Constant shifting of key

Constant shifting of key can be dealt with so far as retuning it to
meantone goes by first obtaining or creating a midi file with the key
data in it. Scala will then notate it following the key changes. As a
first pass at Wagner, this is an essential step.

I'm rendering such a preliminary version of the first act of Tristan
in 31 equal right now. Since it is nearly 80 minutes long, it is going
to be too long to put up on a web page, though of an excellent size
for sticking on a CD. The trouble is, if I give my conclusions on how
well it worked out who will believe them?

Gene,

Would the first 10-15 minutes of Tristan adequately
demonstrate the effects of tuning it in 31-EDO?

Regards,
Yahya

-----Original Message-----
Date: Mon, 02 May 2005 22:06:15 -0000
From: "Gene Ward Smith" <gwsmith@...>

Constant shifting of key can be dealt with so far as retuning it to
meantone goes by first obtaining or creating a midi file with the key
data in it. Scala will then notate it following the key changes. As a
first pass at Wagner, this is an essential step.

I'm rendering such a preliminary version of the first act of Tristan
in 31 equal right now. Since it is nearly 80 minutes long, it is going
to be too long to put up on a web page, though of an excellent size
for sticking on a CD. The trouble is, if I give my conclusions on how
well it worked out who will believe them?

________________________________________________________________________

--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.308 / Virus Database: 266.11.0 - Release Date: 29/4/05

--- In tuning@yahoogroups.com, "Yahya Abdal-Aziz" <yahya@m...> wrote:
>
> Gene,
>
> Would the first 10-15 minutes of Tristan adequately
> demonstrate the effects of tuning it in 31-EDO?

I think so. While I havn't gotten a page for it yet, I've uploaded two
files for when I do have a page. One is the prelude to the first act
of Tristan in 31:

http://66.98.148.43/~xenharmo/mp3/wagner/prelt31.mp3

Another is the Meistersinger overture in 55:

http://66.98.148.43/~xenharmo/mp3/wagner/meist55.mp3

I figure something in 50 and something in 43 would round this out.

Gene,

Thanks, I'm downloading them for a listen. Will take a while
at 28.8kbps, but!

Call me silly, but I find it easier to compare like with like ... :-)
How about putting the _same_ music into 31, 43, 50 and 55?

Regards,
Yahya

-----Original Message-----
________________________________________________________________________
From: "Gene Ward Smith" <gwsmith@...>
Subject: Re: Constant shifting of key

--- In tuning@yahoogroups.com, "Yahya Abdal-Aziz" <yahya@m...> wrote:
>
> Gene,
>
> Would the first 10-15 minutes of Tristan adequately
> demonstrate the effects of tuning it in 31-EDO?

I think so. While I havn't gotten a page for it yet, I've uploaded two
files for when I do have a page. One is the prelude to the first act
of Tristan in 31:

http://66.98.148.43/~xenharmo/mp3/wagner/prelt31.mp3

Another is the Meistersinger overture in 55:

http://66.98.148.43/~xenharmo/mp3/wagner/meist55.mp3

I figure something in 50 and something in 43 would round this out.

________________________________________________________________________

--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.308 / Virus Database: 266.11.5 - Release Date: 4/5/05

I think Gene has already done an excellent job by putting digital samples of Wagner's provocatively ingenious music online. Still, brother Yahya is right, a comparison feature would be very well suited to `microtonal listeners` like us.

Congratulations Gene.
Ozan
----- Original Message -----
From: Yahya Abdal-Aziz
To: tuning@yahoogroups.com
Sent: 06 Mayıs 2005 Cuma 16:17
Subject: [tuning] RE: Constant shifting of key

Gene,

Thanks, I'm downloading them for a listen. Will take a while
at 28.8kbps, but!

Call me silly, but I find it easier to compare like with like ... :-)
How about putting the _same_ music into 31, 43, 50 and 55?

Regards,
Yahya

-----Original Message-----
________________________________________________________________________
From: "Gene Ward Smith" <gwsmith@...>
Subject: Re: Constant shifting of key

--- In tuning@yahoogroups.com, "Yahya Abdal-Aziz" <yahya@m...> wrote:
>
> Gene,
>
> Would the first 10-15 minutes of Tristan adequately
> demonstrate the effects of tuning it in 31-EDO?

