Sims playback in Sibelius

Well, seemingly there is some progress here on the pitch-bend
Sibelius front regarding microtone playback.

Using 128 bits as the "unit" (actually, it should be 127, but
they "round it up") they "suggest" a bend of 16 units for a
quartertone.

Given that scenario, and dividing into 128, the sixth-tone is 10.6
units and the twelfth tone is 5.3 units.

B0 means the command to bend, and 64 (half of 128, obviously, is
the "normal" 12-tET pitch)

So we get *this* for the "uppers" (I dare not say "sharps..." :)

normal = B0, 64

^ = B0, 69.5

> = BO, 74.6

] = BO, 80

Quite frankly, I'm thinking of taking a rather "rash" step and using
the "Tartini" symbols for quartertones, since they are *already* the
default in Sibelius. I know maybe some of the Sims players won't
like that, but I'm not entirely enamored of the Sims symbols for
quartertones.

Now, oddly, the "downers" seem to work somewhat differently, at least
on *my* sound card. The quartertone flats are assigned B0, 80 *as
well as* the quartertone sharps. That's *not* the way it's supposed
to work according to the manual! According to the manual it would be
a *lower* number, 48...

Anyway, I was finding that in order to get the *smaller* microtonal
divisions one actually had to *ADD* to the quartertone number, in
order for it to work right, so I have been getting the following on
the "down" side:

normal = B0, 64

v = BO, 95.9

< = BO, 90.6

[ = BO, 80

Well, that sounds right to *me* anyway, and the numbers can be edited
right on the screen once they have been created using the default
quartertone plugin....

jp

Joe,

Looks like you are making good progress on the Sibelius stuff (I'm only scanning the posts quickly) - good for you! One small point:

--- In tuning@y..., "jpehrson2" <jpehrson@r...> wrote:
> Using 128 bits as the "unit" (actually, it should be 127, but
> they "round it up") they "suggest" a bend of 16 units for a
> quartertone.

I'm not sure, but when you say 127 you might want to see if it matches a lot of programming logic: quite frequently when you set up some structure, say an array, that you want to have 128 units, they are referenced (numbered) 0-127, because the binary code/math works that way.

Don't know if this is the issue here, but whenever I see a 127/128 question, I check on this first.

Later,
Jon

--- In tuning@y..., "jonszanto" <JSZANTO@A...> wrote:

/tuning/topicId_35603.html#35608

> Joe,
>
> Looks like you are making good progress on the Sibelius stuff (I'm
only scanning the posts quickly) - good for you! One small point:
>
> --- In tuning@y..., "jpehrson2" <jpehrson@r...> wrote:
> > Using 128 bits as the "unit" (actually, it should be 127, but
> > they "round it up") they "suggest" a bend of 16 units for a
> > quartertone.
>
> I'm not sure, but when you say 127 you might want to see if it
matches a lot of programming logic: quite frequently when you set up
some structure, say an array, that you want to have 128 units, they
are referenced (numbered) 0-127, because the binary code/math works
that way.
>
> Don't know if this is the issue here, but whenever I see a 127/128
question, I check on this first.
>
> Later,
> Jon

Hey Jon!

Thanks for the tip. When it comes to computer biz, you always "chip"
in! :)

Well, *they're* telling me here that for bends 16=quarter-tone and
32=semitone, and the "default" is the whole step 64, right in the
middle, or half of 128, a "two whole step" range in total so I'm only
going "by the book..."

So, that's obviously, how I get the twelfth of a whole tone as 5.3
and the sixth of a whole tone as 10.6...

THEY were the ones suggesting the 128... I never would have made
that "round up" myself...

Thanks Jon!

jp

> From: jpehrson2 <jpehrson@rcn.com>
> To: <tuning@yahoogroups.com>
> Sent: Tuesday, March 12, 2002 9:36 AM
> Subject: [tuning] Sims playback in Sibelius
>
>
> Well, seemingly there is some progress here on the pitch-bend
> Sibelius front regarding microtone playback.
>
> Using 128 bits as the "unit" (actually, it should be 127, but
> they "round it up") they "suggest" a bend of 16 units for a
> quartertone.

