#############
# The following is 'midimerge' from the midi-archive.
#############

This is a collection of MMML messages concerning MIDI merging.
It's sorted chronologically, mostly.
===============================================================
>From uucp Mon Feb 13 14:15 EST 1989
>From rft Mon Feb 13 13:20 EST 1989 remote from cblpe

To: All MMMLers
Subject: Merging MIDI sources together

I need to merge two (possibly more) MIDI out signals together.
Any recommendations?

Rick Taylor
cblpe!rft
CB 1D343
614-860-2006

>From uucp Mon Feb 13 16:08 EST 1989
>From wbf Mon Feb 13 16:04 EST 1989 remote from cbema
Subject: MIDI Merging

Rick,

	Come see my Digital Music MX-8 to see if it is what you want.  If
you don't really need merging, a swith box is all you need.  These are less
expensive to buy and you can build your own if you like.

Bill Fox	cbema!wbf


>From tjt Mon Feb 13 23:32 EST 1989
To: midi
Subject: Re:  MIDI Merging

>From mosc  Mon Feb 13 22:41:17 1989 remote from aloft
From: aloft!mosc (52232-H. S. Moscovitz)

I have read some comments on midi-mergers on the MMML.  I will share
my experience on the subject.  Understanding these devices is fairly 
simple, not quite as tricky as microcode programming a floating point 
DSP chip.

I have the J. L. Cooper MSB Plus midi-merge/router/filter/transposer
box.  It provides 8X8 midi routing and switching, and id can merge any
two inputs to any of 8 outputs.  It is very programmable and stores
about 64 settings (I use only two).  It has two independent midi
processors that can do channel bumping, transposition and midi
filtering.  You can filter note, pitch bend, controller, after-touch,
program changes, real time, and/or systems exclusive/systems common
commands.

The MSB is hard to beat in the user friendliness department.  This unit
is sure to please any macho-hard-core-techno electronic musician.  You
won't find any whimpy graphical liquid crystal displays on this
device.  Here's the scoop on one of the refreshing highlights of the user
interface directly from the owners manual:  "The program memory and
display of the MSB Plus use a modified Octal from, rather than the
normal decimal due to the 8 select push buttons.  That is, instead of
showing (and selecting) patches 1 thru 64, you will use 11 thru 88.  In
this system, for example, patch 21 is directly after 18."  After
careful study, I have determined that the modified octal system is
very much like the conventional octal system, except that the 0 digit
is not used, but 8 is.  Pretty nice huh?  But wait, there's more:
For the sake of consistency, the MSB Plus uses the modified hexidecimal 
number system for representing midi channels.  I am pleased to see the 
manufacturer boldly offer a novel new number system where 'F'==15, 
and '0'==16.

An unexpected feature is the red panic button.  It is a momentary
on/off switch that generates a very complex sequence of midi events to
intended to turn off all stuck-on notes in your midi network.  It
starts off with a simple midi "all notes off" message and then sends a
complex sequence of midi events on all channels for five seconds, or
until you release the button.  This sequence is a paranoid midi
musician's dream.

When I first got the unit home I was curious, so I began setting up
various pathological midi hook-ups to see if I could find the panic
button's limitations.  The MSB's flexible routing capabilities made it
easy to set up variations on infinite midi loops, even with automatic
transposition!  The waves of sound generated by my synthesizers were
electrifying.  (I'm sure Jimi Hendrix would have switched to keyboards
had he heard this.)  The MSB has shown me there are vast new musical
horizons. The panic button does its assigned job flawlessly.

					Howard Moscovitz
					att!aloft!mosc

>From uucp Tue Feb 14 12:12 EST 1989
>From gjm Tue Feb 14 11:50:10 EST 1989 remote from coma
Subject: Re: JLCooper MSB+

I can second Howard Moscovitz's recommendation of the MSB+.  I have found
it essential for coordinating many synths and dealing with various pieces
of software (different provisions for echo, etc.).  If you have two devices
that are the same (e.g. TX81Z's, or TF1 modules in a TX rack), it is useful
for configuring them separately (e.g. separating them out of lock-step with
the same system channel to different system channels).