I think so. While I havn't gotten a page for it yet, I've uploaded two
files for when I do have a page. One is the prelude to the first act
of Tristan in 31:

http://66.98.148.43/~xenharmo/mp3/wagner/prelt31.mp3

Another is the Meistersinger overture in 55:

http://66.98.148.43/~xenharmo/mp3/wagner/meist55.mp3

I figure something in 50 and something in 43 would round this out.

--- In tuning@yahoogroups.com, "Yahya Abdal-Aziz" <yahya@m...> wrote:

> Call me silly, but I find it easier to compare like with like ... :-)
> How about putting the _same_ music into 31, 43, 50 and 55?

One possibility would be to make the Scala seq files available. Once
you've tuned to one brand of meantone, it is trival to switch to
another. You'd need to install not only Scala but Timidity, and
download a good soundfont such as SGN180 or Merlin Vienna.

Gene,

You wrote:
________________________________________________________________________
Date: Sat, 07 May 2005 01:17:34 -0000

--- In tuning@yahoogroups.com, "Yahya Abdal-Aziz" <yahya@m...> wrote:

> Call me silly, but I find it easier to compare like with like ... :-)
> How about putting the _same_ music into 31, 43, 50 and 55?

One possibility would be to make the Scala seq files available. Once
you've tuned to one brand of meantone, it is trival to switch to
another. You'd need to install not only Scala but Timidity, and
download a good soundfont such as SGN180 or Merlin Vienna.
________________________________________________________________________

Yes, I've got Timidity downloaded but not yet had enough reason
to install it. Now I will.

Thanks for the suggestions on soundfonts, too.

Regards,
Yahya

--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.308 / Virus Database: 266.11.6 - Release Date: 6/5/05

--- In tuning@yahoogroups.com, "Gene Ward Smith" <gwsmith@s...> wrote:
> One is the prelude to the first act
> of Tristan in 31:
>
> http://66.98.148.43/~xenharmo/mp3/wagner/prelt31.mp3
>
> Another is the Meistersinger overture in 55:
>
> http://66.98.148.43/~xenharmo/mp3/wagner/meist55.mp3
>
> I figure something in 50 and something in 43 would round this out.

I spent a couple of hours of modem time downloading the Prelude. I
find it very difficult to listen to because of the incessant wide
electronic vibrato. This degree of vibrato is certainly not either
authentic or realistic (as if every cello in the section vibrated
together!). When you listen to a good orchestral string section you
find the individual players' vibratos are inaudible.

Now, I understand why the vibrato is there, because electronic notes
sound 'dead' - but surely there must be some other way of enlivening
the sound? Adjustment of the attack-sustain-decay business? What
actually happens in a good performance is that the note gradually and
subtly changes dynamic and timbre while being sustained, without the
need for vibrato.

I do find it almost impossible to judge tuning if all the parts are
vibrating at once.