Joe, it can't possibly be 127 ... i has to be 128 because
128 = 2^7, and internal software calculations are always
ultimately based on powers of 2 because computers are binary
beasts.

i think you're confused because binary counting always
starts with zero and not with 1. so there are 128 discrete
units, between 0 and 127.

(i've been delayed sending this post -- i see that Jon
Szanto has already set you straight on this.)

and there seems to be something else wrong here ...
more below...

> Given that scenario, and dividing into 128, the sixth-tone is 10.6
> units and the twelfth tone is 5.3 units.
>
> B0 means the command to bend, and 64 (half of 128, obviously, is
> the "normal" 12-tET pitch)
>
> So we get *this* for the "uppers" (I dare not say "sharps..." :)
>
> normal = B0, 64
>
> ^ = B0, 69.5
>
> > = BO, 74.6
>
> ] = BO, 80

Joe, these figures don't jive.

if the pitch-bend quantization is set to 128 units per
*whole-tone*, where the "whole-tone" is 2^(1/6) ratio,
then were dealing with 6 * 128 = 768edo, which i point
out on my "equal temperaments" Dictionary entry is common
for many electronic instruments and software apps.

The breakdown for 72edo in units of 768edo is as follows:

72edo cents 768edo (= "Sibelius units")

0 0 0
1 16&2/3 10&2/3
2 33&1/3 21&1/3
3 50 32
4 66&2/3 42&2/3
5 83&1/3 53&1/3
6 100 64
7 116&2/3 74&2/3
8 133&1/3 85&1/3
9 150 96
10 166&2/3 106&2/3
11 183&1/3 117&1/3
12 200 128

so the correct figures for your example, rounded
to one decimal place, are:

normal = B0, 64

^ = B0, 74.7

> = BO, 85.3

] = BO, 96

hopefully, i've interpreted correctly what you've said
about Sibelius, and am not leading you further afield...

-monz

_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com

--- In tuning@y..., "monz" <joemonz@y...> wrote:

/tuning/topicId_35603.html#35613

>
> Joe, it can't possibly be 127 ... i has to be 128 because
> 128 = 2^7, and internal software calculations are always
> ultimately based on powers of 2 because computers are binary
> beasts.
>
> i think you're confused because binary counting always
> starts with zero and not with 1. so there are 128 discrete
> units, between 0 and 127.
>
> (i've been delayed sending this post -- i see that Jon
> Szanto has already set you straight on this.)
>

***Thanks, Monz, for both you and Jon setting me straight. Well,
zero certainly does have a useful function sometimes!

>
> and there seems to be something else wrong here ...
> more below...
>
>
> > Given that scenario, and dividing into 128, the sixth-tone is
10.6
> > units and the twelfth tone is 5.3 units.
> >
> > B0 means the command to bend, and 64 (half of 128, obviously, is
> > the "normal" 12-tET pitch)
> >
> > So we get *this* for the "uppers" (I dare not say "sharps..." :)
> >
> > normal = B0, 64
> >
> > ^ = B0, 69.5
> >
> > > = BO, 74.6
> >
> > ] = BO, 80
>
>
> Joe, these figures don't jive.
>
> if the pitch-bend quantization is set to 128 units per
> *whole-tone*,

***Monz... here is where the misunderstanding is. The 128 units go
over TWO whole tones, not just one. That's why the resolution is
only 3 cents!

In other words it's 400/128 = 3.1 cents resolution.
>
>
> so the correct figures for your example, rounded
> to one decimal place, are:
>
> normal = B0, 64
>
> ^ = B0, 74.7
>
> > = BO, 85.3
>
> ] = BO, 96
>
>

The way you have it here, ergo, Monz is the quarter tone at 96.

However, it's the SEMI-tone that's at 96. That's right here in the
manual.

The quarter-tone is at 16, the half-step is at 32.

64+16 = 80, and I didn't make that up, either, since that's the
default that the quartertone plug-in automatically assigns
quartertones.

