscribe, send an email to:
>metatuning-unsubscribe@yahoogroups.com
>
>Web page is http://groups.yahoo.com/groups/metatuning/
>
>To post to the list, send to
>metatuning@yahoogroups.com
>
>You don't have to be a member to post.
>
> >Yahoo! Groups Links
>
>
>
> >
>
>
> >

--
Kraig Grady
North American Embassy of Anaphoria Island <http://anaphoria.com/>
The Wandering Medicine Show
KXLU <http://www.kxlu.com/main.html> 88.9 FM Wed 8-9 pm Los Angeles

🔗Carl Lumma <clumma@...>

9/14/2005 11:04:30 PM

Yup.

-Carl

> oh i see , you do a whole pitch bend for the whole channel, is
> that right?
>
> >16 MIDI channels, 1 for drums. -C.
> >
> >
> >
> >>why only 15?
> >>
> >>>But with monotimbral music, up to 15 tones/octave can be
> >>>delivered. Each note class is associated with a MIDI channel.
> >>>The appropriate offsets are sent to each channel before the
> >>>piece begins. Any expressive pitch bends in the piece can
> >>>ruin things, but it doesn't look like Tones scores generate
> >>>expressive bends yet.

🔗Kraig Grady <kraiggrady@...>

9/14/2005 11:12:40 PM

what a pain that you can't get two pitches from the same channel except for flukes like 22 ET for example which gives you the tritone

Carl Lumma wrote:

>Yup.
>
>-Carl
>
> >
>>oh i see , you do a whole pitch bend for the whole channel, is
>>that right?
>>
>> >>
>>>16 MIDI channels, 1 for drums. -C.
>>>
>>> >>>
>>> >>>
>>>>why only 15?
>>>>
>>>> >>>>
>>>>>But with monotimbral music, up to 15 tones/octave can be
>>>>>delivered. Each note class is associated with a MIDI channel.
>>>>>The appropriate offsets are sent to each channel before the
>>>>>piece begins. Any expressive pitch bends in the piece can
>>>>>ruin things, but it doesn't look like Tones scores generate
>>>>>expressive bends yet.
>>>>> >>>>>
>
>
>
>
>
>
>Meta Tuning meta-info:
>
>To unsubscribe, send an email to:
>metatuning-unsubscribe@yahoogroups.com
>
>Web page is http://groups.yahoo.com/groups/metatuning/
>
>To post to the list, send to
>metatuning@yahoogroups.com
>
>You don't have to be a member to post.
>
> >Yahoo! Groups Links
>
>
>
> >
>
>
>
> >

--
Kraig Grady
North American Embassy of Anaphoria Island <http://anaphoria.com/>
The Wandering Medicine Show
KXLU <http://www.kxlu.com/main.html> 88.9 FM Wed 8-9 pm Los Angeles

🔗Jon Szanto <jszanto@...>

9/14/2005 11:26:51 PM

--- In metatuning@yahoogroups.com, "Carl Lumma" <clumma@y...> wrote:
> > 1. The mp3 thing - just who implements the render instrument (you've
> > mentioned Kontakt 2): the end user or them and they just shoot out
> > an mp3 file?
>
> Them.

Wow. Weird.

> > 2. Why a sampler? That would mean buying a sample library as well.
>
> Because they're using a wide variety of instruments, and GM
> in particular. Name one other GM synth that can handle full-
> scale retunings (besides Timidity).

a. It's a sampler, not a synth. We should at least be clear in our
terminology
b. Typical for their site, I can't find a blinking thing about
microtuning in detail, not least to mention "full-scale retunings". I
take your word for it, but I also guess this is all through their
scripting, right?
c. Kontakt didn't used to come with content, IIRC (but could be
wrong). Already mapped for GM use? (I'm asking some of these questions
for my own general knowledge, as you seem to know the product pretty
deeply).

> > and I sure wish they could find a way to code it as an applet,
> > maybe a Java thing you could download or something.
>
> Why would you want to download it?

So that I could use it with the sounds of *my* choice! And what sweet
people they would be if they would release the code in open source
(yeah, right - check out list prices for Mathematica. These folks like
commerce!).

Besides, my DAW is not connected to the outside world, and it likes it
that way.

Cheers,
Jon

🔗Jon Szanto <jszanto@...>

9/14/2005 11:28:05 PM

--- In metatuning@yahoogroups.com, Kraig Grady <kraiggrady@a...> wrote:
> what a pain that you can't get two pitches from the same channel except
> for flukes like 22 ET for example which gives you the tritone

I'll supply the "yup" this time. Pitch-bend microtonality sucks
gigantic eggs.