I regularly use the midi merge capability, but have experienced data loss
under heavy load (sequencer + keyboard), so be forwarned.  The KX88 merges
its input to its output port -- this is the most common merging that I want
(sequencer + KX88 keyboard) and it seems to be able to handle the merge
with fewer problems than the MSB+.  Both can run into problems with large
data dumps (an obvious midi merge problem).

How well do some of the other midi switch/merge boxes behave under heavy
load or sysex dump conditions?

-Gary

P.S. Sorry about the duplicate article (Research mailer vs. BSD Mail confusion).

>From tjt Tue Feb 14 12:38 EST 1989
To: midi
Subject: Re: JLCooper MSB+

> >From gjm Tue Feb 14 11:50:10 EST 1989 remote from coma
> I regularly use the midi merge capability, but have experienced data loss
> under heavy load (sequencer + keyboard), so be forwarned.  The KX88 merges
> its input to its output port .... and it seems to be able to handle the merge
> with fewer problems than the MSB+.  Both can run into problems with large
> data dumps (an obvious midi merge problem).

The reaction of the KX88 to large data dumps is pretty amusing - on mine,
at least, it just goes completely bananas, LED's blinking wildly, and refuses
to do anything unless you power cycle it.  Does the MSB+ react in the
same catastrophic way?    ...Tim...

>From tjt Tue Feb 14 13:30 EST 1989
To: midi
Subject: re: COoPer OctAl pAnic

>From mosc  Tue Feb 14 13:13:54 1989 remote from aloft
From: aloft!mosc (52232-H. S. Moscovitz)

In response to Andy McDonough: 

    do they charge money for this?  are they available?

Yes one can purchase a Cooper MPU Plus for somewhere around $350 (if 
my memory serves me well, which it usually doen't).  It really is a
very nice box that makes it much easier to manage a midi network with
several possible controllers and many sound generators.  It is really
overkill if all that is needed is a midi merger.   

I have been told that nobody makes a merger that can mix more than
two midi inputs.  Does anybody know for sure the correctness of this
statement?

>From uucp Mon Feb 20 16:19 EST 1989
>From gjm Mon Feb 20 16:14:21 EST 1989 remote from coma
Subject: RE: JLCooper MSB+

I haven't observed catastrophic failure with the MSB+ due to overloading
the merge function.  I have observed lost data, mostly audible by loosing
a Note Off (On 0) message which results in stuck notes (time to hit the
panic button).

-Gary

>From uucp Thu Jan  4 13:42 EST 1990
>From nsw  Thu Jan  4 13:41:50 1990 remote from cord
Received: by cord.garage.att.com; 9001041841
Date: 4 Jan 90 13:41:50 EST (Thu)
From: Neil Weinstock <nsw@garage.att.com>
Subject: Re: MIDI Thru
Message-Id: <9001041841.AA12828@cord.garage.att.com>
To: twitch!midi


A MIDI thru is essentially just a straight through connection from the MIDI
input.  No processor or UART gets involved; it's just a very simple
bit of wiring.  That's why so-called "MIDI Thru boxes" are very cheap.

A more appropriate place for a device's internally generated MIDI data to be
merged with the input data would be at a MIDI output, though this function
is seldom seen on synthesizers.  It is found on some MIDI master controllers,
which may (as in the Roland A-80) perform some massaging of the input data
before merging it into the output.

If you want this function but your synth doesn't support it, you could
get a similar effect by placing the synth's MIDI out and thru into a merger
(e.g., Pocket Merge).

