Help-Midi Kits?
Colonel Mustard
Midi kits for accordions, preferably the new magnetic sensor type.
I would also appreciate any help in finding any net/web articles on how to
fit the kits. I am 'reasonably' competent with a soldering iron and wonder
if the switches are difficult to fit?
Any info would be greatly appreciated.
Thanks
David Batty
<m011...@mcmail.com> writes
the page until you get to "Accordion Newsnet UK", select shops without
internet websites and you will find WEM or Watkins Electric Music.
He is an expert on midi on the accordion.
David Batty
Visit Leyland Accordion Club website at:
http://www.sectorsoftware.demon.co.uk/accord.htm
email: da...@sectorsoftware.demon.co.uk
Dan Lavry
David Batty <da...@sectorsoftware.demon.co.uk> wrote in article
<sGeeGWAs...@sectorsoftware.demon.co.uk>...
> In article <35d48...@news2.mcmail.com>, Colonel Mustard
> <m011...@mcmail.com> writes
> >Can anyone help me. I'm trying to locate a supplier, (U.K. if poss.), of
> >Midi kits for accordions, preferably the new magnetic sensor type.
> >I would also appreciate any help in finding any net/web articles on how
to
> >fit the kits. I am 'reasonably' competent with a soldering iron and
wonder
> >if the switches are difficult to fit?
> >Any info would be greatly appreciated.
> >Thanks
I am no expert. The way I would go about it is to check the distributors of
midi retrofit gear.
I belive Boonton (spelling?) and Master have many models and types.
Accordion International in the US has a few retrofit items as well, and I
am sure that there are more.
Some knowlagble folks have warned me that there are a number of possible
problems to be taken into account. One such problem has to do with the fact
that when you do it yourself, you may be covering too large of the grill
area with a control panel, thus blocking a lot of the accoustic sound. I am
sure that a lot of the accordion gurues can say alot about that. I do not
know if you will get their intrest, but some of the regulars know a lot.
Robert Berta as amazingly knowladgble regarding the midi stuff. I learned a
lot from him and he is a real gentelman as well. I am sure Don and some
others can contribute a lot. I can not take the space here to mention the
names of all my "hidden mentors", but if you get them going, you will learn
a lot very fast.
Dan Lavry
Gary Blair
>Can anyone help me. I'm trying to locate a supplier, (U.K. if poss.), of=
Midi kits for accordions, preferably the new magnetic sensor type.
I would also appreciate any help in finding any net/web articles on how t=
o
fit the kits. I am 'reasonably' competent with a soldering iron and wonde=
r
if the switches are difficult to fit?
Any info would be greatly appreciated.
Thanks<
Geoff Holter is your best bet for magnetic sensor Midi. He can be
contacted at
01325 381223 He is in Darlington, England
Or you can also contact Rolston Accordions of Motherwell , Scotland. 016=
98
265350
You could speak to Asimov electronics 01576 470222. He has quite a variet=
y
of Midi Kits available and can even make one to order ( any spec).
They are in Dumfries, Scotland.
Regards
Gary Blair (Scotland)
Bruce Rodger
breaks, bits missed out.... here's a repost.
Bruce.
>Message text written by Colonel Mustard
>I would also appreciate any help in finding any net/web articles on how to
>fit the kits. I am 'reasonably' competent with a soldering iron and wonder
>if the switches are difficult to fit?
As well as the suppliers Gary mentioned, have a chat with Charlie Watkins
on 0181 679 5575.
I installed the watkins midi kit a few years ago - relatively straightforward,
but don't consider tackling it unless you're fairly confident that you'll be
able to put your box all back together :-)
There's no real electronics skill needed, as the logic circuitry comes
pre-assembled - all you've got to do (!) is fabricate what is
effectively a big matrix switch. You need to know what to do when
confronted with the hot end of a soldering iron , of course, but the
skills required are more akin to those of a model maker...
The kit I installed used mechanical switches - i believe newer one use
magnetic sensors, or optical switching. Can anyone give me more info on
this - is it simply a bank of reed switches and small actuating
magnets, or is it something more sophisticated?
Bruce.
--
Bruce Rodger |Bruce....@strath.ac.uk PGP key available
IT Services |http://www.strath.ac.uk/CC/People/bruce.html
The University of Strathclyde | +44 (0)141 548 3460
Glasgow G4 0LN, Scotland. | Fax 553 4100
Hugh Barwell
>The kit I installed used mechanical switches - i believe newer one use
>magnetic sensors, or optical switching. Can anyone give me more info on
>this - is it simply a bank of reed switches and small actuating
>magnets, or is it something more sophisticated?
I too would love to know how the magnetic systems are arranged. It seems a
very good idea. The contacts are, presumably, completely sealed against
dust?
Hugh Barwell, Nottingham
DoN. Nichols
I have never been inside a MIDI equipped accordion, but I do know
some ways that it could be done. Reed switches are one possiblity, but
eventually they will develop contact problems, unless they are
mercury-wetted reed switches.
What I would consider the most reliable technique would be to use
Hall-effect transducers. Essentially, they are devices which allow you to
measure the strength of a magnetic field. Thus, you could determine not
just the fact that the button or key had moved, but also how fast it had
moved, based on the rate of change of the magnetic field. Some MIDI
devices, I understand, do make use of such information. I don't see it in a
MIDI emulation of the accordion sound, but for emulating a piano's sound,
the rate of change would determine how loud to start the note.
Squeeze On,
DoN.
--
NOTE: spamblocking on against servers which harbor spammers.
Email: <dnic...@d-and-d.com> | Donald Nichols (DoN.)|Voice (703) 938-4564
My Concertina web page: | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---
Bruce Rodger
>Bruce wrote:
>
>>The kit I installed used mechanical switches - i believe newer one use
>>magnetic sensors, or optical switching. Can anyone give me more info on
>>this - is it simply a bank of reed switches and small actuating
>>magnets, or is it something more sophisticated?
>
>
>I too would love to know how the magnetic systems are arranged. It seems a
>very good idea. The contacts are, presumably, completely sealed against
>dust?
>
Reed switches certainly are sealed within a glass envelope - but can
someone confirm or deny that this is the switching technology that
people use? I'm just guessing.....
Gary Blair
>As well as the suppliers Gary mentioned, have a chat with Charlie Watkin=
s =
on 0181 679 5575.
=
I installed the watkins midi kit a few years ago - relatively
straightforward,
but don't consider tackling it unless you're fairly confident that you'll=
be
able to put it all back together :-)
=
There's no real electronics skill needed, as the logic circuitry comes
pre-assem
bled - <>>
How could I have forgotten my old mate Charlie!! ( I hope he will forgive=
me!)
For the folks out there who don't know Charlie, he is one amazing chappy.=
He was the man that invented the W.E.M. copycat! He was also one of the
first Amp makers to supply rigs for stadium work. He toured extensively
with the Rolling Stones during their hey-day. The Accordion world is VERY=
lucky to have him.......He will keep you right on the MIdi side of things=
,
no problem.
BTW don't mention ''Blaze Away'' nor the Uk's finest Accordion ( accordin=
g
to Charlie)
the ''Super Dominator!''