Cheers,
Jon

🔗Carl Lumma <clumma@...>

9/15/2005 12:05:31 AM

Not sure what you're referring to here... in any case, this is
just the most basic method of tuning via pitch bend... the rest
of my letter had more flexible methods.

-Carl

> what a pain that you can't get two pitches from the same channel
> except for flukes like 22 ET for example which gives you the
> tritone
>
> Carl Lumma wrote:
>
> >Yup.
> >
> >-Carl
> >
> >
> >
> >>oh i see , you do a whole pitch bend for the whole channel, is
> >>that right?
> >>
> >>
> >>
> >>>16 MIDI channels, 1 for drums. -C.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>>why only 15?
> >>>>
> >>>>
> >>>>
> >>>>>But with monotimbral music, up to 15 tones/octave can be
> >>>>>delivered. Each note class is associated with a MIDI channel.
> >>>>>The appropriate offsets are sent to each channel before the
> >>>>>piece begins. Any expressive pitch bends in the piece can
> >>>>>ruin things, but it doesn't look like Tones scores generate
> >>>>>expressive bends yet.
> >>>>>
> >>>>>
> >
> >
> >
> >
> >
> >
> >Meta Tuning meta-info:
> >
> >To unsubscribe, send an email to:
> >metatuning-unsubscribe@yahoogroups.com
> >
> >Web page is http://groups.yahoo.com/groups/metatuning/
> >
> >To post to the list, send to
> >metatuning@yahoogroups.com
> >
> >You don't have to be a member to post.
> >
> >
> >Yahoo! Groups Links
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
>
> --
> Kraig Grady
> North American Embassy of Anaphoria Island <http://anaphoria.com/>
> The Wandering Medicine Show
> KXLU <http://www.kxlu.com/main.html> 88.9 FM Wed 8-9 pm Los Angeles

🔗Carl Lumma <clumma@...>

9/15/2005 12:12:23 AM

> > Because they're using a wide variety of instruments, and GM
> > in particular. Name one other GM synth that can handle full-
> > scale retunings (besides Timidity).
>
> a. It's a sampler, not a synth. We should at least be clear in
> our terminology

There's really no distinction to be made here.

> b. Typical for their site, I can't find a blinking thing about
> microtuning in detail, not least to mention "full-scale
> retunings". I take your word for it, but I also guess this
> is all through their scripting, right?

I don't actually know. I have heard from that all of their
flagship instruments have always been full-scale tunable.
However, since I have it here, maybe I can check and see.

> c. Kontakt didn't used to come with content, IIRC (but could
> be wrong).

My impression from their website is that it comes with content.

> Already mapped for GM use?

I don't know about that, but I'm sure GM banks are available.

> (I'm asking some of these questions for my own general
> knowledge, as you seem to know the product pretty deeply).

Nothing could be further from the truth.

> > > and I sure wish they could find a way to code it as an
> > > applet, maybe a Java thing you could download or something.
> >
> > Why would you want to download it?
>
> So that I could use it with the sounds of *my* choice!

You can download the MIDI file.

> And what sweet people they would be if they would release
> the code in open source (yeah, right - check out list prices
> for Mathematica. These folks like commerce!).

All these scientific packages are priced this way... including
very affordable student pricing. Of course, they are changing
to download the ringtones.

-Carl

🔗Jon Szanto <jszanto@...>

9/15/2005 8:36:46 AM

--- In metatuning@yahoogroups.com, "Carl Lumma" <clumma@y...> wrote:
> There's really no distinction to be made here.

Ahem: sloppy.

> I don't actually know. I have heard from that all of their
> flagship instruments have always been full-scale tunable.
> However, since I have it here, maybe I can check and see.

Ah, yes, _please_ do. I have been watching their products for just
such a feature, and neither Kontakt 1, nor Kompakt (which I believe is
was the first of their sampler line to include some form of
microtuning). Be aware that NI's line is wildly inconsistent, with
some of the instrument having been developed outside and then bought
by the company. Even when instruments use altered tuning they
frequently do it in different ways, and some of the instruments don't
support it at all.

> My impression from their website is that it comes with content.

Yes, looks like plenty for the current needs, I sit corrected. I
downloaded a demo last night and will try to see whatsup (with tuning,
etc).

> I don't know about that, but I'm sure GM banks are available.

If its not a pain to map them. But with guidance it should be possible.

> You can download the MIDI file.

