Keyboard designs

Alright Carl (and anyone else who was curious), I've restored Paul's
designs to my photo-space:

http://photobucket.com/albums/v170/igliashon_jones/

This weekend I'll upload the square-based designs that Stevie and I
worked with last weekend...when using square keys in straight rows
some of those ideal-looking designs actually become pretty hopeless.
Doublewide has the easiest chords out of the bunch but the most
difficult/non-intuitive scales. Porcupine is great for scales but
chords are much more stretched and uncomfortable. All in all, Stevie
seemed to prefer George's orignal Bosanquet-based design as the best
compromise between chords and scales; his fingers are much longer
than mine so chords that I thought would be unplayable he has no
problem with (though that's probably just as much due to his 13+
years of piano training as his finger size). Also, we decided that
we're not going to use buttons directly over the key switches, but
rather use a type-writer/cash-register-esque system of levers and
fulcrums, with square shaped keys at on end, springs at the other,
and a pivot axle in the middle. We'll have parallel pivot axles
running along one of the diagonal pitch axes, and they will be
vertically offset to allow for full key motion. This will allow for
square-shaped keys that still have the action of regular piano keys.

However, we're going to build a stryofoam model before making any
committments.

We are also going to re-key an old MIDI controller to the
Pentadecaphonic 8-white 7-black layout for use with 15-equal, or
maybe even Porcupine-15 from 22.

Meanwhile, Stevie is busy building sample patches on his Korg Triton
that are tuned to 22-equal...god bless the retunability of samplers.

-Igliashon

>Alright Carl (and anyone else who was curious), I've restored Paul's
>designs to my photo-space:
>
>http://photobucket.com/albums/v170/igliashon_jones/

Swweeet!

Paul:

() Heavens, did you draw these with the voronoi cell function? :)

() And say, how are you turning temperaments ('porcupine-based')
into x/y values?

>This weekend I'll upload the square-based designs that Stevie and I
>worked with last weekend...

Cool. Keep us posted.

>when using square keys in straight rows some of those ideal-looking
>designs actually become pretty hopeless.

Why are you so wedded to square/straight then?

>Also, we decided that we're not going to use buttons directly over
>the key switches, but rather use a type-writer/cash-register-esque
>system of levers and fulcrums, with square shaped keys at on end,
>springs at the other, and a pivot axle in the middle. We'll have
>parallel pivot axles running along one of the diagonal pitch axes,
>and they will be vertically offset to allow for full key motion.
>This will allow for square-shaped keys that still have the action
>of regular piano keys.

And why would you want that? I've never played on one, so I can't
be sure, but I think buttons shaped exactly like piano keys would be
better than levers even on conventional keyboards. (I tend to hate
MIDI controllers, and one of the reasons is that the keys are so
much harder to use at the back of the keyboard, since the fulcrum
is usually right there. My favorite keyboards, on pianos and organs,
have longer levers. Why not extend this all the way?)

>However, we're going to build a stryofoam model before making any
>committments.

Very good idea.

-Carl

--- In MakeMicroMusic@yahoogroups.com, Carl Lumma <ekin@l...> wrote:

> >Alright Carl (and anyone else who was curious), I've restored
Paul's
> >designs to my photo-space:
> >
> >http://photobucket.com/albums/v170/igliashon_jones/
>
> Swweeet!
>
> Paul:
>
> () Heavens, did you draw these with the voronoi cell function? :)

Yup. I also mentioned other details in a previous post here.

> () And say, how are you turning temperaments ('porcupine-based')
> into x/y values?

Same way as George Secor: x = pitch, y = generators.

>> >Alright Carl (and anyone else who was curious), I've restored
>> >Paul's designs to my photo-space:
>> >
>> >http://photobucket.com/albums/v170/igliashon_jones/
>>
>> Swweeet!
>>
>> Paul:
>>
>> () Heavens, did you draw these with the voronoi cell function? :)
>
>Yup. I also mentioned other details in a previous post here.

Der, I sure missed that!

>> () And say, how are you turning temperaments ('porcupine-based')
>> into x/y values?
>
>Same way as George Secor: x = pitch, y = generators.

Oh.

-Carl

> Why are you so wedded to square/straight then?

One part aesthetics, one part ergonomics (Stevie really didn't like
the angle he had to set his hands at to play the offset/diagonal
designs I originally worked up). Also, for something like Magic, the
keys must be substantially squat, and that makes it tricky to use the
thumb to strike them for someone with big fingers like Stevie. We
could use hexagonal keys, but that doesn't do much better for any of
them. Doublewide in particular needs to be substantially angled to
make scales ergonomic, so changing key shape doesn't really help.
Also remember that we have limited resources for the constuction of
this thing; we don't have the benefit of having a corporation (or
even a team of experienced engineers) develop and manufacture it for
us. This is all out of our own quite shallow pockets, and whatever
we can't buy ready-made we have to make from scratch ourselves.
We're talking low-budget here. Hence the design needs to be simple.

