weirdness (or user error) with pianotech, gene's scale, and keyboard maps

(Brief intro: I like this kind of sound: listen to 'pianotech improv #0.mp3' here:

http://www.box.net/shared/m37jhti1og#/shared/m37jhti1og/3/17990454

This is the sound of small in-tune ratios without temperament, changing fundamentals. The scale is one I practiced for a long time years ago, with 36 notes per octave. Mostly 11-limit.)

Next part: I'm trying out Gene's 46 note 'epimorphic' scale. I'm having some problems getting it to sound exactly in tune. When I play along with the sequencer playing my old scale, there is obvious beating: Listen to 'pianotech gene's scale test.mp3'

Obviously, something isn't working right.

It's probably some mistake I'm making, but I can't figure out what it is.

If I try adjacent notes in Gene's scale, these aren't closer, they're farther away.

Plus, I can't get anything starting on A to have that no-beating overtone buzz sound.

According to the keyboard map, A should be the beginning of the scale, yes?

Here is Gene's scala file, with pitch-class names next to the ratios. I ought to be able to play
a=1/1, f=9/8, c=5/4, f#=11/8, c=3/2 and have it match my old 'scale' exactly.

To my ear, the old scale I used to use works fine in Pianotech--the partials lock in, and you get that overtone 'buzz'. (For now we are not debating whether that is desirable, rather, I'm saying that it's how I can tell that the numbers match the sound)

note: the actual scala file does not have the added pitch-class names:

! caleb46.scl
46 note 13-limit epimorphic scale
46
!
49/48 bb
36/35 b
21/20 c
16/15 c#
13/12 d
11/10 d#
10/9 e
9/8 f
8/7 f#
7/6 g
13/11 g#
6/5 a
11/9 a#
16/13 b
5/4 c
14/11 c#
9/7 d
21/16 d#
4/3 e
27/20 f
11/8 f#
7/5 g
45/32 g#
10/7 a
16/11 a#
40/27 b
3/2 c
32/21
14/9
11/7
8/5
13/8
18/11
5/3
22/13
12/7
7/4
16/9
9/5
11/6
13/7
15/8
21/11
35/18
55/28
2/1

and here's the keyboard map:

! Size of map:
128
! First MIDI note number to retune:
0
! Last MIDI note number to retune:
127
! Scale degree 0 is mapped to MIDI note:
9
! Reference MIDI note for which frequency is given:
45
! Frequency of the above MIDI note:
110.0000000000
! Scale degree to consider as formal octave:
128
! Mapping:
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127

So, why am I getting inconsistent results?

I'm also now trying another way of testing for accuracy: running the piano through the distortion processing that Logic provides. You can hear difference tones clearly this way.

But the problem is audible without distortion in any case.

I've also set the unison width to 0, the octave stretching to 1.00, and the string length to max, to reduce enharmonicity.

The scale seems to be accurate starting on C1!

My tentative conclusion is that for some reason the scale is starting on pitch C1, or note 036, no matter what keymap.

Why this should be happening is anybody's guess.

???

Caleb

On Jul 15, 2010, at 10:26 AM, calebmrgn wrote:

> (Brief intro: I like this kind of sound: listen to 'pianotech improv #0.mp3' here:
>
> http://www.box.net/shared/m37jhti1og#/shared/m37jhti1og/3/17990454
>
> This is the sound of small in-tune ratios without temperament, changing fundamentals. The scale is one I practiced for a long time years ago, with 36 notes per octave. Mostly 11-limit.)
>
> Next part: I'm trying out Gene's 46 note 'epimorphic' scale. I'm having some problems getting it to sound exactly in tune. When I play along with the sequencer playing my old scale, there is obvious beating: Listen to 'pianotech gene's scale test.mp3'
>
> Obviously, something isn't working right.
>
> It's probably some mistake I'm making, but I can't figure out what it is.
>
> If I try adjacent notes in Gene's scale, these aren't closer, they're farther away.
>
> Plus, I can't get anything starting on A to have that no-beating overtone buzz sound.
>
> According to the keyboard map, A should be the beginning of the scale, yes?
>
> Here is Gene's scala file, with pitch-class names next to the ratios. I ought to be able to play
> a=1/1, f=9/8, c=5/4, f#=11/8, c=3/2 and have it match my old 'scale' exactly.
>
> To my ear, the old scale I used to use works fine in Pianotech--the partials lock in, and you get that overtone 'buzz'. (For now we are not debating whether that is desirable, rather, I'm saying that it's how I can tell that the numbers match the sound)
>
> note: the actual scala file does not have the added pitch-class names:
>
> ! caleb46.scl
> 46 note 13-limit epimorphic scale
> 46
> !
> 49/48 bb
> 36/35 b
> 21/20 c
> 16/15 c#
> 13/12 d
> 11/10 d#
> 10/9 e
> 9/8 f
> 8/7 f#
> 7/6 g
> 13/11 g#
> 6/5 a
> 11/9 a#
> 16/13 b
> 5/4 c
> 14/11 c#
> 9/7 d
> 21/16 d#
> 4/3 e
> 27/20 f
> 11/8 f#
> 7/5 g
> 45/32 g#
> 10/7 a
> 16/11 a#
> 40/27 b
> 3/2 c
> 32/21
> 14/9
> 11/7
> 8/5
> 13/8
> 18/11
> 5/3
> 22/13
> 12/7
> 7/4
> 16/9
> 9/5
> 11/6
> 13/7
> 15/8
> 21/11
> 35/18
> 55/28
> 2/1
>
> and here's the keyboard map:
>
> ! Size of map:
> 128
> ! First MIDI note number to retune:
> 0
> ! Last MIDI note number to retune:
> 127
> ! Scale degree 0 is mapped to MIDI note:
> 9
> ! Reference MIDI note for which frequency is given:
> 45
> ! Frequency of the above MIDI note:
> 110.0000000000
> ! Scale degree to consider as formal octave:
> 128
> ! Mapping:
> 0
> 1
> 2
> 3
> 4
> 5
> 6
> 7
> 8
> 9
> 10
> 11
> 12
> 13
> 14
> 15
> 16
> 17
> 18
> 19
> 20
> 21
> 22
> 23
> 24
> 25
> 26
> 27
> 28
> 29
> 30
> 31
> 32
> 33
> 34
> 35
> 36
> 37
> 38
> 39
> 40
> 41
> 42
> 43
> 44
> 45
> 46
> 47
> 48
> 49
> 50
> 51
> 52
> 53
> 54
> 55
> 56
> 57
> 58
> 59
> 60
> 61
> 62
> 63
> 64
> 65
> 66
> 67
> 68
> 69
> 70
> 71
> 72
> 73
> 74
> 75
> 76
> 77
> 78
> 79
> 80
> 81
> 82
> 83
> 84
> 85
> 86
> 87
> 88
> 89
> 90
> 91
> 92
> 93
> 94
> 95
> 96
> 97
> 98
> 99
> 100
> 101
> 102
> 103
> 104
> 105
> 106
> 107
> 108
> 109
> 110
> 111
> 112
> 113
> 114
> 115
> 116
> 117
> 118
> 119
> 120
> 121
> 122
> 123
> 124
> 125
> 126
> 127
>
> So, why am I getting inconsistent results?
>
>

On 15 July 2010 18:17, caleb morgan <calebmrgn@...> wrote:

> My tentative conclusion is that for some reason the scale is starting on pitch C1, or note 036, no matter what keymap.

Can you locate the pure octaves and fifths? That's what I do to debug
keyboard mappings.

Graham

All the 2/1's sound pure.

There are too many 3/2's that are close.

But I was able to make a list of the accurate 5/4's by ear, starting with A-1 (lowest note on my MIDI controller). The numbers are from the Pianotech 'Options' display.

Good 5/4's presently occurring with this scale:

A-1 (21) with C1 (36)
E0 (28) " G1/43
F#0 and A1
G0 and Bb1
C1 and Eb2
D1 and F2
E1 and G2
G2 and Bb3
A2 and C4
==========
G3 and A#4 (cycle repeats, more or less)

I'm not smart enough right now (or perhaps ever) to figure out where the tuning base (first note) of Gene's scale is, given this.

