Reply to John Starrett

John, my apologies for taking so long to reply to your posts. The
earliest non-realtime tuning methods that I posted, from about August
1999 to January 2000, specified exact JI vertical intervals and had
fairly poor drift management (compared to the spring model), which
caused a fair amount of short-term drift in overall tuning, that had to
be recovered fairly quickly and noticeably. You are one of the few who
has said he prefers this treatment to the later, much more smooth,
spring model's output.

I'm going to run a test right now to see if that option still works - it
does! In my "comma pump" sequence, tuning is driven down to almost 70
cents below nominal, highly noticeable (though it takes several times
around the pump to happen, as you'd expect).

Here's what I'm curious about: can your desire for more motion be
satisfied using the spring model, but very rigid vertical springs and
much softer horizontal and/or grounding springs? I still _do_ have
problems relaxing the spring matrix when the ratios of stiffness get
extreme, but can achieve results within a cent or two of exact JI, which
I _think_ should sound vertically pretty good. It's not at all clear,
though, whether that would leave enough of that bouncy horizontal motion
to satisfy you.

Don't know if you'd have time or interest, but I'd like to do up two
or three tunings of some piece, including one pre-spring, to see where
the edge of your desire lies. Fun?

JdL

--- In crazy_music@y..., "John A. deLaubenfels" <jdl@a...> wrote:
> John, my apologies for taking so long to reply to your posts. The
> earliest non-realtime tuning methods that I posted, from about
August
> 1999 to January 2000, specified exact JI vertical intervals and had
> fairly poor drift management (compared to the spring model), which
> caused a fair amount of short-term drift in overall tuning, that had
to
> be recovered fairly quickly and noticeably. You are one of the few
who
> has said he prefers this treatment to the later, much more smooth,
> spring model's output.
>
> I'm going to run a test right now to see if that option still works
- it
> does! In my "comma pump" sequence, tuning is driven down to almost
70
> cents below nominal, highly noticeable (though it takes several
times
> around the pump to happen, as you'd expect).
>
> Here's what I'm curious about: can your desire for more motion be
> satisfied using the spring model, but very rigid vertical springs
and
> much softer horizontal and/or grounding springs? I still _do_ have
> problems relaxing the spring matrix when the ratios of stiffness get
> extreme, but can achieve results within a cent or two of exact JI,
which
> I _think_ should sound vertically pretty good. It's not at all
clear,
> though, whether that would leave enough of that bouncy horizontal
motion
> to satisfy you.
>
> Don't know if you'd have time or interest, but I'd like to do up two
> or three tunings of some piece, including one pre-spring, to see
where
> the edge of your desire lies. Fun?
>
> JdL

Yes, fun. I would be glad to test your method. I know there was at
least one other member who liked the effect. But I warn you, it is
novelty I crave. I also enjoy a pure comma pump.

John Starrett

[John wrote:]
>Yes, fun. I would be glad to test your method. I know there was at
>least one other member who liked the effect. But I warn you, it is
>novelty I crave. I also enjoy a pure comma pump.

Have you heard the negative melody spring Chopin on my web site? That's
got some "novelty" in it. Pure comma pump could mean very loose
grounding springs, allowing a lot of drift either up or down. But if by
this you mean that the comma is recovered suddenly (as I seem to recall
you saying), my methods may be strangely "deficient".

What sequence would you like to hear comparisons of?

JdL

Hi John,

> Don't know if you'd have time or interest, but I'd like to do up two
> or three tunings of some piece, including one pre-spring, to see where
> the edge of your desire lies. Fun?

I'm interested in this experiment too; don't think I've heard the
early pre-spring tunings :-).

Robert

[I wrote:]
>>Don't know if you'd have time or interest, but I'd like to do up two
>>or three tunings of some piece, including one pre-spring, to see where
>>the edge of your desire lies. Fun?

[Robert wrote:]
>I'm interested in this experiment too; don't think I've heard the
>early pre-spring tunings :-).

No, I don't think you have! You, too, are welcome to pick a piece.
The one I remember posting long ago was Beethoven's Pathetique Sonata,
quite vivid but in truth a rather poor sequence (much over-quantized).

JdL

Hi John,

Bach's c major prelude from the WTC might sound interesting in strict
adaptive j.i.

http://pmclassical.iicinternet.com/midi.shtml

Would it still work with the repeating notes changing pitch
as the chords change?