The rapid scale passages in the violins sound not legato enough -
again I don't know if there is some way to make it sound as if they
are slurred together. Also the timpani roll should be faster (and
quieter?). But these are minor points.

~~~T~~~

--- In tuning@yahoogroups.com, "Tom Dent" <tdent@a...> wrote:
> I spent a couple of hours of modem time downloading the Prelude. I
> find it very difficult to listen to ...

All of your points are well taken. Gene's particular manner of
creating electronic versions of orchestral/chamber pieces, via midi
files, uses one approach. However, carefully crafted orchestral
emulations can, and do, create a much better simulation of the sound
world of an orchestra. It isn't easy, and takes a lot of time, and
most of these - naturally - are done in standard 12tet. It would add
another layer of complexity, and certainly another order of magnitude
of time spent, to do a more faithful version in alternate tunings. But
quite musical results can be accomplished, with layered samples, the
playing of each line individually with a attention to nuances applied,
etc.

The craft of mimicing an orchestra through electronic means is
standard fare in film work these days, and can be shockingly close to
reality. It would be really interesting to someday hear the results of
those kind of craft skills applied to non-12tet interpretations of
orchestral scores.

Cheers,
Jon

Dear Joe, what is the minimal list of electronic equipment and software required to produce this "shockingly-close to reality" microtonal-orchestral performance? A price tag included will be most appreciated.

Thank you,
Ozan
----- Original Message -----
From: Jon Szanto
To: tuning@yahoogroups.com
Sent: 16 Mayıs 2005 Pazartesi 3:20
Subject: [tuning] Re: Constant shifting of key

--- In tuning@yahoogroups.com, "Tom Dent" <tdent@a...> wrote:
> I spent a couple of hours of modem time downloading the Prelude. I
> find it very difficult to listen to ...

All of your points are well taken. Gene's particular manner of
creating electronic versions of orchestral/chamber pieces, via midi
files, uses one approach. However, carefully crafted orchestral
emulations can, and do, create a much better simulation of the sound
world of an orchestra. It isn't easy, and takes a lot of time, and
most of these - naturally - are done in standard 12tet. It would add
another layer of complexity, and certainly another order of magnitude
of time spent, to do a more faithful version in alternate tunings. But
quite musical results can be accomplished, with layered samples, the
playing of each line individually with a attention to nuances applied,
etc.

The craft of mimicing an orchestra through electronic means is
standard fare in film work these days, and can be shockingly close to
reality. It would be really interesting to someday hear the results of
those kind of craft skills applied to non-12tet interpretations of
orchestral scores.

Cheers,
Jon

Dear Jon, what is the minimal list of electronic equipment and software required to produce this "shockingly-close to reality" microtonal-orchestral performance? A price tag included will be most appreciated.

Thank you,
Ozan
----- Original Message -----
From: Jon Szanto
To: tuning@yahoogroups.com
Sent: 16 Mayıs 2005 Pazartesi 3:20
Subject: [tuning] Re: Constant shifting of key

--- In tuning@yahoogroups.com, "Tom Dent" <tdent@a...> wrote:
> I spent a couple of hours of modem time downloading the Prelude. I
> find it very difficult to listen to ...

All of your points are well taken. Gene's particular manner of
creating electronic versions of orchestral/chamber pieces, via midi
files, uses one approach. However, carefully crafted orchestral
emulations can, and do, create a much better simulation of the sound
world of an orchestra. It isn't easy, and takes a lot of time, and
most of these - naturally - are done in standard 12tet. It would add
another layer of complexity, and certainly another order of magnitude
of time spent, to do a more faithful version in alternate tunings. But
quite musical results can be accomplished, with layered samples, the
playing of each line individually with a attention to nuances applied,
etc.

The craft of mimicing an orchestra through electronic means is
standard fare in film work these days, and can be shockingly close to
reality. It would be really interesting to someday hear the results of
those kind of craft skills applied to non-12tet interpretations of
orchestral scores.

Cheers,
Jon

--- In tuning@yahoogroups.com, "Jon Szanto" <jszanto@c...> wrote:

> The craft of mimicing an orchestra through electronic means is
> standard fare in film work these days, and can be shockingly close to
> reality. It would be really interesting to someday hear the results of
> those kind of craft skills applied to non-12tet interpretations of
> orchestral scores.

Film composers, I'm afraid, have technical resources I don't have.
Even if I spent the money, one great difficulty is overcoming the 12et
straightjacket. Timidity is the only program I know which fully
implements the midi tuning standard, and I don't know how to get good