> And why would you want that? I've never played on one, so I can't
> be sure, but I think buttons shaped exactly like piano keys would be
> better than levers even on conventional keyboards.

I agree, but those would be big-ass buttons, which on a 22-tone
Bosanquet layout would make tetrads VERY difficult. With our design
the playing surfaces of the keys are still a square grid-type thing,
it's just that the squares are really the outer-ends of levers.
Picture something like an old-fashioned non-electric typewriter. You
won't have the problems typically associated with playing regular
keyboards since the playable area is equivalent to the very outer
edge of the keys. That way the action will be similar to a
traditional keyboard--in fact, we might even be able to weight the
keys. I may eventually post a diagram of the mechanical aspects, but
university is now in session so my time is significantly lessened.

-Igs

>> Why are you so wedded to square/straight then?
>
>One part aesthetics, one part ergonomics (Stevie really didn't like
>the angle he had to set his hands at to play the offset/diagonal
>designs I originally worked up). Also, for something like Magic, the
>keys must be substantially squat, and that makes it tricky to use the
>thumb to strike them for someone with big fingers like Stevie. We
>could use hexagonal keys, but that doesn't do much better for any of
>them. Doublewide in particular needs to be substantially angled to
>make scales ergonomic, so changing key shape doesn't really help.
>Also remember that we have limited resources for the constuction of
>this thing; we don't have the benefit of having a corporation (or
>even a team of experienced engineers) develop and manufacture it for
>us. This is all out of our own quite shallow pockets, and whatever
>we can't buy ready-made we have to make from scratch ourselves.
>We're talking low-budget here. Hence the design needs to be simple.

Sounds good!

>> And why would you want that? I've never played on one, so I can't
>> be sure, but I think buttons shaped exactly like piano keys would
>> be better than levers even on conventional keyboards.
>
>I agree, but those would be big-ass buttons, which on a 22-tone
>Bosanquet layout would make tetrads VERY difficult.

No no, I was just saying, even for conventional keyboards, I think
they're doing it all wrong! You should of course use smaller and
differently-shaped keys for 22.

>With our design the playing surfaces of the keys are still a
>square grid-type thing, it's just that the squares are really the
>outer-ends of levers.

Right, I got that. It just seems like a weird choice -- it seems
much harder to build, and it sounds like you agree with me that
levers don't make playing any better.

>Picture something like an old-fashioned non-electric typewriter.
>You won't have the problems typically associated with playing regular
>keyboards since the playable area is equivalent to the very outer
>edge of the keys. That way the action will be similar to a
>traditional keyboard--in fact, we might even be able to weight the
>keys.

For crying out loud, this obsession with "weight" is driving me
nuts. The best piano makers and technicians strive to make keys
feel as weightless as possible. And every other keyboard instrument,
from harpsichords to organs to clavichords to clavinets to well-
regulated electric pianos, don't have weight.

Sorry [deep breath], I review keyboards all day with sluggish actions
that manufacturers and musicians spend a lot of money on, and it
gets to me. :)

>I may eventually post a diagram of the mechanical aspects, but
>university is now in session so my time is significantly lessened.

Good luck! Where do you attend?

-Carl

On Wednesday 02 February 2005 05:15 pm, Carl Lumma wrote:

> For crying out loud, this obsession with "weight" is driving me
> nuts. The best piano makers and technicians strive to make keys
> feel as weightless as possible. And every other keyboard instrument,
> from harpsichords to organs to clavichords to clavinets to well-
> regulated electric pianos, don't have weight.

??? resistence then ??? You don't propose those terrible spring actions are
the zenith of keyboard action?

Maybe we should talk objectively: give an example of a keyboard with an action
you admire. I like weighted, yet light and repetitive piano keys like on a
fine Steinway or Baldwin grand. And by my book, the Yamaha P-200 digital
piano has the best action and sound of any digital piano, bar none, including
Kurzweil. That's wy I bought a Yamaha. Unfortunately, no microtuning tables
or 12-note scales :(

> Sorry [deep breath], I review keyboards all day with sluggish actions
> that manufacturers and musicians spend a lot of money on, and it
> gets to me. :)

Which ones do you dislike?

--
Aaron Krister Johnson
http://www.akjmusic.com
http://www.dividebypi.com

>> For crying out loud, this obsession with "weight" is driving me
>> nuts. The best piano makers and technicians strive to make keys
>> feel as weightless as possible. And every other keyboard
>> instrument, from harpsichords to organs to clavichords to clavinets
>> to well-regulated electric pianos, don't have weight.
>
>??? resistence then ???