I'm curious to find out.

This is one popular piece that I never tire of hearing however
often I hear it.

Robert

<snip>
> Pure comma pump could mean very loose
> grounding springs, allowing a lot of drift either up or down.
<snip>
> JdL

That is the meaning I was going for. I am listening to some of your
Chopin now, but I think what I's like to hear is Scarlatti. How about
Sankey's sequencing of K51? It has several I V I V sequences centered
on the tonic that are then repeated centered on the dominant. That
should stretch your springs! I know Sankey was a little pissy about
you messing with his tuning, but I don't care. This is for science!

John Starrett

[I wrote:]
>>Pure comma pump could mean very loose grounding springs, allowing a
>>lot of drift either up or down.

[John Starrett:]
>That is the meaning I was going for. I am listening to some of your
>Chopin now, but I think what I's like to hear is Scarlatti. How about
>Sankey's sequencing of K51? It has several I V I V sequences centered
>on the tonic that are then repeated centered on the dominant. That
>should stretch your springs! I know Sankey was a little pissy about
>you messing with his tuning, but I don't care. This is for science!

John Sankey was kind enough to grant permission for me to post retunings
of his MIDI sequences. When I wrote back thanking him and admitting
that I don't know who d'Alembert is/was, he wrote back suggesting that
I stick with modern instruments, without, however, withdrawing his
permission.

So, I have extracted the Scarlatti K051 sequence and posted several
options. The original sequence has the following tuning, in cents
relative to 12-tET:

G#/Ab 8: E8 00 3C (-12.50)
D#/Eb 3: E3 1F 3D (-8.62)
A#/Bb A: EB 30 3E (-5.08)
F 5: E5 30 3F (-1.95)
C 0: E0 00 40 (0.00)
G 7: E7 7E 3E (-3.17)
D 2: E2 70 3D (-6.64)
A 9: EA 30 3C (-11.33)
E 4: E4 50 3B (-13.67)
B B: EC 57 3B (-13.50)
F#/Gb 6: E6 0F 3B (-15.26)
C#/Db 1: E1 32 3B (-14.40)

Whew! I'd have to let Mr. Sankey explain that tuning set; it's not what
I'd pick. I first made a 12-tET version, then tuned that in various
ways. John Starrett, did you want anything 5-limit? I assumed, wild
man that you are, that you'd want 7, but I can do others!

Besides the original, I've included:

k051_12z7: reduced to 12-tET, then tuned by JdL the _old_ way,
without springs, to 7-limit.

k051_12dr7: reduced to 12-tET, then tuned by JdL in 7-limit with
fairly rigid vertical springs and quite soft grounding springs,
allowing for quite a bit of drift of the tuning over time.

k051_12vvs7: reduced to 12-tET, then tuned by JdL in 7-limit with
extremely rigid vertical springs and normal grounding and
horizontal springs.

Go to the JdL directory of this group's files area and download the
zip; feedback welcome!

JdL

> John Starrett, did you want anything 5-limit? I assumed,
> wild
> man that you are, that you'd want 7, but I can do others!
>
> Besides the original, I've included:

The original in Sankey's tuning or in 12tet?

> k051_12z7: reduced to 12-tET, then tuned by JdL the _old_ way,
> without springs, to 7-limit.

I laughed my ass off. John Sankey is going to have you killed.

> k051_12dr7: reduced to 12-tET, then tuned by JdL in 7-limit with
> fairly rigid vertical springs and quite soft grounding
> springs,
> allowing for quite a bit of drift of the tuning over time.

Oh yeah.

> k051_12vvs7: reduced to 12-tET, then tuned by JdL in 7-limit with
> extremely rigid vertical springs and normal grounding and
> horizontal springs.

The second half contrasts so sharply with the nice purity of the first
half, this is my least favorite.

Z7 is my favorite for the sheer preverse pleasure of hearing funny
stuff. With a few hand tweaks, dr7 would no longer be funny, but a
really interesting serious tuning choice. Truth is, if I had to choose
one tuning (with which I am familiar) to hear this piece in for the
rest of my life, I would choose Sankey's. Nevertheless is it great to
hear these experiments.
Yes, I would like to hear the five limit versions.

John Starrett