Regards
Gary Blair (scotland)
David Batty
<cra...@ccsun.strath.ac.uk> writes
>ed mechanical switches - i believe newer one use
>magnetic sensors, or optical switching. Can anyone give me more info on
>this - is it simply a bank of reed switches and small actuating
>magnets, or is it something more sophisticated?
>
>
When I opened Charlies own accordion in March the switches were a
springy piece of metal wire connected to the appropriate spot so that
when actuated they simply bent to one side and touched another wire,
simple and efficient.
Bruce Rodger
>In article <6rflv2$l...@d-and-d.com>, Hugh Barwell <hu...@prima.net> wrote:
>
> I have never been inside a MIDI equipped accordion, but I do know
>some ways that it could be done. Reed switches are one possiblity, but
>eventually they will develop contact problems, unless they are
>mercury-wetted reed switches.
Will they? I've just had a look at the spec sheets for some (in the RS catalog)
which say they have a lifetime of around 10^8 (0ne hundred million) operations.
That's a lot of playing..... :-)
The worry I have is that there is a certain amount of "hysteresis" in the
switch - eg the switches will close when the actuating magnet comes within
(say) 6mm, but do not open till the distance exceeds (say) 11mm.
Having said all that, they're not expensive, so I may buy a few and experiment.
>
> What I would consider the most reliable technique would be to use
>Hall-effect transducers. Essentially, they are devices which allow you to
>measure the strength of a magnetic field. Thus, you could determine not
>just the fact that the button or key had moved, but also how fast it had
>moved, based on the rate of change of the magnetic field. Some MIDI
>devices, I understand, do make use of such information. I don't see it in a
>MIDI emulation of the accordion sound, but for emulating a piano's sound,
>the rate of change would determine how loud to start the note.
True. The other devices worth looking at would be optical slot-switches.
Bruce.
>
> Squeeze On,
> DoN.
>
>--
> NOTE: spamblocking on against servers which harbor spammers.
> Email: <dnic...@d-and-d.com> | Donald Nichols (DoN.)|Voice (703) 938-4564
> My Concertina web page: | http://www.d-and-d.com/dnichols/DoN.html
> --- Black Holes are where God is dividing by zero ---
Bruce Rodger
>>
>When I opened Charlies own accordion in March the switches were a
>springy piece of metal wire connected to the appropriate spot so that
>when actuated they simply bent to one side and touched another wire,
>simple and efficient.
Thats exactly how mine is set up - it uses one of Charlies kits - but I
occasionally get a problem with 'double-bounce' on the switches. Probably
due to my inexperience in installing these things - it takes a bit of practice
to find out the optimum length for the springs....
Bruce.
DoN. Nichols
Bruce Rodger <cra...@ccsun.strath.ac.uk> wrote:
>David Batty <da...@sectorsoftware.demon.co.uk> writes:
>>>
>>When I opened Charlies own accordion in March the switches were a
>>springy piece of metal wire connected to the appropriate spot so that
>>when actuated they simply bent to one side and touched another wire,
>>simple and efficient.
>
>Thats exactly how mine is set up - it uses one of Charlies kits - but I
>occasionally get a problem with 'double-bounce' on the switches. Probably
>due to my inexperience in installing these things - it takes a bit of practice
>to find out the optimum length for the springs....
One trick to help with that is to have two parallel springs of
different lengths to the contact points. Each length will bounce at its own
rate, and you are unlikely to have both bounce open at the same time.
(You've got to adjust them so both close at the same position, of course, or
one can bounce before the other gets a chance to do its job.)
DoN. Nichols
Bruce Rodger <cra...@ccsun.strath.ac.uk> wrote:
>Donald Nichols writes:
>>In article <6rflv2$l...@d-and-d.com>, Hugh Barwell <hu...@prima.net> wrote:
>>
>> I have never been inside a MIDI equipped accordion, but I do know
>>some ways that it could be done. Reed switches are one possiblity, but
>>eventually they will develop contact problems, unless they are
>>mercury-wetted reed switches.
>
>Will they? I've just had a look at the spec sheets for some (in the RS catalog)
>which say they have a lifetime of around 10^8 (0ne hundred million) operations.
>
>That's a lot of playing..... :-)
>
>The worry I have is that there is a certain amount of "hysteresis" in the
>switch - eg the switches will close when the actuating magnet comes within
>(say) 6mm, but do not open till the distance exceeds (say) 11mm.
And -- if you have DC current above a certain level through them,
that will generate enough magnetic field to hold them in. Not too likely
that you'll be using that level of current, but it is something to be wary
of.
>Having said all that, they're not expensive, so I may buy a few and experiment.
That's the thing to do.
One idea for the hysterisis is to have the magnet and the reed
switch both mounted in fixed positions, and a steel plate which is
interposed between them attached to the lever. That might give a faster
change in magnetic flux at the switch.
Dan Lavry
DoN. Nichols <dnic...@d-and-d.com> wrote in article
<6rlfri$5...@izalco.d-and-d.com>...
> In article <6rjbsm$13u$1...@rockall.cc.strath.ac.uk>,
> Bruce Rodger <cra...@ccsun.strath.ac.uk> wrote:
> >David Batty <da...@sectorsoftware.demon.co.uk> writes:
> >>>
> >Thats exactly how mine is set up - it uses one of Charlies kits - but I
> >occasionally get a problem with 'double-bounce' on the switches.
Probably
> >due to my inexperience in installing these things - it takes a bit of
practice
> >to find out the optimum length for the springs....
>
> One trick to help with that is to have two parallel springs of
> different lengths to the contact points. Each length will bounce at its
own
> rate, and you are unlikely to have both bounce open at the same time.
> (You've got to adjust them so both close at the same position, of course,
or
> one can bounce before the other gets a chance to do its job.)
It is a common practice in switch based electronic design to overcome
switch bounce with a rather simple "debouncer" circuit, There are various
"flavors" available. In the case of SPDT (single pole double throw) the
circuit recognizes the first contact and latches (holds that information)
an ON state, until the center piece (moving contact) touches the other
pole. At that point an OFF condition is latched.
This works because real SPDT switches do not bounce "all the way" between
the 2 positions. They bounce between a position (contact) and the middle
(no contact - open) until they settle. The same is true in both directions.
Chances are that with so many switches the midi accordion uses an SPST
switch, just an on-off type.
In such cases, one may resort to a time delay scheme, where for a given
time after a contact is first made, or un-made, no more inputs (changes)
are recognized. Such time is typically a few milliseconds (thousand of a
second), depending on the particular switch type. A balance between fast
settling mechanical switch with short electronic delay, and what the ear
can hear (very short note ect.) may yield reasonable results.
One may think this is aseteric but virtually any control switch in
equipment all around us uses debouncing techniques. I would hope the Midi
accordion industry does ignore such common practice. The costs are very
small. In the case of many switches, you typically have the IC (integrated
circuit) run at a very high clock rate (millions or tens of millions per
seconds) and once an input is recognized, it can set the proper time delay
for that switch.
I myself am more intrigued by Hall effect or Photo (light) types. The Hall
effect type could operate under
real adverse conditions including dust. No real contact is needed. No wear
and aging, and little sensitivity to precise positioning. The light
detector type is probably more susceptible to dust covering the lenses, and
proper positioning, but that could be handled. LED's decay in light
intesity over the years, but a good design can give you a robust
performance for 10000 years...