    ________________    __________________    _________________________
////                \\//                  \\//                         \\\\
\\\\ Neil Weinstock //\\ att!cord!nsw  or //\\ "Your hair is so...     ////
//// AT&T Bell Labs \\// nsw@cord.att.com \\//  lustre-laden." - Moss  \\\\
\\\\________________//\\__________________//\\_________________________////

>From tjt Fri Aug  4 16:03 EDT 1989
To: midi
Subject: re: Midi Masters and Slaves

>From arpa!STONY-BROOK.SCRC.Symbolics.COM!ESC Fri Aug  4 13:15 EDT 1989 remote from att
From: Eric S. Crawley <ESC@STONY-BROOK.SCRC.Symbolics.COM>

    Date: Thu, 3 Aug 89 19:35 EDT
    From: twitch!midi-request@att.att.com

    #### Forwarded from the Mostly MIDI mailing list (twitch!midi) ####
    >From uucp Thu Aug  3 18:22 EDT 1989
    >From srm Thu Aug  3 17:21 CDT 1989 remote from ihlpm
    To: twitch!midi
    Subject: re: Midi Masters and Slaves

    If you want to get fancier, there are a variety of MIDI patch bay/merger/processor
    boxes which cost much more.  I haven't really investigated these much further
    since all I really want is a cheap switcher.  Anyone care to comment here?

    Sam Mullins

Sure, I'll comment:

I have been using a JLCooper MSB+ switcher as the center of my MIDI
network for almost a year and can't live without.  I initally bought a
DMC MX-8 switcher/processor and actually liked it better than the MSB+
but I needed the 2 extra inputs provided by the MSB+.  

Both units allow switching (MSB+ is 8x8 and the MX-8 is 6x8) and both
have 2 "processors" that allow you to process data from 2 separate
inputs by filtering out unwanted data (SYSEX, realtime, program change,
aftertouch, etc.) and by changing the channel of the data.  The MX-8
goes much further by allowing you to do velocity cross-switching, delay
and MIDI echo, and keyboard range mapping.  The MX-8 was also cheaper at
the time.  Both units allow the outputs of the "processors" to be merged
at any output.

Both units are programmable and allow you to select an input and channel
to send program change commands that change the program of the switcher.
Each program can send a number of program change commands to every
output on various channels.  So, you could have the switcher set up
every slave when you select a program on the switcher.

I have my MSB+ set up so it responds to program change commands from my
master controller keyboard.  I just select different programs on the
controller to select different configurations.  I leave the MIDI
transmit channel of both of my controllers set to channel 1 and use the
MSB+ to change the channel for a particular slave.  That way, I don't
have to worry about changing things on my controllers.  

A switcher is a must if you use a computer as a sequencer and patch
editor.  You would have to recable every time you wanted 2 way
communication between a slave module and the computer if the slave
doesn't have a keyboard on it.

I could go on, but I think this is a good summary.  If anyone wants more
details, I would be happy to provide them.

>From uucp Mon Aug  7 09:06 EDT 1989
>From gfd Mon Aug  7 09:00:38 1989 remote from mtdca
FROM:       g.f.demarest
TO:         twitch!midi
DATE:        7 Aug 1989   9:00 EDT
SUBJECT:    switcher

> I have been using a JLCooper MSB+ switcher as the center of my MIDI
> network for almost a year and can't live without.  I initally bought a
> DMC MX-8 switcher/processor and actually liked it better than the MSB+
> but I needed the 2 extra inputs provided by the MSB+.  


Any prices on these boxes?  Is that 6 in 8 out?  Thrus?  Merging?

>From uucp Thu Mar  1 23:02 EST 1990
>From upheisei!rick Fri Mar  2 12:04 JST 1990 remote from attunix
To: upheisei!attunix!twitch!midi
Date: 1990 Feb 24 Thu 1.27.82 EMT
Subject: midi boxes

One quick item: I bought a "midi through" box and quickly discovered to
my chagrin that it can't merge MIDI channels.  I recommend against buying
any box that can't merge channels.  When I bought it, I wasn't thinking about
that eventuality, but I turned out to really need it; now I flip knobs all
the time because I don't have it.
	Rick