?

>You don't propose those terrible spring actions are the zenith
>of keyboard action?

Nope, I sure don't -- they're some of the worst, and I mentioned
one of the reasons why elsewhere on this list today. Note also
that much of what we perceive as "action" is actually "response".
The biggest reason those actions are so bad is that many of their
responses don't do anything to the sound.

>Maybe we should talk objectively: give an example of a keyboard
>with an action you admire. I like weighted, yet light and
>repetitive piano keys like on a fine Steinway or Baldwin grand.

Piano keys are weighted only to counterbalance the weight of the
action parts -- in other words, to make them feel less "weighted".

My favorite keyboard actions ever can be found on any piano at
Faust Harrison pianos, 58th near Broadway, New York. If you get
a full Stanwood job on a good grand you can come close.

It's probably true that our fingers can achieve the finest amounts
of control when pushing against a force -- which all keyboards have
to have if the keys are to come back up. In fact, many "synth"
actions require a breakaway force much higher than that of a fine
piano, even though it may not feel that way.

My position on "weighted" synthesizer actions is that they are
utter crap. The only exception is the semi-weighted action made
by Fatar and now found only on Kurzweil keyboards. The worst is
Roland's fully-weighted action, which is unbelievably slow and
klunky -- the keys actually percuss violently several times in a
cycle.

>And by my book, the Yamaha P-200 digital piano has the best action
>and sound of any digital piano, bar none, including Kurzweil.
>That's why I bought a Yamaha. Unfortunately, no microtuning tables
>or 12-note scales :(

I don't remember playing the P-200 by name, but I suspect it's the
same action they use in their S90, which is middle-of-the-road for
these things. There are at least three actions available on
Kurzweil keyboards; are you sure you tried them all?

-Carl

--- In MakeMicroMusic@yahoogroups.com, "Igliashon Jones"
<igliashon@s...> wrote:
>
> This weekend I'll upload the square-based designs that Stevie and I
> worked with last weekend...when using square keys in straight rows
> some of those ideal-looking designs actually become pretty
hopeless.
> ... All in all, Stevie
> seemed to prefer George's orignal Bosanquet-based design as the
best
> compromise between chords and scales; his fingers are much longer
> than mine so chords that I thought would be unplayable he has no
> problem with (though that's probably just as much due to his 13+
> years of piano training as his finger size). ...
>
> -Igliashon

I've redone the last figure, after making some calculations to
determine the proper offset and angles for a slanted square-key
Bosanquet layout with smaller keys (20 mm on a side, aligned corner-
to-corner by *row* rather than column):

/makemicromusic/files/secor/GKbd22s2.gif

I think this is a bit more pleasing to the eye than my previous
design. Keys for tones 2deg apart have the same x-axis placement,
which makes this layout equally suitable for tunings with wide and
narrow fifths.

After I finished this, I was a bit surprised at how close this comes
to a slanted checkerboard layout (with *both* rows and columns lined
up), which would slightly favor tunings with fifths wider than 700c --
something to consider if one of your constraints is to make the
construction as simple as possible.

Best,

--George

George wrote...

>I've redone the last figure, after making some calculations to
>determine the proper offset and angles for a slanted square-key
>Bosanquet layout with smaller keys (20 mm on a side, aligned corner-
>to-corner by *row* rather than column):
>
>/makemicromusic/files/secor/GKbd22s2.gif
>
>I think this is a bit more pleasing to the eye than my previous
>design. Keys for tones 2deg apart have the same x-axis placement,
>which makes this layout equally suitable for tunings with wide and
>narrow fifths.

Heya, sorry for all the messages aimed at you, George, but did
you make this so that it would display actual-size on your
monitor? If so, can you tell me the size of the keys on your
screen? That way, I can adjust the image to be actual size here...

-Carl

--- In MakeMicroMusic@yahoogroups.com, Carl Lumma <ekin@l...> wrote:
> ...
> Heya, sorry for all the messages aimed at you, George, but did
> you make this so that it would display actual-size on your
> monitor? If so, can you tell me the size of the keys on your
> screen? That way, I can adjust the image to be actual size here...
>
> -Carl

There's a ruler scale in both mm and inches at both the bottom and
left edges.

--George

>> Heya, sorry for all the messages aimed at you, George, but did
>> you make this so that it would display actual-size on your
>> monitor? If so, can you tell me the size of the keys on your
>> screen? That way, I can adjust the image to be actual size here...
>>
>> -Carl
>
>There's a ruler scale in both mm and inches at both the bottom and
>left edges.

D'oh!

-Carl