Regarding mercury wetted mechanical switches, I do not have a strong
opinion yet. These switches are typically position sensitive, so you
probably could not play it upside down. I guess this is not a real problem
for most of us....
Dan Lavry
DoN. Nichols
Dan Lavry <danl...@halcyon.com> wrote:
>
>
>DoN. Nichols <dnic...@d-and-d.com> wrote in article
><6rlfri$5...@izalco.d-and-d.com>...
[ ... ]
>It is a common practice in switch based electronic design to overcome
>switch bounce with a rather simple "debouncer" circuit, There are various
>"flavors" available. In the case of SPDT (single pole double throw) the
>circuit recognizes the first contact and latches (holds that information)
>an ON state, until the center piece (moving contact) touches the other
>pole. At that point an OFF condition is latched.
>This works because real SPDT switches do not bounce "all the way" between
>the 2 positions. They bounce between a position (contact) and the middle
>(no contact - open) until they settle. The same is true in both directions.
Agreed. As long as you have SPDT types available.
>Chances are that with so many switches the midi accordion uses an SPST
>switch, just an on-off type.
>In such cases, one may resort to a time delay scheme, where for a given
>time after a contact is first made, or un-made, no more inputs (changes)
>are recognized. Such time is typically a few milliseconds (thousand of a
>second), depending on the particular switch type. A balance between fast
>settling mechanical switch with short electronic delay, and what the ear
>can hear (very short note ect.) may yield reasonable results.
Again, reasonable -- but somewhat difficult as a retrofit, rather
than a part of the original design. Since we have just read about problems
experienced by someone with an already-built instrument, perhaps doubling
the contact springs is the simplest idea likely to be useful.
[ ... ]
>I myself am more intrigued by Hall effect or Photo (light) types. The Hall
>effect type could operate under
>real adverse conditions including dust. No real contact is needed. No wear
>and aging, and little sensitivity to precise positioning.
Plus the ability to get information about key velocity, and/or
absolute key position, should that be desired.
> The light
>detector type is probably more susceptible to dust covering the lenses, and
>proper positioning, but that could be handled. LED's decay in light
>intesity over the years, but a good design can give you a robust
>performance for 10000 years...
Quite adequate, indeed. The worst problem, I suspect, woudl be in a
very smoky environment, such as some night clubs. (Of course, this is a
problem with the human as well. :-)
>Regarding mercury wetted mechanical switches, I do not have a strong
>opinion yet. These switches are typically position sensitive, so you
>probably could not play it upside down. I guess this is not a real problem
>for most of us....
You seem to be thinking of the switches with a pool of mercury,
which is displaced into the contacts. What I am suggesting has a film of
mercury plated on each reed in the encapsulated switch, and no loose pool of
mercury. When the switch reeds get close enough, the mercury film on each
flows to join the other, making excellent contact. Once connected, the
reeds must move farther apart before the film of mercury is stretched enough
to separate. This results in excellent immunity to bounce problems. (But
is of course more expensive, and results in a hazardous waste problem when
the instrument reaches the end of its life.
Dan Lavry
DoN. Nichols <dnic...@d-and-d.com> wrote in article
<6rnavf$7...@izalco.d-and-d.com>...
> In article <01bdcde7$de508400$ee29...@danlavry.halcyon.com>,
> Dan Lavry <danl...@halcyon.com> wrote:
> >
> >
> >DoN. Nichols <dnic...@d-and-d.com> wrote in article
> ><6rlfri$5...@izalco.d-and-d.com>...
>
> Agreed. As long as you have SPDT types available.
I do not know much about accordions or midi accordions, but I am an EE
(electronics engineer) and my area is hardware design. Been doing that for
30 years. The selection of switches is Huge. One can get almost anything
you can dream of...
> Again, reasonable -- but somewhat difficult as a retrofit, rather
> than a part of the original design. Since we have just read about
problems
> experienced by someone with an already-built instrument, perhaps doubling
> the contact springs is the simplest idea likely to be useful.
I am very disappointed to hear that one has experienced it with an already
built instrument. Does anyone know the brand and model of such instrument?
I would call it inadequate design and avoid such midi implementation all
together....
> >I myself am more intrigued by Hall effect or Photo (light) types. The
Hall
> >effect type could operate under
> >real adverse conditions including dust. No real contact is needed. No
wear
> >and aging, and little sensitivity to precise positioning.
>
> Plus the ability to get information about key velocity, and/or
> absolute key position, should that be desired.
Excellent point. I do not know if I would personally wish key velocity
information. I would probably prefer and overall bellows air pressure
transducer of overall volume, but others may wish to play with individual
key velocity (such as for midi piano). For me, having a right hand acoustic
accordion bassoon, plus say, a midi trumpet, and a left hand accordion
stradella, plus say an electric bass or tuba, plus say guitar chords, plus
overall bellow dynamics sounds like a lot of stuff to control already. But
be my guest and use key velocity as well. Pretty soon, you can get rid of
the rest of the band...
One thing that I wonder about: I do not have any real experience wit midi
accordions yet (very soon I will), but I have played with a lot of midi
synth and midi computer software. It seems to me that there are cases where
one may easily "overload" the midi channel. What I mean is that the data
packets gets "bunched up" when a lot of activity happens simultaneously.
The data gets transmitted on a "first come first serve" bases. I never
experienced such problems with a lot of synth as long as I did not use
continues controllers (such as pitch band and after touch). Both pitch band
and after touch are real data hogs, and some midi controllers are more
offensive then others. In fact, examine the various high end midi software
for sequencing and you will find that many offer features such as "thin
aftertouch". The user gets to get rid of say 50% of aftertouch data or pre
decided amount. Then you listen to it and hope it is not too coarse. Horrid
stuff! After a while, I stopped using aftertouch and used the pitch band
very sparingly, and only during un-busy passages.
I think one can play 8 channels of midi real fast, lots of chords etc. and
be OK, as long as the continues controllers are avoided. Key velocity is
not a great offender at all, but it adds a data byte to each played note.
It would probably work fine under normal circumstances.
Your inputs are good and interesting. Thanks.
Dan Lavry
Joe Kesselman, yclept Keshlam
design in a time delay so the circuit ignores momentary breaks
in the connection. This assumes that the delay can be made long
enough to absorb the bounce but short enough to be unnoticable
to the user. Harder with musical instruments than with an ordinary
key or pushbutton, admittedly.
(My degree claims I'm an electrical engineer. Boy, is it confused.)
------------------------------------------------------
Joe Kesselman, http://www.lovesong.com/people/keshlam/
New URL for Walkabout Clearwater Coffeehouse and Chorus:
http://www.lovesong.com/walkabout/
Hugh Barwell
>The technology is so mature, and so I wonder why
>any midi accordion keys yield muliple contacts.
Yes, it's most disappointing. I suffer from it quite a bit, and it's very
noticeable if I am using a guitar or harpsichord sound, or anything with a
hard attack. It also shows up if I play MIDI accordion into Cubase Score to
create a score. Fortunately there is a function -delete doubles- which
removes the re-iterated notes, but that doesn't help me when I am playing,
live.