tuning results, other than using Csound, unless I use MTS.

Hi Ozan,

--- In tuning@yahoogroups.com, "Ozan Yarman" <ozanyarman@s...> wrote:
> Dear Joe, what is the minimal list of electronic equipment and
software required to produce this "shockingly-close to reality"
microtonal-orchestral performance? A price tag included will be most
appreciated.

I'm sorry, that isn't quite possible. Usually, this amounts to "the
sky is the limit". When I've got more time I can post a couple of past
links (if they are still good) that gave examples of a couple people
who do this that were quite illustrative.

But you know what? A lot of it has to do with 2 things:

1. Using your ears (and I *don't* mean that facetiously!)
2. Understanding the thing you are trying to replicate

This kind of answer that I'm giving will probably be frustrating, but
the above 2 elements can lead you in quite a number of different
directions. For example, some of the articles I've posted in the past
(mostly on MMM, I'm afraid) were dealing with hardware synthesizers.
As anyone who frequents MMM knows, we're going through a fairly large
shift in electronic music right now from hardware to software. I don't
mean to imply that hardware synths will be replaced by soft, but that
there are now alternatives that are more cost - and space - effective.

If you concentrate only on point #2 above, one could think about Tom's
comment: that an entire violin section can't _musically_ be emulated
by one line or pass through a synth. [Please note that while I am
using the term "synth", you could substitute any sound generation
device, for example Gene uses soundfonts frequently]. If you were to
record a number of passes of the same violin line, with slightly
differing velocities, attacks, aftertouch, etc, you come closer to
mixing the infinite variety of nuances (and fallicies!) happening at
any time in a violin section.

I fear I might babble too much, and that the tuning list might not be
the appropriate forum. But some composers I know have been able to get
good orchestral simulations using a small number of samplers layered
with other instruments, while big-time Hollywood types have racks and
racks of these. The bottom line is listening to your result, and
spending time trying to build in the phrasings and nuances of the type
ensemble you are attempting to represent. Taking a basic midi file and
dumping it through a generalised instrument, even with a little
processing, is probably going to yield lesser results.

Two last things:

1. Trying to recreate orchestral/classical stuff isn't the only way to
go, and Gene has done other stuff with the same tools that I like.

2. I haven't even mentioned the next hurdle - doing all this
convincingly *and* making it microtonal!

Cheers,
Jon

--- In tuning@yahoogroups.com, "Gene Ward Smith" <gwsmith@s...> wrote:
> Film composers, I'm afraid, have technical resources I don't have.

Yes, I understand that and sympathise as well. And their methods might
*not* be the ideal thing for you. On the other hand, I don't think
that you've tried any of the options that I patiently put on MMM a
long time ago for you, WRT phrasing, multiple lines, etc. It would
also require a good midi editing environment and very likely a good
midi input device, like a keyboard. I know this isn't your preferred
method.

> Even if I spent the money, one great difficulty is overcoming the 12et
> straightjacket. Timidity is the only program I know which fully
> implements the midi tuning standard, and I don't know how to get good
> tuning results, other than using Csound, unless I use MTS.

These are the methods if you simply want to pour in a midi file and
have an audio file come out the other end. I think I was pointing to
other methods that would be more labor intensive. Also, as I mentioned
to Ozan, I am not exactly sure how anyone could best go about doing
this; I *know* it can be done in 12, but you would have to be willing
to do a lot of work to get a similar result in non12.

It is a crossroads right now. I'm not convinced we'll see a microtonal
nirvana anytime soon, but at least an awareness of how well a
simulation can be done, along with the awareness of the weaknesses of
the current methods, is good for pondering.

Jon

--- In tuning@yahoogroups.com, "Jon Szanto" <jszanto@c...> wrote:

> If you concentrate only on point #2 above, one could think about Tom's
> comment: that an entire violin section can't _musically_ be emulated
> by one line or pass through a synth. [Please note that while I am
> using the term "synth", you could substitute any sound generation
> device, for example Gene uses soundfonts frequently]. If you were to
> record a number of passes of the same violin line, with slightly
> differing velocities, attacks, aftertouch, etc, you come closer to
> mixing the infinite variety of nuances (and fallicies!) happening at
> any time in a violin section.

It seems to me it could if controller messages were created and
implemented to do that. For instance, a midi controller which for