Watch, I'll do it again... :)

Thanks for the comments, though!

jp

----- Original Message -----
From: jpehrson2 <jpehrson@rcn.com>
To: <tuning@yahoogroups.com>
Sent: Tuesday, March 12, 2002 2:32 PM
Subject: [tuning] Re: Sims playback in Sibelius

> --- In tuning@y..., "monz" <joemonz@y...> wrote:
>
> /tuning/topicId_35603.html#35613
>
> >
> > Joe, it can't possibly be 127 ... i has to be 128 because
> > 128 = 2^7, and internal software calculations are always
> > ultimately based on powers of 2 because computers are binary
> > beasts.
> >
> > i think you're confused because binary counting always
> > starts with zero and not with 1. so there are 128 discrete
> > units, between 0 and 127.
> >
> > (i've been delayed sending this post -- i see that Jon
> > Szanto has already set you straight on this.)
> >
>
> ***Thanks, Monz, for both you and Jon setting me straight. Well,
> zero certainly does have a useful function sometimes!
>
>
> >
> > and there seems to be something else wrong here ...
> > more below...
> >
> >
> > > Given that scenario, and dividing into 128, the sixth-tone is
> 10.6
> > > units and the twelfth tone is 5.3 units.
> > >
> > > B0 means the command to bend, and 64 (half of 128, obviously, is
> > > the "normal" 12-tET pitch)
> > >
> > > So we get *this* for the "uppers" (I dare not say "sharps..." :)
> > >
> > > normal = B0, 64
> > >
> > > ^ = B0, 69.5
> > >
> > > > = BO, 74.6
> > >
> > > ] = BO, 80
> >
> >
> > Joe, these figures don't jive.
> >
> > if the pitch-bend quantization is set to 128 units per
> > *whole-tone*,
>
>
> ***Monz... here is where the misunderstanding is. The 128 units go
> over TWO whole tones, not just one. That's why the resolution is
> only 3 cents!
>
> In other words it's 400/128 = 3.1 cents resolution.
> >
> >
> > so the correct figures for your example, rounded
> > to one decimal place, are:
> >
> > normal = B0, 64
> >
> > ^ = B0, 74.7
> >
> > > = BO, 85.3
> >
> > ] = BO, 96
> >
> >
>
> The way you have it here, ergo, Monz is the quarter tone at 96.
>
> However, it's the SEMI-tone that's at 96. That's right here in the
> manual.
>
> The quarter-tone is at 16, the half-step is at 32.
>
> 64+16 = 80, and I didn't make that up, either, since that's the
> default that the quartertone plug-in automatically assigns
> quartertones.
>
> Watch, I'll do it again... :)
>
> Thanks for the comments, though!

hmmm... that's really weird ... so then Sibelius's tuning
resolution is 384edo. that's a new one for my ET table!

as i said, 768edo is fairly common as a tuning resolution
in computer music, and i'm quite familiar with it because
it's also the resolution used by Texture, the sequencer
program i used to use in the late 1980s-early 1990s, before
i became a fan of Cakewalk.

just to be sure, if i were you, i'd double/triple/quadruple
check those numbers, because 384 is a real oddball AFAIK.
anyone else out there familiar with other apps that use it?

-monz

_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com

--- In tuning@y..., "monz" <joemonz@y...> wrote:

/tuning/topicId_35603.html#35617

>
> hmmm... that's really weird ... so then Sibelius's tuning
> resolution is 384edo. that's a new one for my ET table!
>
> as i said, 768edo is fairly common as a tuning resolution
> in computer music, and i'm quite familiar with it because
> it's also the resolution used by Texture, the sequencer
> program i used to use in the late 1980s-early 1990s, before
> i became a fan of Cakewalk.
>
> just to be sure, if i were you, i'd double/triple/quadruple
> check those numbers, because 384 is a real oddball AFAIK.
> anyone else out there familiar with other apps that use it?
>

***Hello Monz!

Well, without any doubt, on the "course" resolution, which happens to
be the only one I can get to work at the moment, I will quote from
the manual, page 285:

"Bend-by is a number between 0 and 127, where each integer represents
1/32nd of a half-step (semitone). B0,64 procudes a note at its
written pitch; values lower than 64 flatten the note, and values
higher than 64 sharpen it. To make a note sound one half-step
(semitone) higher than written, use B0,96; to make it sound one half-
step (semitone) lower, use B0,32."

Personally, I would *much* rather have a resolution of 128 per WHOLE
tone. Then, of course, I get a resolution of 1.56 cents, but alas...

Now I have a 3.1 cent resolution... unless I can get the "fine"
tuning feature to work which isn't happening...

So, yes, I guess that means 32 divisions per semitone, or 384edo.

Strange, admittedly.

jp