I looked at making my own MIDI interface for accordion, and found there is a
chip that can do most of the task, E 510, I believe it is. Not being a
professional in electronics, and quite unable to program a microprocessor
myself, it seemed a solution. It is used in my Fatar Midi keyboard. But it
is normally used for velocity sensing applications, which obviously require
changeover contacts. These also provide the debouncing function, as
described in an earlier post in this thread. You can fool it into working
with single normally-open contacts, but then contact bounce may occur.
I wonder if the reason we suffer contact bounce is that the interface
designers are adapting bits of programming from keyboard designs which
normally have changeover contacts, adapting it for single contacts without
adding any alternative means of suppressing or ignoring bounce?
That's just a hypothesis.
Hugh Barwell, Nottingham, UK
-----Original Message-----
From: Dan Lavry <danl...@halcyon.com>
Newsgroups: rec.music.makers.squeezebox
Date: 23 August 1998 02:01
Subject: Re: Help-Midi Kits?
>
>Joe Kesselman, yclept Keshlam <kes...@alum.mit.edu> wrote in article
><6rnmkm$103s$1...@mdnews.btv.ibm.com>...
>> Another standard debouncing technique, purely electronic, is to
>> design in a time delay so the circuit ignores momentary breaks
>> in the connection. This assumes that the delay can be made long
>> enough to absorb the bounce but short enough to be unnoticable
>> to the user. Harder with musical instruments than with an ordinary
>> key or pushbutton, admittedly.
>>
>> (My degree claims I'm an electrical engineer. Boy, is it confused.)
>
>Hi Joe
>
>I guess two EE's in the group had the comments rgarding debounce and time
>delay. It is very standard stuff indeed. Every inexprnsive computer
>keyboard (some cost less then $30 US) have debouncing capabilities. If it
>were not so, then your name could be typed JJJJJJJJJoooooooooeeeeeeee. The
>technology is so mature, and so I wonder why any midi accordion keys yield
>muliple contacts. I belive a 10-20 msec delay is long enough to overcome
>the bounce of many type of switches, but fast enough for the quickest hand.
>Any comments?
>
>Dan Lavry
>------------------------------------------------------------------------
DoN. Nichols
Dan Lavry <danl...@halcyon.com> wrote:
>
>
>DoN. Nichols <dnic...@d-and-d.com> wrote in article
><6rnavf$7...@izalco.d-and-d.com>...
>> In article <01bdcde7$de508400$ee29...@danlavry.halcyon.com>,
>> Dan Lavry <danl...@halcyon.com> wrote:
>> >
>> >
>> >DoN. Nichols <dnic...@d-and-d.com> wrote in article
>> ><6rlfri$5...@izalco.d-and-d.com>...
>>
>> Agreed. As long as you have SPDT types available.
>I do not know much about accordions or midi accordions, but I am an EE
>(electronics engineer) and my area is hardware design. Been doing that for
>30 years. The selection of switches is Huge. One can get almost anything
>you can dream of...
Of course -- but a lot of times the switch actually gets constructed
into the accordion, rather than a pre-made switch being installed. The one
which was beign described somewhere back in this thread is one such, with
the switches being made out of stiff pieces of wire, positioned so they
would touch as the key was depressed, and then bend as necessary to allow
the key full motion. Thus, contact bounce was highly likely.
>> Again, reasonable -- but somewhat difficult as a retrofit, rather
>> than a part of the original design. Since we have just read about
>problems
>> experienced by someone with an already-built instrument, perhaps doubling
>> the contact springs is the simplest idea likely to be useful.
>
>I am very disappointed to hear that one has experienced it with an already
>built instrument. Does anyone know the brand and model of such instrument?
>I would call it inadequate design and avoid such midi implementation all
>together....
By already-built I don't mean factory-built, but just one which, at
this point in time, has been converted (perhaps by the owner, himself), and
the switches are bounce-prone (see above), in combination with no (or
inadequate) de-bouncing circuitry.
>> >I myself am more intrigued by Hall effect or Photo (light) types. The
>Hall
[ ... ]
>> Plus the ability to get information about key velocity, and/or
>> absolute key position, should that be desired.
>Excellent point. I do not know if I would personally wish key velocity
>information. I would probably prefer and overall bellows air pressure
>transducer of overall volume, but others may wish to play with individual
>key velocity (such as for midi piano). For me, having a right hand acoustic
>accordion bassoon, plus say, a midi trumpet, and a left hand accordion
>stradella, plus say an electric bass or tuba, plus say guitar chords, plus
>overall bellow dynamics sounds like a lot of stuff to control already. But
>be my guest and use key velocity as well. Pretty soon, you can get rid of
>the rest of the band...
Well ... I'm not likely to use *any* of these, since a treble
concertina using the English system is too crowded to hold any but the most
custom circuitry. So -- I just speculate. :-)
>One thing that I wonder about: I do not have any real experience wit midi
>accordions yet (very soon I will), but I have played with a lot of midi
>synth and midi computer software. It seems to me that there are cases where
>one may easily "overload" the midi channel. What I mean is that the data
>packets gets "bunched up" when a lot of activity happens simultaneously.
>The data gets transmitted on a "first come first serve" bases. I never
>experienced such problems with a lot of synth as long as I did not use
>continues controllers (such as pitch band and after touch). Both pitch band
>and after touch are real data hogs, and some midi controllers are more
>offensive then others. In fact, examine the various high end midi software
>for sequencing and you will find that many offer features such as "thin
>aftertouch". The user gets to get rid of say 50% of aftertouch data or pre
>decided amount. Then you listen to it and hope it is not too coarse. Horrid
>stuff! After a while, I stopped using aftertouch and used the pitch band
>very sparingly, and only during un-busy passages.
Hmm ... I find myself wondering whether anyone has investigated
using ethernet instead of MIDI for an interface. It could handle a *lot*
more data without getting in trouble. The 10BaseT connections (twisted-pair
ethernet) use modular connectors (RJ-45), like telephone connectors, but 8
conductor instead of the usual 4 (RJ-11) or 6 (RJ-13). That makes for
inexpensive cables, and quick connect/disconnect. The bandwidth is a *lot*
better than MIDI is, and could easily be hooked up to my Sun Workstations,
which don't particularly like normal MIDI connections.
>I think one can play 8 channels of midi real fast, lots of chords etc. and
>be OK, as long as the continues controllers are avoided. Key velocity is
>not a great offender at all, but it adds a data byte to each played note.
>It would probably work fine under normal circumstances.
Two parallel MIDI connections, or the 10BaseT ethernet could do it.
>Your inputs are good and interesting. Thanks.
As are yours,
Thanks as well,
Joe Kesselman, yclept Keshlam
> Hmm ... I find myself wondering whether anyone has investigated
>using ethernet instead of MIDI for an interface.
Interesting question. Main drawbacks:
Ethernet isn't really set up for realtime operation; data transmission can be
delayed by packet collision. (Ethernet retries if the packets are garbled,
MIDI, as far as I can tell, just keeps going on the basis that a late note is
not much better than no note.) Higher bandwidth might offset that, or
might not.