value n would say to take whatever instrument is playing on the given
track, tweak it a bit, and play it n times with these differing,
tweaked values. This wouldn't be the same as midi controller 93,
"chorus", which says to try to fake it by detuning a bit; it would be
more computationally intensive, but might be expected to lead to much
more realistic results.

> 2. I haven't even mentioned the next hurdle - doing all this
> convincingly *and* making it microtonal!

And that could be a biggie. I'd be interested to know how to get Giga
Studio stuff to deal with microtonality; so far as I know it doesn't.

--- In tuning@yahoogroups.com, "Gene Ward Smith" <gwsmith@s...> wrote:
> It seems to me it could if controller messages were created and
> implemented to do that.

You can try it if you want, but it seems to once again beg the issue:
if you are trying to create a sound environment that is recreational
of a grouped human activity, you need to figure out what the
inflections are that cause it to have that sound.

I've gone over this with you before, so I won't belabor it. It
frequently seems that to recreate a beautiful musical phrase, you have
to input a beautiful musical phrase. I'm not going to posit that you
can't do some manipulation mechanically or algorithmically that will
come closer, but those things might work best for another new kind of
music. But shaping phrases of a violin line with an understanding of
how a violinist plays is paramount to coming closer to the real.

> > 2. I haven't even mentioned the next hurdle - doing all this
> > convincingly *and* making it microtonal!
>
> And that could be a biggie. I'd be interested to know how to get Giga
> Studio stuff to deal with microtonality; so far as I know it doesn't.

It *is* a biggie, and samplers are only going to make it happen in one
way: if you create your own sample sets in the tuning you want. If you
simply find a way to alter the pitch of a really greatly recorded and
produced 12tet sample set, you are going to be mightily dissapointed
in the results, as pitch shifts of a very few cents have a drastic
effect on the timbre of a sample, and trying to have a smooth and
coherent melody line (or seamless balances in a tutti chord) will be
elusive at best.

Jon

--- In tuning@yahoogroups.com, "Jon Szanto" <jszanto@c...> wrote:
> --- In tuning@yahoogroups.com, "Gene Ward Smith" <gwsmith@s...> wrote:

> > It seems to me it could if controller messages were created and
> > implemented to do that.
>
> You can try it if you want, but it seems to once again beg the issue:
> if you are trying to create a sound environment that is recreational
> of a grouped human activity, you need to figure out what the
> inflections are that cause it to have that sound.

Before you had said that people have already figured something of
importance out, namely the laying down of multiple versions of a
single sample. Now you seem to be saying how to improve matters is not
known and requires experimentation.

If the multiple versions business is correct, the problem can be
framed as one of how to render a midi file: the existing midi
controller 93 could be used to define how the multiple versions
business will work. I don't see why this is avoiding the issue--I
think it defines what the issue really is. Presumably we are not going
to adopt the position that things ought to be done in a way as
difficult as possible, so that we suffer for our art?

> But shaping phrases of a violin line with an understanding of
> how a violinist plays is paramount to coming closer to the real.

This raises a new issue, but an interesting one. How do you propose
doing it?

> It *is* a biggie, and samplers are only going to make it happen in one
> way: if you create your own sample sets in the tuning you want. If you
> simply find a way to alter the pitch of a really greatly recorded and
> produced 12tet sample set, you are going to be mightily dissapointed
> in the results, as pitch shifts of a very few cents have a drastic
> effect on the timbre of a sample, and trying to have a smooth and
> coherent melody line (or seamless balances in a tutti chord) will be
> elusive at best.

Another new issue--the desirability of finely divided (and of course
very carefully tuned) samples for midi music. Soundfonts are already
big, and this proposal would make them bigger. Meanwhile, we don't
seem to have any soundfont experts ready to step in and do it. None of
these problems are things that being incredibly rich could not solve,
of course.

--- In tuning@yahoogroups.com, "Gene Ward Smith" <gwsmith@s...> wrote:
> --- In tuning@yahoogroups.com, "Jon Szanto" <jszanto@c...> wrote:
> > --- In tuning@yahoogroups.com, "Gene Ward Smith" <gwsmith@s...> wrote:
>
> > > It seems to me it could if controller messages were created and
> > > implemented to do that.
> >
> > You can try it if you want, but it seems to once again beg the issue:
> > if you are trying to create a sound environment that is recreational
> > of a grouped human activity, you need to figure out what the
> > inflections are that cause it to have that sound.
>
