advice :: retuning a disklavier

hi all

i'm looking for some advice on retuning a disklavier..

i'll be recording and performing premieres of three new james tenney
spectral variations for player piano in the spring.

the two issues i'd like help with are:

(i) in the upper octave of the piece there are 12 pitch classes. i
have devised a scordatura where
(a) as many strings as possible are tuned down and never as much as
a semitone, and
(b) the few that are rasied are raised no more than +4¢

does that seem reasonable and safe for the instrument?

(ii) the curators have suggested transporting the instrument to the performance venue on the morning of the recording (the performance is
that same evening) and then retuning it there. i've had advice that
it should be retuned, allowed settle and retuned again, so i'm a
little worried about it settling back out of tune.

can anyone advise on the best course of action? (i think there're
cost implications re access to the tuner, but we really need to get
this right as we'll not get a chance to do it again for some time).

any help would be very much appreciated.
ciarán.

ciarán maher
rhizomecowboy.com

> i'll be recording and performing premieres of three new james
> tenney spectral variations for player piano in the spring.

Awesome!

> the two issues i'd like help with are:
>
> (i) in the upper octave of the piece there are 12 pitch classes.
> i have devised a scordatura where (a) as many strings as
> possible are tuned down and never as much as a semitone, and
> (b) the few that are rasied are raised no more than +4¢

Generally speaking, if you want a good result from retuning a
piano you must have 12 pitch classes in every octave. They
don't have to be the same in every octave, but there should be
12, and they must of course be ascending, and definitely no
note should be retuned more than *half* a semitone, or 50
cents.

A good procedure is to take your scale, find offsets from
the nearest ET notes, then find the average (mean) offset, and
finally add this average offset to every offset. This way,
the total tension on the piano changes as little as possible.
Do this for every octave if you have a different pitch classes
in different octaves.

> (ii) the curators have suggested transporting the instrument
> to the performance venue on the morning of the recording (the
> performance is that same evening) and then retuning it there.
> i've had advice that it should be retuned, allowed settle and
> retuned again, so i'm a little worried about it settling back
> out of tune.

Yes, expect to have to tune it at least twice to get a good
result with the new scale, and expect to tune it at least once
after being moved.

-Carl

hey carl

thanks a lot for getting back so quickly.

On 18 Ean 2007, at 17:10, Carl Lumma wrote:

> Generally speaking, if you want a good result from retuning a
> piano you must have 12 pitch classes in every octave. They
> don't have to be the same in every octave, but there should be
> 12, and they must of course be ascending, and definitely no
> note should be retuned more than *half* a semitone, or 50
> cents.
>

right. i was worried a full semitone might be a lot.

the piece needs the first 24 partials of a series based on a low A
(55Hz).
my main difficulty is allocating strings to partials 12 to 24.

i had assumed that the crucial thing was having strings tuned down.
where possible. i'll have another look at it and try to redistribute.
i think it'll require tuning a few more pitches up a little. can you
suggest a safe maximum for retunign upward in that register? i'm
guessing it can't be as much as 50¢?

ciarån
>

ciarán maher
rhizomecowboy.com

> > Generally speaking, if you want a good result from retuning a
> > piano you must have 12 pitch classes in every octave. They
> > don't have to be the same in every octave, but there should be
> > 12, and they must of course be ascending, and definitely no
> > note should be retuned more than *half* a semitone, or 50
> > cents.
>
> right. i was worried a full semitone might be a lot.

I've moved pianos to A=415 before, but I've never raised
that much.

> the piece needs the first 24 partials of a series based on a
> low A (55Hz).
> my main difficulty is allocating strings to partials 12 to 24.

I guess the appropriate E-E octave will map directly to 12-24.
Is A=55 specified by the composer, or can you move it?

> can you suggest a safe maximum for retunign upward in that
> register? i'm guessing it can't be as much as 50¢?

I think it should be OK, but a real piano tuner :) like
Paul or Clark could probably tell you better.

-Carl

On 18 Ean 2007, at 19:50, Carl Lumma wrote:

> I guess the appropriate E-E octave will map directly to 12-24.

it doesn't map very well actually.
the problem is the big gap between 12º and 13º and the cluster at the
other end

if you start with:

12º = E+2 tuned to E-string (+2)

and keep on mapping from there, everything has to go *way* up

i.e. if you mapped 13º to the F-string it'd have to be well over a
semitone etc.

the scord i worked out had to have almost a semitone detune for 24º
and 23º, but everything else was out by no more than about a 1/4tone.
i.e.

24ºl to F-98
23º to E-71
22º to D#-49

and so on.
i'd hoped that the register of the higher pitches might mean you can
get away with a slightly larger detune.

> Is A=55 specified by the composer, or can you move it?

no, i don't think so. it's just a guide.

ciarán maher
rhizomecowboy.com

> > I guess the appropriate E-E octave will map directly to 12-24.
>
> it doesn't map very well actually.
> the problem is the big gap between 12º and 13º and the cluster
> at the other end

Yes, well, there's nothing to be done about that. It's the
nature of the harmonic series.