Ethernet takes a lot more hardware to support. MIDI's basically just a standard
(though fast) serial port, and a UART is a much cheaper chip than an Ethernet
controller.
DoN. Nichols
Joe Kesselman, yclept Keshlam <kes...@alum.mit.edu> wrote:
>In <6rpv2o$a...@izalco.d-and-d.com>, dnic...@d-and-d.com (DoN. Nichols) writes:
>>using ethernet instead of MIDI for an interface.
>
>
>Ethernet isn't really set up for realtime operation; data transmission can be
>delayed by packet collision. (Ethernet retries if the packets are garbled,
>MIDI, as far as I can tell, just keeps going on the basis that a late note is
>not much better than no note.) Higher bandwidth might offset that, or
>might not.
Given that the majority of data will be flowing from the controller
to the various midi sound generators, there should be very little
collision. The *much* higher data rate will also allow the retries to occur
quickly enough so there is unlikely to be enough dealy to be a problem. The
packets could be designed to bear a timestamp. and one could be discarded by
the receiving end, if too late.
Yes -- in congested networks the delay can sometimes be significant,
since each subsequent retry is delayed by a multiple of the original retry
delay (all modified by a random number generator, to minimize the chance of
each retry at each end being the same delay, and colliding again.
>Ethernet takes a lot more hardware to support. MIDI's basically just a standard
>(though fast) serial port, and a UART is a much cheaper chip than an Ethernet
>controller.
Granted. But for a high-end system, where there are more data
packets going out (because of all the options being turned on), it might be
worthwhile. (Say, if someone is trying to synthesize a whole orchestra from
a single accordion controller -- with pitch-bend. :-)
Listening to MIDI can be an interesting experience -- such as when
you hear an entire choir pause to find the next note -- all at once. :-)
Enjoy,
Hans Palm
>
> In <6rpv2o$a...@izalco.d-and-d.com>, dnic...@d-and-d.com (DoN. Nichols) writes:
> > Hmm ... I find myself wondering whether anyone has investigated
> >using ethernet instead of MIDI for an interface.
>
> Interesting question. Main drawbacks:
>
> Ethernet isn't really set up for realtime operation; data transmission can be
> delayed by packet collision. (Ethernet retries if the packets are garbled,
> MIDI, as far as I can tell, just keeps going on the basis that a late note is
> not much better than no note.) Higher bandwidth might offset that, or
> might not.
A couple of collisions wouldn't matter on a 10 or 100 Mbit/s connection.
MIDI is actually only 38.5 kbit/s, causing delay effects if you've got
a lot of info to transfer.
> Ethernet takes a lot more hardware to support. MIDI's basically just a standard
> (though fast) serial port, and a UART is a much cheaper chip than an Ethernet
> controller.
Yes, the hardware would probably be too expensive.
A much better solution would be to use the new USB interfaces that all
new
PC's are equipped with. You have 12 Mbit/s, "hot-swapping" and even
enough
power to supply the circuitry inside the accordion.
With a USB accordion, you don't have to switch off the PC when you're
connecting the cable. You don't need batteries or a separate power
source. The cable can't be longer than 5 meters but I'm shure there's
a solution for that too. You'll need a device driver to convert from
USB to your sound card but that's pretty straight-forward.
There are already USB to MIDI interface being developed but why use MIDI
instead of USB directly. Check out http://www.usb.org/ to find out
more about USB.
So, in a couple of years, will we be discussing USB accordions instead?
The problems with the key switches remains. What is the best
(simple/cheap)
solution?
Mechanical - No!
Magnetic (Hall effect, reed or other) - Maybe.
Optical - Maybe.
The last couple of mmonth's, I've been trying to find a cheap reedless
accordion. Most companies I've tried contacting have not been very
interested
in answering my questions. I've tried email (most companies never reply)
and
fax (somewhat better). I've found a couple of models but they're all
very expensive.
I've found the Cavagnolo Odyssee, the Petosa Millennium, a couple of
models
from Master and a few others. And there's the Concerto DA-100 from
Accordions International (also sold by master) but I don't need the
reeds.
OK they may be removed, but the bellows dynamics wouldn't work properly.
And they don't sell it without the reed blocks!
I don't know if I'm able to afford one of those so I haven't given up
building one myself. The microcontroller or whatever wouldn't be
too easy to design. A pressure sensor could be fitted to get bellows
dynamics. I'm able to solve most problems except for the key swithes.
I want something cheap, reliable that's easy to fit into most
accordions.
Any ideas?
Hans Palm
http://accordion.simplenet.com
Hugh Barwell
>The problems with the key switches remains. What is the best
>(simple/cheap)
>solution?
>
>Mechanical - No!
>Magnetic (Hall effect, reed or other) - Maybe.
>Optical - Maybe.
There is one technology that hasn't been mentioned. I don't know how it
works exactly, but I believe Yamaha, Bosendorfer and other top end, hi tech
piano manufacturers use some sort of light beam system to scan piano actions
to detect notes being played, and translate this into MIDI. I understand
this uses no contacts whatsoever: it just sort of looks at the hammers!
It's not optical in the normal sense of having an opto device for each note.
Hugh Barwell, Nottingham, UK
-----Original Message-----
From: Hans Palm <hans...@mailbox.swipnet.se>
Date: 24 August 1998 20:55
Subject: Re: Key switch technology (Was: Re: Help-Midi Kits?)
Dan Lavry
Hans Palm <hans...@mailbox.swipnet.se> wrote in article
<6rsfjd$1...@d-and-d.com>...
> Joe Kesselman, yclept Keshlam wrote:
> MIDI is actually only 38.5 kbit/s, causing delay effects if you've got
> a lot of info to transfer.
That was my "guess also, but not kbits/s instead Kbytes/s The basic clock
is 1MHz (last I looked, and it used to be 500KHz in the early days). With 8
bits per byte you have 125KHz. Now, most normal actions take 2 bytes (such
as note on + which note, note off + which note ect...) so you are at about
62.5KHz. True, if you send multiple channels of data in an interleaved
fashion, such as base followed by a chord followed by a note on the treble,
the midi inserts a channel destination byte and now you have 3 bytes per
"note" and you are approaching 40KHz. Ad some more features and you are
even worse (slower). That is why I think that after touch and pitch bend
are out of place, and perhaps individual key velocity as well.
> So, in a couple of years, will we be discussing USB accordions instead?
Or fire wire?
> The problems with the key switches remains. What is the best
> (simple/cheap)
> solution?
>
> Mechanical - No!
> Magnetic (Hall effect, reed or other) - Maybe.
> Optical - Maybe.
Agree. You got to choose something. I like the last two as well.
> The last couple of mmonth's, I've been trying to find a cheap reedless
> accordion. Most companies I've tried contacting have not been very
> interested
> fax (somewhat better). I've found a couple of models but they're all
> very expensive.
Yes they are
> I've found the Cavagnolo Odyssee, the Petosa Millennium, a couple of
> models
> from Master and a few others. And there's the Concerto DA-100 from
> Accordions International (also sold by master) but I don't need the
> reeds.
> OK they may be removed, but the bellows dynamics wouldn't work properly.
> And they don't sell it without the reed blocks!