I did not get that - it looked to me that all you could do was
download the ringtone file, directly to your phone, for a charge of
$2. If you know of a way to download a MIDI file, please post how.

> All these scientific packages are priced this way... including
> very affordable student pricing. Of course, they are changing
> to download the ringtones.

I'm not a student (unless you count "student of life"). And yes, it
seems a bit greedy to make money off a little app like this, but
that's capitalism for you.

Cheers,
Jon

🔗Carl Lumma <clumma@...>

9/15/2005 10:10:42 AM

> > There's really no distinction to be made here.
>
> Ahem: sloppy.

There is no difference between a "synthesizer" and a
"sampler" as the terms are commonly used. There's no
technical difference in what they do or in what people
do with them. One might say "sampler" if it is
actually used to record samples -- Kontakt isn't used
that way. It does have time stretching and pitch
shifting of samples using granular synthesis, it has
filters and LFOs. Meanwhile "synths" like your favorite
z3ta+ have wavetable "oscillators".

> > My impression from their website is that it comes with content.
>
> Yes, looks like plenty for the current needs, I sit corrected. I
> downloaded a demo last night and will try to see whatsup (with
> tuning, etc).

Please let us know how you make out!

> > You can download the MIDI file.
>
> I did not get that - it looked to me that all you could do was
> download the ringtone file, directly to your phone, for a charge
> of $2. If you know of a way to download a MIDI file, please post
> how.

I believe the ringtones are GM MIDI files.

> > All these scientific packages are priced this way... including
> > very affordable student pricing. Of course, they are changing
> > to download the ringtones.
>
> I'm not a student (unless you count "student of life").

People like you and I fall between the cracks... there's very
few mathematical hobbyists out there. Really, they want to
sock big corporations for $$$, since they use the software to
solve problems that save them $$$$.

> And yes, it seems a bit greedy to make money off a little app
> like this, but that's capitalism for you.

Well I'm sure it wasn't cheap to build, and the server isn't
free either. Meanwhile, ringtones are outselling bands like
Coldplay. $2 is a little high, but probably dictated by the
billing system.

-Carl

🔗Jon Szanto <jszanto@...>

9/15/2005 12:11:05 PM

--- In metatuning@yahoogroups.com, "Carl Lumma" <clumma@y...> wrote:
> There is no difference between a "synthesizer" and a
> "sampler" as the terms are commonly used. There's no
> technical difference in what they do or in what people
> do with them.

OK, I won't beat this, but I disagree. If you were to load up Kontakt
with no samples whatsoever (because none are directly built into the
thing) would it make any sounds? Could you use it to create sounds
from scratch, i.e. synthesize them? I just find it odd that you crave
precision in terms, both linguistic and numeric, in your tuning
discussions but that the fundamental differences here don't bother you.

But not a big issue.

> Please let us know how you make out!

I will. I remember when I demo'd Kompakt (cute spelling) just to try
the microtuning scripting it was one of the features NOT included in
the demo...!

> I believe the ringtones are GM MIDI files.

Yeah, but what do you do if all they do is download to a phone? And
since I don't have a phone that will use downloadable ringtones, I
can't tell. (I'm assuming it downloads to a phone - why else would
they ask for the phone #?)

> Well I'm sure it wasn't cheap to build, and the server isn't
> free either.

This one doesn't seem to be a big deal to me. I've seen Java applets
at least this complex in JMSL. Server costs? Well, maybe, but I doubt
it's a big hit.

> Meanwhile, ringtones are outselling bands like
> Coldplay. $2 is a little high, but probably dictated by the
> billing system.

No idea. I just hate most cellphone ringage. I couldn't believe it:
the other night we went with friends to a performance (stage) at the
Civic Theatre here, and literally as soon as the crowd got on their
feet I could count almost 1 out of every 2 people around me, flipping
on their phone and starting to gab. Its pathological at this point.

Cheers,
Jon

🔗Carl Lumma <clumma@...>

9/15/2005 1:45:11 PM

> > Please let us know how you make out!
>
> I will. I remember when I demo'd Kompakt (cute spelling) just to
> try the microtuning scripting it was one of the features NOT
> included in the demo...!

D'oh!

> > I believe the ringtones are GM MIDI files.
>
> Yeah, but what do you do if all they do is download to a phone? And
> since I don't have a phone that will use downloadable ringtones, I
> can't tell. (I'm assuming it downloads to a phone - why else would
> they ask for the phone #?)

It does. Ok, so that's a feature request: let me download the
MIDI file.

> > Well I'm sure it wasn't cheap to build, and the server isn't
> > free either.
>
> This one doesn't seem to be a big deal to me.