>From uucp Mon Aug  7 17:16 EDT 1989
>From STONY-BROOK.SCRC.Symbolics.COM!ESC Mon Aug  7 15:21 EDT 1989 remote from arpa
Received: from DJINN.SCRC.Symbolics.COM by STONY-BROOK.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 637653; 7 Aug 89 15:31:47 EDT
Date: Mon, 7 Aug 89 15:21 EDT
From: Eric S. Crawley <ESC@STONY-BROOK.SCRC.Symbolics.COM>
To: midi@twitch.att.com
In-Reply-To: The message of 7 Aug 89 10:14 EDT from twitch!midi-request@att.att.com
Message-ID: <19890807192135.0.ESC@DJINN.SCRC.Symbolics.COM>

    Date: Mon, 7 Aug 89 10:14 EDT
    From: twitch!midi-request@att.att.com

    #### Forwarded from the Mostly MIDI mailing list (twitch!midi) ####
    >From uucp Mon Aug  7 09:06 EDT 1989
    >From gfd Mon Aug  7 09:00:38 1989 remote from mtdca
    FROM:       g.f.demarest
    TO:         twitch!midi
    DATE:        7 Aug 1989   9:00 EDT
    SUBJECT:    switcher

    > I have been using a JLCooper MSB+ switcher as the center of my MIDI
    > network for almost a year and can't live without.  I initally bought a
    > DMC MX-8 switcher/processor and actually liked it better than the MSB+
    > but I needed the 2 extra inputs provided by the MSB+.  


    Any prices on these boxes?  Is that 6 in 8 out?  Thrus?  Merging?

As I recall, they were around $350+ or so but my memory isn't too great.
The MSB+ was closer to $400.  The MX-8 is 6 in, 8 out and the MSB+ is 8
in, 8 out.  There are no MIDI-Thrus and if you think about it for a
minute, you realize that a thru on a switcher is not a good idea.  Both
units will merge any two inputs to any number of outputs (that is one of
the most important features!).

>From uucp Sun Jun 17 20:00 EDT 1990
>From westmark!s4mjs!mjs Sun Jun 17 18:33:24 1990 remote from att
Received: by westmark.UUCP (smail2.5)
	id AA17184; 17 Jun 90 18:33:24 EDT (Sun)
Received: by s4mjs.uucp (/\=-/\ Smail3.1.16.1 #16.20)
	id <m0hh7qR-0000vZC@s4mjs.uucp>; Sun, 17 Jun 90 18:11 EDT
Message-Id: <m0hh7qR-0000vZC@s4mjs.uucp>
Date: Sun, 17 Jun 90 18:11 EDT
From: mjs@s4mjs.uucp (M. J. Shannon)
To: midi@twitch.att.com
Subject: Confusion?

Here's my current setup:

 (Computer)             (drums)                 (tones)
+---------+          +------------+          +------------+
|      out|--------->|in HR-16 out|--------->|in FB-01 out|--\
| MQX-16S |          +------------+          +------------+  |
|       in|<-------------------------------------------------/
+---------+

No matter which order I put the modules in, I lose the capability of
getting system exclusive responses from *one* of them!  Surely I'm not
the first person in the world to come to this conclusion, so I guess my
question is: what piece of hardware do I need to get *both* sets of
system exclusive responses (and more, if/as/when I manage to expand my
setup)?

I assume my setup must change to:

+---------+          +------------+          +------------+
|      out|--------->|in HR-16 out|--------->|in FB-01 out|--\
| MQX-16S |          +------------+          +------------+  |
|         |                                        |thru     |
|       in|<---------------------------\         inV         |
+---------+                             \    +------------+  |
                                         \---|out MAGIC in|<-/
                                             +------------+

or (I guess this is preferable):

+---------+          +-------------+         +------------+
|      out|--------->|in FB-01 thru|-------->|in HR-16 out|--\
| MQX-16S |          +-------------+         +------------+  |
|         |                 |out                             |
|       in|<----\         inV                                |
+---------+      \    +------------+                         |
                  \---|out MAGIC in|<------------------------/
                      +------------+

Now, given that such a MAGIC box exists, it would probably come in two
different flavors: merger (always merges its inputs) and switch (allows
reception of a single input).  For each flavor, how many inputs could I
get merged/switched, and what's it gonna cost me?  Do any of you have
such a beasty that you would recommend (for or against!)?