> Any ideas?
Hans, I may be wrong, but I thought that the Cavagnolo and Petosa you
mentioned are reedless.
I am sure the Petosa is (I tried it at their Seattle store).
Speaking of switches, Do you or anyone know which manufacturers use
mechanical, which use optical and magnetic?
I know the list of manufacturers of midi accordions is long, but would be
good to know who is doing what.
I am aware of Petosa, Cavagnolo, Concerto DA, Monarch, Bell-Bugari,
Excelsior, Iorio, and Alessandrini
(new name for me), Then come the Midi kits such as Masters..... but I could
not tell the switch types on more then one or two of these brands.
Best regards
Dan Lavry
Hans Palm
>
> > MIDI is actually only 38.5 kbit/s, causing delay effects if you've got
> > a lot of info to transfer.
> That was my "guess also, but not kbits/s instead Kbytes/s The basic clock
> is 1MHz (last I looked, and it used to be 500KHz in the early days). With 8
> bits per byte you have 125KHz. Now, most normal actions take 2 bytes (such
> as note on + which note, note off + which note ect...) so you are at about
> 62.5KHz. True, if you send multiple channels of data in an interleaved
We're both wrong actually since it's 31.5 kBaud = 31.5 kbits/s.
Since there's a start bit, eight data bits and a stop bit, the
effective byte rate is only 3150 bytes/s. Most messages
(Note On, Note Off) use three bytes (message/channel, e.g. pitch number,
e.g. velocity). So a single key down/up includes 6 bytes and will take
about 2 ms to transfer.
> > So, in a couple of years, will we be discussing USB accordions instead?
> Or fire wire?
Why not... The MIDI 1.0 specification is from October 1983 and
a lot has happened since then. The protocol may still be used
but a higher transfer rate wouldn't hurt...
> > The problems with the key switches remains. What is the best
> > (simple/cheap)
> > solution?
> >
> > Mechanical - No!
> > Magnetic (Hall effect, reed or other) - Maybe.
> > Optical - Maybe.
> Agree. You got to choose something. I like the last two as well.
>
> > The last couple of mmonth's, I've been trying to find a cheap reedless
> > accordion. Most companies I've tried contacting have not been very
> > interested
> > fax (somewhat better). I've found a couple of models but they're all
> > very expensive.
> Yes they are
>
> > I've found the Cavagnolo Odyssee, the Petosa Millennium, a couple of
> > models
> > from Master and a few others. And there's the Concerto DA-100 from
> > Accordions International (also sold by master) but I don't need the
> > reeds.
> > OK they may be removed, but the bellows dynamics wouldn't work properly.
> > And they don't sell it without the reed blocks!
> > Any ideas?
>
> Hans, I may be wrong, but I thought that the Cavagnolo and Petosa you
> mentioned are reedless.
Well, they are, except for the Concerto (also sold by Master),
that's actually a MIDI accordion with reeds. I've been in contact
with the designer and it's not designed to be used without the
reed blocks. You may take them out to save weight but the air
will flow in an uncontrolled manner and the bellows dynamics
will not work properly. I asked him if I could buy it without
the reed blocks to save a couple of dollars but he said no.
> I am sure the Petosa is (I tried it at their Seattle store).
>
> Speaking of switches, Do you or anyone know which manufacturers use
> mechanical, which use optical and magnetic?
Well, the Concerto uses optical. With Master it depends on how
old the MIDI interface is. I think that the latest models are optical.
> I know the list of manufacturers of midi accordions is long, but would be
> good to know who is doing what.
> I am aware of Petosa, Cavagnolo, Concerto DA, Monarch, Bell-Bugari,
> Excelsior, Iorio, and Alessandrini
According to Master, most most Italian accordion manufacturers install
the Master MIDI.
Dan Lavry
Hans Palm <hans...@mailbox.swipnet.se> wrote in article
<6rv6th$d...@d-and-d.com>...
> Dan Lavry wrote:
> > Hans Palm <hans...@mailbox.swipnet.se> wrote in article
> > > Joe Kesselman, yclept Keshlam wrote:
> We're both wrong actually since it's 31.5 kBaud = 31.5 kbits/s.
> Since there's a start bit, eight data bits and a stop bit, the
> effective byte rate is only 3150 bytes/s. Most messages
> (Note On, Note Off) use three bytes (message/channel, e.g. pitch number,
> e.g. velocity). So a single key down/up includes 6 bytes and will take
> about 2 ms to transfer.
I would hope that the individual note velocity byte be avoided, and all
notes on all channels may have a volume control data inputted via the
overall bellow dynamics pressure sensor. But in general, I will not "split
hair" with you, and take you word regarding 2mS. That delay does not sound
like much to me. Even playing 6-8 simultaneous notes (events) is probably
within reason. I would think that going for more "bells and whistles" may
start causing a lot of problems.
> Why not... The MIDI 1.0 specification is from October 1983 and
> a lot has happened since then. The protocol may still be used
> but a higher transfer rate wouldn't hurt...
I do not know the limiting factor today. Cost has always been an issue
here. A few years ago the limitation was the opto-coupler speed. The most
commonly used device was a Sharp device (if I recall it was P900 or P9000).
One can always speed things way beyond 1Mbits/sec, but the Midi standards
committee would always try to be inclusive of all manufacturers including
the very low cost stuff where every cent counts. The $50 low cost midi
compatible keyboard is part of the reason why we are speed deprived.
> Well, the Concerto uses optical. With Master it depends on how
> old the MIDI interface is. I think that the latest models are optical.
So who is using magnetic (Hall effect?)
I saw something on the web: Limex Electronics. I believe they are selling
midi retrofit they refer to as contactless. I wish I could offer more.
Once again, I heard a number of experts say that "doing your own" is fine
if you want to have an "all midi" and perhaps that is what you wish to
have. But it was stated that one needs to be careful when retrofitting an
instrument with reeds, mostly because of blocking the air passage (grill)
with a control panel. I would assume that any self respecting manufacturer
would integrate the midi with plenty of thought as to the acoustic impact.
If all you want is a midi accordion (reedless), then maybe a good
compromise would be to retrofit a ready made controller (such as Master)
into an inexpensive accordion. I would hate to buy a Concerto (I thing it
is a Piermaria) for a lot of money(!!!) even without the reeds, because you
are paying for good wood work that serves no purpose, and for bellows that
are too good for what you will need.
Best wishes
Dan Lavry
Loren Schaeffler
system. The key scanning sequence takes time also. The keyboards or
modules take alot more time than this to actually sound a tone after it
is received. I had alot of information about this when I started
designing midi systems for chemnitzer concertina's 8 years ago. I
believe the tone generator delay is about 8 or 10 milliseconds. The
total delay is about 15 milliseconds which is enough to be heard. <This
is not a problem> because a reed also takes time to make a sound after a
key is played. I actually slowed my system down. I don't recall what the
actual time lapse is from when a key is pressed until a note is heard
because that isn't important. What is more important is that the sound
from the tone generator, keyboard, or module start sounding at about the
same time as the reeds. An accordian or concertina player is accustomed
to sound delay because of the delay in the action of the reed. My first
midi system was so fast that the sound of a piano, for example, started
noticably sooner than the sound of the reed which, in my opinion, made
it harder to play because the electronic and reed sounds started at
different times.