> Before you had said that people have already figured something of
> importance out, namely the laying down of multiple versions of a
> single sample. Now you seem to be saying how to improve matters is not
> known and requires experimentation.

I don't believe you understood me correctly, and there are multiple
issues here. I never said multiple versions of a _sample_, for
instance. I was referring to the fact that multiple 'tracks' (if you
will) of a violin part could be sequenced slightly differently and
then rendered to audio, giving more of a simulation of a section than
simply one track ever could.

I'm not saying experimentation. I'm saying you need to understand what
makes an orchestral section sound and phrase the way it does, if what
you are doing is trying to emulate an orchestra.

But, really, this is nearly identical to a conversation we had a
couple years ago. No need to go over it again.

> If the multiple versions business is correct

Not the point.

> Presumably we are not going
> to adopt the position that things ought to be done in a way as
> difficult as possible, so that we suffer for our art?

Do what you want. All that matters is results.

> > But shaping phrases of a violin line with an understanding of
> > how a violinist plays is paramount to coming closer to the real.
>
> This raises a new issue, but an interesting one. How do you propose
> doing it?

That should be obvious.

> Another new issue--the desirability of finely divided (and of course
> very carefully tuned) samples for midi music.

No kidding. That is, if you want sample based music. If you don't know
how to massage, layer, and manipulate samples, you will still end up
with mechanical sounding music, even with good samples.

> Soundfonts are already
> big, and this proposal would make them bigger.

Not that much serious music uses the soundfont format anyway. Most are
utilizing some form of standard samples, so unless you plan on
spending a lot of time with a conversion unit, sample collections
might be the way to go. But they are expensive (the good ones) and
also require a lot of time and learning to use well. I am not one who
uses them well, have simply spent a lot of time in the company (and
studios) of some who do.

Jon

I made an example.
http://www.geocities.jp/imyfujita/wtc1011p.mid
Shifting Cmajor to Dmajor!!!
Can you find its secret?

Jon Szanto wrote:
>It *is* a biggie, and samplers are only going to make it happen in one
>way: if you create your own sample sets in the tuning you want. If you
>simply find a way to alter the pitch of a really greatly recorded and
>produced 12tet sample set, you are going to be mightily dissapointed
>in the results, as pitch shifts of a very few cents have a drastic
>effect on the timbre of a sample,

That, in my experience, is a great exaggeration. Timbre shifts occur with pitch shifts of a few semitones, perhaps, but not with pitch shifts of cents.

"Ozan Yarman" <ozanyarman@s...> wrote:
> Dear Joe, what is the minimal list of electronic equipment and
>>software required to produce this "shockingly-close to reality"
>microtonal-orchestral performance? A price tag included will be most
> appreciated.

I wouldn't say "shockingly" close, but almost reasonably close these days. And here's a very rough price list:

$1500 modern home-built windows computer w/2 gigabytes ram, PIV 3.2, big fast disk drives, moderately good sound card
$500 gigastudio 3
$500 sequencer and audio recording software
$1000 speakers for computer output
$2000 ($1000-$5000, orchestral sample libraries, prices and comprehensiveness vary greatly)
--------
$5500 plus a LOT of time.
Guglielmo

Guglielmo

Hi,

--- In tuning@yahoogroups.com, Guglielmo <gugliel@g...> wrote:
> That, in my experience, is a great exaggeration. Timbre shifts occur
> with pitch shifts of a few semitones, perhaps, but not with pitch
shifts
> of cents.

Not in my experience, but it may also be what kind of sounds we are
(each) using. For instance, if I take a very nice piano (instrument,
not dynamic) sample: take a C natural and bend it up 100 cents, and
then alternate between that note and a C sharp sample, one can hear a
distinct difference. I leave open the possibility that doing a lot of
editing of individual samples in an audio editor, especially if one
could work at higher sample rates and bit depths, might yield better
results. I believe what I was referring to, however, would be if
someone utilized a standard sample library with a soft sampler, and
depended on midi relaying and pitch bends to effect the changes - I
don't believe that musical results would always be the end of that chase.

> I wouldn't say "shockingly" close, but almost reasonably close these
> days.

Probably my "shockingly" comes from having been with electronics for
over 3 decades, and what passes for an orchestral mockup these days,
compared to what it was like quite a while ago, is shocking. But in
the bigger picture, "reasonably" is most certainly a better term!

Cheers,
Jon

P.S.
> $5500 plus a LOT of time.
> Guglielmo

Guglielmo, do you have any musical examples online somewhere?