There's a fair bit of smarts in there as far as actually
generating the scores, auto-selecting the instrumentation.
It's a very slick bit of algocomp.

> I've seen Java applets

This isn't an applet; it's a web application. Big difference.

> at least this complex in JMSL.

What do you mean by complex? Figuring out how to download
ringtones to phones... make MIDI files compatible with phones...
and most of all, a rich UI with lots of interactivity.

> > Meanwhile, ringtones are outselling bands like
> > Coldplay. $2 is a little high, but probably dictated by the
> > billing system.
>
> No idea. I just hate most cellphone ringage. I couldn't believe
> it: the other night we went with friends to a performance (stage)
> at the Civic Theatre here, and literally as soon as the crowd
> got on their feet I could count almost 1 out of every 2 people
> around me, flipping on their phone and starting to gab. Its
> pathological at this point.

Yup. And 10X worse in Asia.

-Carl

🔗Jon Szanto <jszanto@...>

9/15/2005 2:52:59 PM

C,

--- In metatuning@yahoogroups.com, "Carl Lumma" <clumma@y...> wrote:
> It does. Ok, so that's a feature request: let me download the
> MIDI file.

Good idea.

> There's a fair bit of smarts in there as far as actually
> generating the scores, auto-selecting the instrumentation.
> It's a very slick bit of algocomp.

Well... I'm not sure about the exact algo end of things, but there has
been a lot of codebase like that built up over the years, and the
auto-selecting, etc, could be easily done in look-up tables. I mean,
I've coded stuff like this before (non-music, but similar generative
stuff).

> What do you mean by complex? Figuring out how to download
> ringtones to phones... make MIDI files compatible with phones...
> and most of all, a rich UI with lots of interactivity.

The ringtones thing is a very specific part of it, to be sure. But if
all we were talking about was the midi stuff, and the UI, I still
don't think it is a big deal (in the sense that it could also live in
an applet). Don't get me wrong: I happen to think it is very well done
in this instance. Almost looks like some of Edward Tufts diagrams come
to life!

> Yup. And 10X worse in Asia.

Ow.

🔗Carl Lumma <clumma@...>

9/15/2005 3:41:46 PM

> > It does. Ok, so that's a feature request: let me download
> > the MIDI file.
>
> Good idea.

You wanna write them or should I? I'm happy to pay the $2 fee.

> > There's a fair bit of smarts in there as far as actually
> > generating the scores, auto-selecting the instrumentation.
> > It's a very slick bit of algocomp.
>
> Well... I'm not sure about the exact algo end of things, but
> there has been a lot of codebase like that built up over the
> years, and the auto-selecting, etc, could be easily done in
> look-up tables. I mean, I've coded stuff like this before
> (non-music, but similar generative stuff).

Keep in mind, Jon, that the thing is generalized for a huge
variety of 1-D CA, and can instantly cough up music that sounds
roughly like it's in one of a dozen or more different styles.
Not a coffee-break project, dude.

> > What do you mean by complex? Figuring out how to download
> > ringtones to phones... make MIDI files compatible with phones...
> > and most of all, a rich UI with lots of interactivity.
>
> The ringtones thing is a very specific part of it, to be sure.
> But if all we were talking about was the midi stuff, and the UI,
> I still don't think it is a big deal (in the sense that it could
> also live in an applet).

I could live in an applet -- that would be easy. The point is,
it's all running server-side, and delivered with straight html
and javascript.

-Carl

🔗Jon Szanto <jszanto@...>

9/15/2005 4:05:35 PM

--- In metatuning@yahoogroups.com, "Carl Lumma" <clumma@y...> wrote:
> You wanna write them or should I? I'm happy to pay the $2 fee.

I'm not going to get around to it real soon, so you can suggest. $2
fee? You mean suggesting that a fee-based system, either downloaded
direct to phone, or download a midi file to computer? (Just making
sure I'm understanding you).

> Keep in mind, Jon, that the thing is generalized for a huge
> variety of 1-D CA, and can instantly cough up music that sounds
> roughly like it's in one of a dozen or more different styles.
> Not a coffee-break project, dude.

Well, maybe. CA coding isn't that difficult - I did my first on paper
in high school for a biology project (inspired by Martin Gardner). But
I'm sure it is a nicely sophisticated system, and even more sure that
it is _elegant_. But you know about the style stuff? I honestly don't
think they sound all that different, with the exception of patch choices!

> I could live in an applet -- that would be easy. The point is,
> it's all running server-side, and delivered with straight html
> and javascript.