I'm still using contacts in my systems simply because they are easier to
install. I reduce or eliminate contact bounce by software in the control
unit.
Loren.....
--
New Website Address
http://prairie.lakes.com/~lorens
Chemnitzer Concertina Information
Peter S. Lupica
useless information. Baud rate does not necessarily equal bit rate. In
fact, they are two different things.
Hans Palm <hans...@mailbox.swipnet.se> wrote in article
<6rv6th$d...@d-and-d.com>...
> Dan Lavry wrote:
> > Hans Palm <hans...@mailbox.swipnet.se> wrote in article
> > > Joe Kesselman, yclept Keshlam wrote:
> >
> > > MIDI is actually only 38.5 kbit/s, causing delay effects if you've
got
> > > a lot of info to transfer.
> > That was my "guess also, but not kbits/s instead Kbytes/s The basic
clock
> > is 1MHz (last I looked, and it used to be 500KHz in the early days).
With 8
> > bits per byte you have 125KHz. Now, most normal actions take 2 bytes
(such
> > as note on + which note, note off + which note ect...) so you are at
about
> > 62.5KHz. True, if you send multiple channels of data in an interleaved
>
> We're both wrong actually since it's 31.5 kBaud = 31.5 kbits/s.
> Since there's a start bit, eight data bits and a stop bit, the
> effective byte rate is only 3150 bytes/s. Most messages
> (Note On, Note Off) use three bytes (message/channel, e.g. pitch number,
> e.g. velocity). So a single key down/up includes 6 bytes and will take
> about 2 ms to transfer.
>
> > > So, in a couple of years, will we be discussing USB accordions
instead?
> > Or fire wire?
>
> Why not... The MIDI 1.0 specification is from October 1983 and
> a lot has happened since then. The protocol may still be used
> but a higher transfer rate wouldn't hurt...
>
> Well, the Concerto uses optical. With Master it depends on how
> old the MIDI interface is. I think that the latest models are optical.
>
Dan Lavry
lot. A tuba takes a lot of time to "build" sound, and on the other side of
the spectrum there are drums that respond quickly. Part of that could be a
problem, and part of it provides "real world charm". A brass bend is never
a perfect thing from time delay stand point. Of course too much added delay
is not good.
My concern is more along the line that having too much simultaneous data,
given the bandwidth limitations of a system creates a bottle neck at busy
time, and the notes come out "one at a time" (first played first to sound).
Of course this is the case here anyway. You play a chord, and the first
note to contact plays first, the second note needs to wait for the first to
be completed and so on. As long as the outcome is fast enough (the notes
begin almost at the sane time), everyone is happy. If each note takes too
long and too many are played at once you may end with too much time between
notes played simultaneously. More midi activity and data may slow the
process further.
Hans Palm
>
> Interesting. One very minor point that may fall under the category of
> useless information. Baud rate does not necessarily equal bit rate. In
> fact, they are two different things.
Yes, I know, but it's equal in the MIDI standard.
Hans Palm
Sweden
http://accordion.simplenet.com
Hans Palm
> I saw something on the web: Limex Electronics. I believe they are selling
> midi retrofit they refer to as contactless. I wish I could offer more.
Yes, I've been there too. I've tried contacting them without any luck...
> Once again, I heard a number of experts say that "doing your own" is fine
> if you want to have an "all midi" and perhaps that is what you wish to
> have. But it was stated that one needs to be careful when retrofitting an
> instrument with reeds, mostly because of blocking the air passage (grill)
> with a control panel. I would assume that any self respecting manufacture
> would integrate the midi with plenty of thought as to the acoustic impact.
> If all you want is a midi accordion (reedless), then maybe a good
> compromise would be to retrofit a ready made controller (such as Master)
> into an inexpensive accordion. I would hate to buy a Concerto (I thing it
> is a Piermaria) for a lot of money(!!!) even without the reeds, because you
> are paying for good wood work that serves no purpose, and for bellows that
> are too good for what you will need.
A control panel isn't necessary. There are accordion MIDI interfaces
without
a control panel. Since in most cases you're hooked up to a sound module,
you may either use the control panel on that or connect foot pedals for
changing sounds or whatever.
You may also use combinations of keys on the keyboard to control the
MIDI
interface inside your accordion. The key combinations must of course be
very unlikely to occur when you're playing a tune...
The most common accordion MIDI interface here in Sweden is the Bonotron.
The AM-2 model is such a model without a control panel. Check it out
at http://www.bono.se/acc_uk.htm
DoN. Nichols
Peter S. Lupica <LAI...@worldnet.att.net> wrote:
>Interesting. One very minor point that may fall under the category of
>useless information. Baud rate does not necessarily equal bit rate. In
>fact, they are two different things.
>
They are different things *if* you are using parallel data paths, or
going through some complex system such as a modern modem.
For a simple serial port, they are identical. A MIDI port
qualifies as a simple serial port for these purposes.
Squeeze On,
Joe Kesselman, yclept Keshlam
someone else's post, please try to trim the quote down a bit to
to make it easier to find your new words. (And to save bandwidth.)
Joe Kesselman, yclept Keshlam
>Baud rate does not necessarily equal bit rate.
True, but the inequality only occurs when you have more than one bit being
sent at a time). Important point for modern modems, which are more complex
than a simple binary signalling system, but generally doesn't apply to straight
port-to-port connections.
Joe Kesselman, yclept Keshlam
>A control panel isn't necessary.
Good point.
>You may also use combinations of keys on the keyboard
My plan on the MIDI'tina is to add a thumb button or two that
would act as "control keys", and use these alone or in
conjunction with the normal buttons. For example, transpose
might be "hold control and tap the note that should become
the new tonic".
This does mean you can't play and enter commands simultaneously,
or at least don't have _all_ the notes available while entering commands.
There are several ways that might be overcome... I expect I'll decide
by experimentation.
MADAVB
know if any box manufacturer has picked up on this.
Also, I agree that the new USB would be better than midi. The new iMAC uses it
and is totally self contained and would make a great carry along computer.
Tony Bertolino
Dan Lavry
Hugh Barwell <hu...@prima.net> wrote in article <6rst83$5...@d-and-d.com>...
> There is one technology that hasn't been mentioned. I don't know how it
> works exactly, but I believe Yamaha, Bosendorfer and other top end, hi
tech
> piano manufacturers use some sort of light beam system to scan piano
actions
> to detect notes being played, and translate this into MIDI. I understand
> this uses no contacts whatsoever: it just sort of looks at the hammers!
> It's not optical in the normal sense of having an opto device for each
note.
I heard of the Bosendorfer. The complete concept is "reversed". While midi
"takes" the key strokes, converts it to data bytes and sends it to a sound
generating module, the Bosendorfer takes the data bytes and sends it to a
mechanical activator to actually play the real acoustic instrument. I
assume that their system would record the motions of the keys in order to
play it back, but the key feature is to have the real piano, and
Bosendorfer is a real piano indeed (probably the best), play and reproduce
the sound acoustically (no microphones, speakers, electronics or acoustic
coloration). If you could invite a great pianist to play the instrument,
you could have an almost real performance 100 years from now.