> if you start with:
>
> 12º = E+2 tuned to E-string (+2)
>
> and keep on mapping from there, everything has to go *way* up
>
> i.e. if you mapped 13º to the F-string it'd have to be well over a
> semitone etc.

You may be using some characters that aren't supported on my
system. I can't read the above.

> > Is A=55 specified by the composer, or can you move it?
>
> no, i don't think so. it's just a guide.

Well you have some flexibility then. Just make sure that
whatever pitch standard you choose is OK with any other
musicians.

-Carl

> You may be using some characters that aren't supported on my
> system. I can't read the above.

sorry carl. i'll pull out the symbols.

if you start with:

12th partial = E+2 tuned to E-string (+2)

and keep on mapping from there, everything has to go way up

i.e. if you mapped the 13th partial to the F-string it'd have to be
well over a semitone up etc.

the scord i worked out had to have almost a semitone detune for the
24th and 23th partials, but everything else was out by no more than
about a 1/4tone.
i.e.

24th partiall to F -98
23rd to E -71
22nd to D-sharp -49

and so on.
i'd hoped that the register of the higher pitches might mean you can
get away with a slightly larger detune.

c

ciarán maher
rhizomecowboy.com

> > You may be using some characters that aren't supported on my
> > system. I can't read the above.
>
> sorry carl. i'll pull out the symbols.
>
> if you start with:
>
> 12th partial = E+2 tuned to E-string (+2)
>
> and keep on mapping from there, everything has to go way up
>
> i.e. if you mapped the 13th partial to the F-string it'd have to be
> well over a semitone up etc.

Yes, I suppose that's true. I've been playing around with 12-24
on MIDI keyboards for years, but I've never included it in my
piano experiments for precisely this reason. I'll defer to a
real piano tuner at this point... I bet there's a way to make
it work well enough.

-Carl

wow

very luckily, clarenz barlow has sent me the original scordatura.

it accords with my own (apart from the 13th which i was tuning down
but is tuned up in the original).

c

On 18 Ean 2007, at 23:21, Carl Lumma wrote:

> Yes, I suppose that's true. I've been playing around with 12-24
> on MIDI keyboards for years, but I've never included it in my
> piano experiments for precisely this reason. I'll defer to a
> real piano tuner at this point... I bet there's a way to make
> it work well enough.

ciarán maher
rhizomecowboy.com

ciarán maher wrote:
>
> wow
>
> very luckily, clarenz barlow has sent me the original scordatura.
>
> it accords with my own (apart from the 13th which i was tuning down
but is tuned up in the original).
>
> On 18 Ean 2007, at 23:21, Carl Lumma wrote:
>
> > Yes, I suppose that's true. I've been playing around with 12-24
> > on MIDI keyboards for years, but I've never included it in my
> > piano experiments for precisely this reason. I'll defer to a
> > real piano tuner at this point... I bet there's a way to make
> > it work well enough.

Ciarán,

Out of curiosity, what was the solution chosen?

Regards,
Yahya

hey yahya

i've attached jim's scordatura (you'll have to zoom in to see it)
ciarán.

On 21 Ean 2007, at 18:49, yahya_melb wrote:

>
> Out of curiosity, what was the solution chosen?

ciarán maher
rhizomecowboy.com



Hi Ciarán,

ciarán maher wrote:
>
> hey yahya
>
> i've attached jim's scordatura (you'll have to zoom in to see it)
> ciarán.
>
>
> On 21 Ean 2007, at 18:49, yahya_melb wrote:
>
> >
> > Out of curiosity, what was the solution chosen?
>
> ciarán maher

Thanks for replying. Sadly, the message says: "Attachment(not
stored) tuning.pdf"; and indeed it is not posible to access any
attachment.

> rhizomecowboy.com

I looked at this entertaining interactive music site, hoping taht I
might find a file there called "tuning.pdf". What I found there
was "something else". ;-)

Perhaps you might upload the file to the Files area of tuning,
tuning_files or its successor? Or just email me the attachment
directly.

Regards,
Yahya

hey yahya

On 25 Ean 2007, at 15:46, yahya_melb wrote:
>

>
> > rhizomecowboy.com
>
> I looked at this entertaining interactive music site, hoping taht I
> might find a file there called "tuning.pdf". What I found there
> was "something else". ;-)

right. that's from a recent installation i made with an animator
friend. there's a gamut of the first 32 partials of a series that're
triggered by the movement and interaction of the objects in the
virtual space. durational.

>
> Perhaps you might upload the file to the Files area of tuning,
> tuning_files or its successor? Or just email me the attachment
> directly.
>

i've uploaded it here:
http://www.rhizomecowboy.com/spectral_variations/tuning.pdf
c

> Regards,

ciarán maher
rhizomecowboy.com