For the record, it is *not* critical the way I currently use the setup,
but it will be in the forseeable future.

	Thanks for any advice,
	Marty Shannon

P.S. Them pictures are a pain to draw!

>From uucp Mon Jun 18 00:40 EDT 1990
>From upheisei!rick Mon Jun 18 11:24 JST 1990 remote from attunix
To: upheisei!attunix!twitch!midi
Date: 1990 Mai 24 Thu 1.00.56 EMT (90/06/18 Mon 02:24 GMT)
Subject: merge?

Marty Shannon:  While I don't have a merge box myself, it sounds like
you need one.  I don't know of any that can merge > 2 inputs, but then
I haven't looked very closely.  I do continue to kick myself for NOT
getting a merge box (I got a switch box).

Drawing diagrams.  Sounds like you need a more powerful text editor.
I happen to know that the AT&T Toolchest has a version of EMACS that
has an over-write mode.  Makes it much easier to draw such pictures.
There are probably other editors out there as well that can do it...

	Rick

>From uucp Mon Jun 18 12:14 EDT 1990
>From wbf Mon Jun 18 11:13 EDT 1990 remote from cbema
Subject: Marty Shannon's Confusion

/-----------\    /---------\    /---------\    /---------------\
| SEQUENCER |    |  DRUMS  |    |  TONES  |    | YOUR NEXT TOY |
| in    out |    | in  out |    | in  out |    |    in  out    |
\-----------/    \---------/    \---------/    \---------------/
   |     |          |   |          |   |             |   |
   |     |          |   |          |   |             |   |
/--------------------------------------------------------------\
| out    in        out  in        out  in           out  in    |
|                  ONE HECKUVA MAGIC BOX!                      |
\--------------------------------------------------------------/

The above figure is what it sounds like you need, Marty.  The magic box
could be a Digital Music MX-8 or a KMX MIDI Central (15 in/16 out) or some
equivalent box.  I'm not familiar with the KMX, but the add says it does
merging.  From the picture, it looks like only inputs 1 and 2 can be merged
but that may be misleading.  Better check that out for yourself.  I have
the Digital Music and it includes two "processors."  Each processor can do
many types of operations like merging, filtering, keyboard splitting,
re-channelizing, etc., and accepts data from any two inputs.  It is,
however, only 6 in by 8 out... or is that 8 x 6?  I forget which way it is.

Note that no THRUs are shown in the diagram above.  You can add things like
and Alesis Midiverb II so that it gets program change commands sent to its
input from the thru output of the drum machine, for example.  Since this
reverb doesn't do bulk dumps (it's not programmable) you won't have to
waste an input/output pair on the switch/merger/magic box.

Bill Fox	cbema!wbf


>From tjt Mon Jun 18 11:52 EST 1990
To: midi
Subject: Re: Confusion?

> > THRU boxes are also a potential solution.  They give you
> Do they exist with multiple INs and a single OUT?

You definitely need at least a merger.  Simple ones can be gotten
for under a hundred dollars (I have a J.B.Cooper one that looks ugly
but has various dip switches for filtering things separately on each
of the 2 input channels).  How much are those tiny Anadek (sp?) ones?
I've never seen a merger with more than 2 inputs, but I assume you
could gang them together - anyone have a system where they need to do
that?    ...Tim...

>From uucp Mon Jun 18 15:48 EDT 1990
>From nsw  Mon Jun 18 14:47:52 1990 remote from edsel
Received: by edsel.garage.att.com; 9006181847
Date: 18 Jun 90 14:47:52 EDT (Mon)
From: Neil Weinstock <nsw@garage.att.com>
Subject: Mergers and Acquisitions (of mergers)
Message-Id: <9006181847.AA13558@edsel.garage.att.com>
To: twitch!midi

Tim asked if anyone has a system where they need to have ganged mergers,
or many lines merged into one.  Well, consider a system (similar to mine, not
so coincidentally ;-)  where the computer is used for sequencing and sys-ex
storage, and there are several slave instruments plus a keyboard or two.
*Everything* needs to have input to the computer (for librarian purposes),
and  this implies a many->one merger.  I think this is similar to the
configuration that started this thread.