Their system is not like the old player piano, where each note was played
at the same volume and duration as defined by a punched piano roll. Each
note is recorded with precise velocity. This is doable with a piano, since
the player can control only the velocity. It would be a lot more difficult
with a violin (less notes but so much subtle motion requiring a lot of
parameters to be recorded).
I think their system has very little to do with a midi accordion.
Best regards to all
Dan Lavry
Ron Natalie
> I heard of the Bosendorfer. The complete concept is "reversed". While midi
> "takes" the key strokes, converts it to data bytes and sends it to a sound
> generating module, the Bosendorfer takes the data bytes and sends it to a
> mechanical activator to actually play the real acoustic instrument.
MIDI works both ways. That's the whole point. Some units send
data, some receive. It's not much fun otherwise.
There are now MIDI grand pianos (both input and output) in the
consumer price range. The Bosendorfer thing was done a long
time ago (at least 12 years) and I don't think was midi (I
attended a talk at a UNIX conference called /dev/bosendorfer).
I
> assume that their system would record the motions of the keys in order to
> play it back, but the key feature is to have the real piano, and
> Bosendorfer is a real piano indeed (probably the best), play and reproduce
> the sound acoustically (no microphones, speakers, electronics or acoustic
> coloration). If you could invite a great pianist to play the instrument,
> you could have an almost real performance 100 years from now.
This is actually being done. You can order data recorded by
famous pianists like Oscar Peterson, etc...
>
> Their system is not like the old player piano, where each note was played
> at the same volume and duration as defined by a punched piano roll. Each
> note is recorded with precise velocity.
Actually, variable durange is available on a standard player piano.
There was a device called a reproducing piano. It had two sets
of 4 bit force encodings on the outside of the 88 key puncings.
Not the per note coloration of MIDI, but it was an improvement.
To hear something interesting, there was a reproducer featuring
Gershwin playing American In Paris. This was (back in the late
70's) restored (removing some added orchestrations added to the
roll) and played opposite an orchestra, giving what is the best
chance of hearing Gerswin playing it today. The record is
available on Columbia (I believe Micheal Tilson Thomas conducted
the orchestra, if any body needs a more precise reference, I'll
dig it up when I get home).
> I think their system has very little to do with a midi accordion.
>
A midi accordion is essentially a keyboard controller (output
only device). While there probably is some subtle attack issues
for the purist, most of the per note coloration is done by the
bellows, so the keyboard scanning can be fairly crude.
Dan Lavry
<35E6C18A...@sensor.com>...
> Dan Lavry wrote:
>
> > I heard of the Bosendorfer. The complete concept is "reversed". While
midi
> > "takes" the key strokes, converts it to data bytes and sends it to a
sound
> > generating module, the Bosendorfer takes the data bytes and sends it to
a
> > mechanical activator to actually play the real acoustic instrument.
>
> MIDI works both ways. That's the whole point. Some units send
> data, some receive. It's not much fun otherwise.
The difference a very large here. MIDI does regeister or record keystrokes
in one direction and converts such strokes (data) to sound. But the sound
is generated by electronics means, typically a synth module. While the
sound quality depends on the synth itself, there is a lot of electronics
involved, and the sound ends up at a speaker system. The Bosendorfer
actually playes the keys (the hammers hit the piano strings of the real
acoustic instrument. The sound comes directly out of the acoustic
instrument.
> This is actually being done. You can order data recorded by
> famous pianists like Oscar Peterson, etc...
That is the idea.
> Actually, variable durange is available on a standard player piano.
> There was a device called a reproducing piano. It had two sets
> of 4 bit force encodings on the outside of the 88 key puncings.
> Not the per note coloration of MIDI, but it was an improvement.
> To hear something interesting, there was a reproducer featuring
> Gershwin playing American In Paris. This was (back in the late
> 70's) restored (removing some added orchestrations added to the
> roll) and played opposite an orchestra, giving what is the best
> chance of hearing Gerswin playing it today. The record is
> available on Columbia (I believe Micheal Tilson Thomas conducted
> the orchestra, if any body needs a more precise reference, I'll
> dig it up when I get home).
Interseting. I did not know that.
Dan Lavry
Robert Berta
block the reeds. It is a short PC board that is about 1/2" wide and as long as
the keyboard. It uses I believe Hall Effect for the pickups. Over the weekend I
saw a mockup that had the board located immediately between the treble keyboard
and under the switch mechanism and parrallel to the tops of the keys. In other
words it would have absolutely no effect that I could see since it was no where
near the reeds or grill openings. But note this does not include the controller
so if you need that I would investigate how large it is. The problem with
Master, Limex and a couple of other designs is that to do the control changes
you need to stop playing, hit a button on the bass side to enact the controller
and than use combinations of various treble keys to do the control changes. I
have never cared for this system and think it would be a worthwhile option to
utilize a foot controller to do the changes instead.
Frankly I much prefer the designs where the controller design is on the grill
of the accordion so that you can play and change the program seamlessley (once
proficient). I own a Iorio Midi II+ which has some custom burned chips for
additional features and also have access to Concerto Accordions. Both of these
are much more user friendly than other designs I have used.
As to the effect of blocking or changing the sound. A correctly designed NEW
MIDI accordion has this allowed for in placement of the grill cutouts, circuit
boards, etc. I have a top of the line Bugari dual tone chamber accordion that
does not have MIDI. In comparison to both my Iorio dual tone chamber MIDI and
the Concerto dual tone chamber models both are LOUDER than the Bugari for the
same pressure. Sound wise the Piermaria, which is the accordion used in both
Concerto and Iorios is very nice and quite warm. On the tone chamber models the
single bassoon and clarinet reeds are really nice and of a warmer character
more like the Bugari compared to say a Petosa which to my ear is cooler and
harsher in tone.
Another advantage of a MIDI factory built accordion compared to a retrofit is
that in some cases and designs a midi pickup system will slightly effect the
touch and spring pressure of the keys. In a new complete MIDI unit this will be
compensated for in the final design so the correct touch is fine. On some
converted units I have tried, the springs are too strong and really need to be
compensated to allow for either the extra tension or mass of the pickups. The
Limex units seem to be very light and small and probably would be the best
choice to minimize this touch change.
As far as reliability of spring bar vs. Hall effect vs. magnetic pickup vs.
optical sensors.... All have their good and bad features but the simplistic
design of the spring bar is probably the most reliable. I have seen sensors
come loose, magnets fall off and strong light sources change programs and
settings on optical sensor MIDIs. Spring bars have been the method of choice in
literally millions of keyboard synths over the years and their reliability is
proven. About the only maintenance is an occasion squirt of Radio Tuner cleaner
and lube once a year or so on the buss bar and springs. Some of the other
versions are utilized because they allow for some measure of touch sensitivity
but I have yet to try ANY brand that was even close to adequate. The problem is
that an accordion only has about a 3/8" stroke comared to the huge stroke of a
real piano. The ability to feather this difference to approximate the feeling
of a piano is marginal at best on a accordion. And in any case touch
sensitivity is only useful for a few select instruments sounds....generally you
don't want this feature on 99% of the sounds available.
Bob Berta