On Jul 15, 2010, at 1:42 PM, Graham Breed wrote:

> On 15 July 2010 18:17, caleb morgan <calebmrgn@...> wrote:
>
> > My tentative conclusion is that for some reason the scale is starting on pitch C1, or note 036, no matter what keymap.
>
> Can you locate the pure octaves and fifths? That's what I do to debug
> keyboard mappings.
>
> Graham
>

I'll throw in this

I am working on an orchestrated version of Kiev Decays into Dust with GPO.

Pianoteq and GPO interpret the tuning I used very differently. I had to set
the center of the tuning in GPO at B4 (or close to that) and still it was
and octave high and some notes caused beats when played together (though
that might be tuning errors of the samples).

to be clear GPO aria allows you to adjust the center of the tuning though I
don't know exactly what that means yet.

Chris

On Thu, Jul 15, 2010 at 3:45 PM, caleb morgan <calebmrgn@...> wrote:

>
>
> All the 2/1's sound pure.
>
> There are too many 3/2's that are close.
>
> But I was able to make a list of the accurate 5/4's by ear, starting with
> A-1 (lowest note on my MIDI controller). The numbers are from the Pianotech
> 'Options' display.
>
> *Good 5/4's presently occurring with this scale:*
>
> A-1 (21) with C1 (36)
> E0 (28) " G1/43
> F#0 and A1
> G0 and Bb1
> C1 and Eb2
> D1 and F2
> E1 and G2
> G2 and Bb3
> A2 and C4
> ==========
> G3 and A#4 (cycle repeats, more or less)
>
>
>
> I'm not smart enough right now (or perhaps ever) to figure out where the
> tuning base (first note) of Gene's scale is, given this.
>
>
>
>
>
> On Jul 15, 2010, at 1:42 PM, Graham Breed wrote:
>
>
>
> On 15 July 2010 18:17, caleb morgan <calebmrgn@yahoo.com<calebmrgn%40yahoo.com>>
> wrote:
>
> > My tentative conclusion is that for some reason the scale is starting on
> pitch C1, or note 036, no matter what keymap.
>
> Can you locate the pure octaves and fifths? That's what I do to debug
> keyboard mappings.
>
> Graham
>
>
>
>

Hi Chris,

I'll throw in this
>
> I am working on an orchestrated version of Kiev Decays into Dust with GPO.
>
> Pianoteq and GPO interpret the tuning I used very differently. I had to set
> the center of the tuning in GPO at B4 (or close to that) and still it was
> and octave high and some notes caused beats when played together (though
> that might be tuning errors of the samples).
>
> to be clear GPO aria allows you to adjust the center of the tuning though I
> don't know exactly what that means yet.
>
> Chris
>

I'm thinking about buying GPO4, and this was in the hope that they tuned the
samples correctly.
I'm running into a lot of basly tuned soundfonts, and haven't for instance
found a good trombone soundfont yet.
So you're actually hearing beats between 2 pure fifths for instance, or do
you mean even when playing a unison or octave?

The center of the tuning adjustment is I think to adjust the frequency of
the 1/1 of the scale.

Marcel

The beating would be for what I think are unisons - BUT this is between GPO
and pianoteq - GPO may very well be self consistent.

I will look into this further. Been very busy today.

Chris

On Thu, Jul 15, 2010 at 4:12 PM, Marcel de Velde <m.develde@...>wrote:

>
>
>
> Hi Chris,
>
>
> I'll throw in this
>>
>> I am working on an orchestrated version of Kiev Decays into Dust with GPO.
>>
>> Pianoteq and GPO interpret the tuning I used very differently. I had to
>> set the center of the tuning in GPO at B4 (or close to that) and still it
>> was and octave high and some notes caused beats when played together (though
>> that might be tuning errors of the samples).
>>
>> to be clear GPO aria allows you to adjust the center of the tuning though
>> I don't know exactly what that means yet.
>>
>> Chris
>>
>
> I'm thinking about buying GPO4, and this was in the hope that they tuned
> the samples correctly.
> I'm running into a lot of basly tuned soundfonts, and haven't for instance
> found a good trombone soundfont yet.
> So you're actually hearing beats between 2 pure fifths for instance, or do
> you mean even when playing a unison or octave?
>
> The center of the tuning adjustment is I think to adjust the frequency of
> the 1/1 of the scale.
>
> Marcel
>
>

> The beating would be for what I think are unisons - BUT this is between GPO
> and pianoteq - GPO may very well be self consistent.
>
> I will look into this further. Been very busy today.
>
> Chris
>

Aaah ok, between GPO and Pianoteq.
Ah and I couldn't wait and got GPO anyhow :)
The trombone samples aren't that great, but the french horn samples may do
better in their place.
Experimenting a bit now.
But many of the other sounds are ok it seems.
With some work this may sound reasonably pleasant for some music, and it's
convenient to work with.