The obvious alternative is to take some MIDI switcher (a la MX8), and use
different patches to allow various instruments to send data to the computer.
That's not too bad, as it only requires some button pushing (infinitely better
than cable-jockeying.)

However, it sure would be nice to have all the outs plugged into a mega-merger,
with the merger output going to the computer.  Even though a single MIDI line   couldn't handle 4 instruments doing sys-ex dumps at the same time, I doubt      that would ever happen.  Typically, you'd be doing serious sys-ex work with
only one box at a time, or at least you could restrict yourself to that in 
order to avoid blowing away the merger.

                                - Neil

--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--
Neil Weinstock @ AT&T Bell Labs         //     What was sliced bread
att!edsel!nsw or nsw@garage.att.com   \X/    the greatest thing since?

>From uucp Mon Jun 18 18:06 EDT 1990
>From rft Mon Jun 18 16:01 EDT 1990 remote from cblpe

To: midi
Subject: merge boxes


I have an Anatek merge box that merges two inputs to a single output.
The list price is $100, but the discount price is about $90.
It does not require batteries (it steals power from the MIDI circuit)
and it is very small, about 3x2x2.

I use it to merge the output of two keyboards.
The merged output is used as input to my sequencer.
I have not noticed any lost data.
I am happy with the Anatek merge box and it seems to suit my needs
just fine.

I am sure there is an upper limit on the box's capacity to merge two
simultaneous data streams, but I am not sure what the limit is.
There may be problems if you are trying to dump the memory of one synth
while trying the play notes/dump data on/from another synth.
You may lose data.
If you just want to do dumps from one synth at a time, everything
should be fine.

You can gang the Anatek mergers if more merging capability is needed.
I seem to recall they recommend using a maximum of 4 before plugging into 
some type of MIDI gear.
This would allow you to merge 5 units into one MIDI data stream.
Also I seem to recall that Anatek uses some sort of data compression
when the two MIDI data streams are merged.
I think the Anatek merge box outputs a data stream that uses running status.
This may cause problems for some applications.
I will try to check on this sometime this week.

Running status is where the only the first message of a data stream targeted
to a given channel includes the status byte.
The status byte includes the message type, e.g. note on , note off, and
the channel number.
The status byte is then sent only when the message type changes.

Anatek also has a new "studio" version that can merge 7 or 8 inputs  (I don't
remember which, I have a spec sheet at home and I can post more info if anyone
wants more details) and also includes some sort of patch bay facilities.
I don't know the price.


Rick Taylor


>From tjt Mon Jun 18 16:32 EST 1990
To: midi
Subject: re: Mergers and Acquisitions (of mergers)

> *Everything* needs to have input to the computer (for librarian purposes),
> and  this implies a many->one merger.
  ....
> The obvious alternative is to take some MIDI switcher (a la MX8), and use
> different patches to allow various instruments to send data to the computer.
> That's not too bad, as it only requires some button pushing (infinitely better
> than cable-jockeying.)
> However, it sure would be nice to have all the outs plugged into a
> mega-merger, with the merger output going to the computer.

I agree.  Makes me wonder why we haven't seen such things.  I guess one
reason might be that if you have 2 identical synths (not unusual - I have
a small setup, but still have 2 TX81Zs), and you send a sysex message
that requests a dump to be sent to the computer, how do you identify
which one you want to send the dump?  Guess we need synth id's something
like SCSI id's.      ...Tim...

>From uucp Mon Jun 18 17:29 EDT 1990
>From gjm Mon Jun 18 16:29:17 EDT 1990 remote from coma
To: twitch!midi
Subject: Re: Mergers and Acquisitions (of mergers)


I second what Neil said about using different patches on your MIDI switch
box.  The JLCooper MSB+ accepts program change commands to switch patches,
I'm sure that the other boxes do to.  So your program should be able to
switch patches to talk to appropriate devices as needed, one (or two at
most) at a time.  I haven't tried any switch box patch changes which do
anything that is time critical, so if you're interested in sw.box patch
changes during a performance, you'll probably have to construct your own
tests and try out a box before you buy it.