Oh, I suppose. But sometimes there isn't a whole lot of difference
between the two, just where the computing actually takes place. I
haven't ever gotten all that excited as to whether things were client-
or server-side. When I'm not a programmer, I'm a user.

Cheers,
Jon

🔗Carl Lumma <clumma@...>

9/15/2005 4:37:05 PM

> > You wanna write them or should I? I'm happy to pay the $2 fee.
>
> I'm not going to get around to it real soon, so you can suggest. $2
> fee? You mean suggesting that a fee-based system, either downloaded
> direct to phone, or download a midi file to computer? (Just making
> sure I'm understanding you).

Yes.

> > Keep in mind, Jon, that the thing is generalized for a huge
> > variety of 1-D CA, and can instantly cough up music that sounds
> > roughly like it's in one of a dozen or more different styles.
> > Not a coffee-break project, dude.
>
> Well, maybe. CA coding isn't that difficult - I did my first on
> paper in high school for a biology project (inspired by Martin
> Gardner).

Running a CA isn't hard, but keeping track of slices of an
infinity of possible CA in a music system that has to extract
musical features from them is a bit more complicated.

> But you know about the style stuff? I honestly don't
> think they sound all that different, with the exception of
> patch choices!

There's some rhythm stuff, but I agree. But the part extraction
and patch choice stuff isn't trivial, Jon. In fact, you'll notice
they don't explain how it works.

> > I could live in an applet -- that would be easy. The point is,
> > it's all running server-side, and delivered with straight html
> > and javascript.
>
> Oh, I suppose. But sometimes there isn't a whole lot of
> difference between the two, just where the computing actually
> takes place. I haven't ever gotten all that excited as to
> whether things were client-or server-side. When I'm not a
> programmer, I'm a user.

So as a user you might be concerned with security risks (in fact
you suggest you crippled your DAW because of them), making sure
you have the right JVM (I have 3 on my machine -- all required
by different apps I run), large code downloads at runtime, and
the fact that they never got the widgets to look decent or
conistent across platforms / display resolutions. As a user
you might care about speed (since Java applets are hideously
slow), and you might care about being able to use the application
at a friend's house or public terminal without having to go
through all of the above.

-Carl

🔗Jon Szanto <jszanto@...>

9/15/2005 11:20:15 PM

--- In metatuning@yahoogroups.com, "Carl Lumma" <clumma@y...> wrote:
> Running a CA isn't hard, but keeping track of slices of an
> infinity of possible CA in a music system that has to extract
> musical features from them is a bit more complicated.

Yes, I realize that. One of the funny things is that is my
explorations in CA, going all the way back to the first column that
Gardner published his "Life" simulation, that caused me to - some 30
years later - decide to get Wolfram's book. Me, a fairly
non-mathematical person!

> There's some rhythm stuff, but I agree. But the part extraction
> and patch choice stuff isn't trivial, Jon. In fact, you'll notice
> they don't explain how it works.

It must seem harder to you, and that's OK (I could be wrong!), but I
don't think their 'secrecy' means anything.

> So as a user you might be concerned with security risks (in fact
> you suggest you crippled your DAW because of them)

Wrong - I didn't cripple it, I just don't have it connected to
anything else, and then I don't have to have any layers of protection,
and I can strip out other things, to have an OS I don't really love be
lean and mean and optimized for digital audio. Lots of people do this.
But I don't take offense at the term cripple, just wanting to be clear.

Some of your client side points are well taken, Carl. It's just that
right now I couldn't use their Tones module on something that isn't
net-connected, and that doesn't make me smile. But it is far from the
most important thing in the world! :)

Cheers,
Jon

🔗Carl Lumma <clumma@...>

9/16/2005 12:41:54 AM

> > There's some rhythm stuff, but I agree. But the part extraction
> > and patch choice stuff isn't trivial, Jon. In fact, you'll notice
> > they don't explain how it works.
>
> It must seem harder to you, and that's OK (I could be wrong!), but I
> don't think their 'secrecy' means anything.

Maybe you can explain to us how to do feature-extraction on 1-D
CA. How many lines of code do you think you'll need?

> > So as a user you might be concerned with security risks (in fact
> > you suggest you crippled your DAW because of them)
>
> Wrong - I didn't cripple it, I just don't have it connected to
> anything else,

Cripp-led.

> and then I don't have to have any layers of protection,
> and I can strip out other things, to have an OS I don't really
> love be lean and mean and optimized for digital audio. Lots of
> people do this.

I don't have any layers of protection.

-Carl