Marcel

I think GPO is good. Only the use of CC1 for volume gets in the way.
Well worth the (relatively) modest cost.

Now Marcel - please tell me - even if a guess - how can your M-JI fit
a serialist music?
I'm quite curious as to the logic behind it.

Chris

> Aaah ok, between GPO and Pianoteq.
> Ah and I couldn't wait and got GPO anyhow :)
> The trombone samples aren't that great, but the french horn samples may do better in their place.
> Experimenting a bit now.
> But many of the other sounds are ok it seems.
> With some work this may sound reasonably pleasant for some music, and it's convenient to work with.
>
> Marcel

Well, working with Gene's scale got me going on making my own 13-limit-ish 48-pitch scale. (It's got one 17). Unequal step-sizes are fine.

I intend to write a series of solo piano pieces for Pianotech with this.

I'm posting this to hold my own feet to the fire, as it were.

If I don't write a bunch of pieces with this, you can all call me a contemptible slug. Or tell me to work in a bank. I should be so lucky.

I might tweak a few notes. Experience with this list has shown me that it's folly to ask for certain kinds of help--each of us is working on an idiosyncratic concept. Little tips, yes. Aesthetics, overall direction, no. Please, no discouraging words--I'm depressed enough as it is.

For my ear, simply put, overtone-series scales have a special power, as does the sound of the scale changing over time (chord changes).

This taste reflects my background.

With four copies of Pianotech running in Logic, I can get four sounding octaves. Oddly, I can't seem to get more--keymaps higher and lower than this don't seem to work.

Here's an annotated version of the scale. I'd be curious if it's nearly identical to any that other people have used. I'd be surprised if it weren't nearly the same.

This seems to play in tune, but for some reason it starts on A2--(I don't understand why--the behavior of Pianotech and other synths with Scala files and keymaps remains mysterious.)

I hope this formats ok, so that the columns of cents, keys, and ratios line up vertically.

below is the annotated version of the Scala file.

caleb

! caleb48.scl
!
48 note 13-limit moronic scale, with a 17
48 PC RATIO
! a2 1/1
104.9 bb2 17/16
111.7 b2 16/15
128.3 c3 14/13
138.6 c#3 13/12
150.6 d3 12/11

165 d#3 11/10
182.4 e3 10/9
203.9 f3 9/8
231.2 f#3 8/7
266.9 g3 7/6
289.2 g#3 13/11

294.1 a3 32/27
315.6 bb3 6/5
347.4 b3 11/9
359.47 c4 16/13
386.3 c#4 5/4
417.5 d4 14/11

435.1 d#4 9/7
454.2 e4 13/10
498 f4 4/3
536.95 f#4 15/11
551.3 g4 11/8
563.4 g#4 18/13

582.5 a4 7/5
617.5 a#4 10/7
636.6 b4 13/9
648.7 c5 16/11
663 c#5 22/15
702 d5 3/2

745.8 d#5 20/13
764.9 e5 14/9
782.5 f5 11/7
813.7 f#5 8/5
840.53 g5 13/8
852.6 g#5 18/11

884.4 a5 5/3
905.9 a#5 27/16
933.1 b5 12/7
968.8 c6 7/4
996.1 c#6 16/9
1017.6 d6 9/5

1035 d#6 20/11
1049.4 e6 11/6
1061.4 f6 24/13
1071.7 f#6 13/7
1088.3 g6 15/8
1115.5 g#6 [4/3 above 10/7] or 40/21

1200 a6 2/1

On Jul 15, 2010, at 3:45 PM, caleb morgan wrote:

>
> All the 2/1's sound pure.
>
> There are too many 3/2's that are close.
>
> But I was able to make a list of the accurate 5/4's by ear, starting with A-1 (lowest note on my MIDI controller). The numbers are from the Pianotech 'Options' display.
>
> Good 5/4's presently occurring with this scale:
>
> A-1 (21) with C1 (36)
> E0 (28) " G1/43
> F#0 and A1
> G0 and Bb1
> C1 and Eb2
> D1 and F2
> E1 and G2
> G2 and Bb3
> A2 and C4
> ==========
> G3 and A#4 (cycle repeats, more or less)
>
>
>
> I'm not smart enough right now (or perhaps ever) to figure out where the tuning base (first note) of Gene's scale is, given this.
>
>
>
>
>
> On Jul 15, 2010, at 1:42 PM, Graham Breed wrote:
>
>>
>> On 15 July 2010 18:17, caleb morgan <calebmrgn@...> wrote:
>>
>> > My tentative conclusion is that for some reason the scale is starting on pitch C1, or note 036, no matter what keymap.
>>
>> Can you locate the pure octaves and fifths? That's what I do to debug
>> keyboard mappings.
>>
>> Graham
>>
>
>
>

I'm changing the question Chris asked Marcel to a more general one, because I've tried to come up with some answers, the question interests me.

These would not be the same as Marcel's answers.

This isn't too technical or too serious...

1) You use 12-tone rows, and they are 'bent' to fit in a JI scale.

2) A completely different approach: The numerical series is treated as a contour of relatively higher and lower pitches instead of as a succession of particular pitch-classes. The contour is mapped to some set of pitches that happens to be tuned to something like JI.

3) You use more or fewer than 12 tones for series.

4) You use a large collection of pitches--extended JI. You live with comma shifts or different versions of pitch-classes.

5) You change your tuning base as needed--this relates to 4.

5) You fudge by using instruments with chorus.

6) You write music of great difficulty, like Ben Johnston.

Personally, I've come to feel that for my personal sanity I'm better off keeping serialism and JI separate, but I've made attempts to combine them in the past.

caleb

On Jul 16, 2010, at 9:39 AM, Chris Vaisvil wrote:

> I think GPO is good. Only the use of CC1 for volume gets in the way.
> Well worth the (relatively) modest cost.
>
> Now Marcel - please tell me - even if a guess - how can your M-JI fit
> a serialist music?
> I'm quite curious as to the logic behind it.
>
> Chris
>
> > Aaah ok, between GPO and Pianoteq.
> > Ah and I couldn't wait and got GPO anyhow :)
> > The trombone samples aren't that great, but the french horn samples may do better in their place.
> > Experimenting a bit now.
> > But many of the other sounds are ok it seems.
> > With some work this may sound reasonably pleasant for some music, and it's convenient to work with.
> >
> > Marcel
>

> 1) You use 12-tone rows, and they are 'bent' to fit in a JI scale.
>
>
>
> Now Marcel - please tell me - even if a guess - how can your M-JI fit
> a serialist music?
> I'm quite curious as to the logic behind it.
>
>
The logic is that any music made in 12edo is interpreted in 2-plane 5-limit
field of pitches.
In one key that gives the 12 tones: 1/1 135/128 9/8 32/27 5/4 4/3 45/32 3/2
405/256 5/3 16/9 15/8 2/1
The full field is infinite, extending the 3/2 fifths infinately in all 4
directions (1/1 down and up, 5/4 down and up)
There is only Major (fundamental bass on the 1/1 row of fifths) and Minor
(fundamental bass on the 5/4 row of fifths).

Serial music thinks it's using 12edo, but it isn't. It's in this 2-plane
5-limit JI aswell.
(I've called it Marcel's Just Intonation btw, M-JI for short)
The only problem is, that with the ultimate serial music it is ultimately
atonal, meaning our ear-brain can't make sense of it.
For this music one can simply play it in any part of the key, it doesn't
matter.
However, I'm guessing no serial music is that perfectly atonal. So even the
most hardcore serial music will still have some tonality, and then rules
start to come into play like fundamental bass and it's movement, and
possible modulations etc.
It'll be tough to analyse though :)

Btw, Carl decided once more to moderate me.
I can't reply any further to this thread as I've decided to unsubscribe from
this list untill I'm put off moderation.

Marcel
www.develde.net