-Gary

>From uucp Mon Jun 18 19:51 EDT 1990
>From greg Mon Jun 18 17:48 CDT 1990 remote from iwtgp
From: iwtgp!greg (Gregory A Youngdahl +1 708 979 0013)
To: twitch!midi
Subject: Midi Switcher Construction

Hi all,

	Just thought I'd add my 2 cents worth and remind everyone about
the Electronic Musician article in the Aug. '87 issue called "MIDI
Switcher Primer".  Its not a merger, and it sounds like that is what
is required to satisfy the need that started this thread, but if a
switch would suffice, and build-it-yourself intrigues you, this article
describes how to do it.  It is an electronic switch, complete with
opto-isolation per the MIDI standard.  It also wouldn't be too hard to
wire up a MIDI version of the RS-232 or Parallel 'T' switches that
select various serial or parallel devices to connect to computer ports.
I'm sure parts for a 'T' switch solution would be available at the local
Parts-R-Us (Radio Shack) for much less than the $100 merger.

	I can supply copies of the Electronic Musician article (reprinted
without permission) if anyone is interested.

Greg Youngdahl    AT&T Bell Laboratories   Naperville, IL
att!iwtgp!greg

>From uucp Mon Jun 18 22:52 EDT 1990
>From druwy!mab Mon Jun 18 19:44:46 1990 remote from mtuxo
FROM:       druwy!mab
TO:         mtuxo!twitch!midi
DATE:       18 Jun 1990  19:44 MDT
SUBJECT:    re: Mergers and Acquisitions (of mergers)

> reason might be that if you have 2 identical synths (not unusual - I have
> a small setup, but still have 2 TX81Zs), and you send a sysex message
> that requests a dump to be sent to the computer, how do you identify
> which one you want to send the dump?  Guess we need synth id's something
> like SCSI id's.      ...Tim...

Most of the synths I've played with include a system channel in their
sysex messages.  If you have two identical synths, you can set them to
different system channels so that you can control them separately.  I've
seen this on several Kawai synths, FB-01, and CZ-1000.  All of them with
the possible exception of the CZ would remember their system channel when
powered off, although that blasted K5 requires you to manually enable
receipt of sysex messages after power-up.  Same for FB-01.  Maybe someday
the designers will realize we want hands-off MIDI control of all of our
toys.

Alan

>From uucp Tue Jun 19 10:10 EDT 1990
>From honet4!dob Tue Jun 19 09:02 EDT 1990 remote from att
From: honet4!dob (Duane O Bowker +1 201 949 2607)
To: att!twitch!midi
Subject: Mergers, etc.

Hi.  Just wanted to add two cents to Neil's comments RE: MX8.  That looks
like the way to go for many inputs to PC in librarian situations.  You CAN
get by without any button pushes since the MX8 is itself switchable via
MIDI program change commands, if you configure it to be.  So you would have
an MX8 set for librarian/editor A, another for librarian/editor B, another
for playing MIDI guitar into synth module C, et cetera...then build into
your librarian/editors appropriate program change commands to MX8 to switch
among the MIDI configurations.  Other cool trick that I use in my lab at
work quite a bit: MX8 set-ups can include up to 8 program change commands
sent to MIDI outputs of MX8.  In other words, I can send one program change
to the MX8 and that will configure essentially all the MIDI stuff in the lab.
(The MX8 holds the "macro set-ups", if you will, for the laboratory).

						Duane Bowker
						att!honet4!dob

>From tjt Tue Jun 19 09:43 EST 1990
To: midi
Subject: Re: Mergers and Acquisitions (of mergers)

>From mosc  Tue Jun 19 08:16:09 1990 remote from aloft
From: aloft!mosc (52842-H. S. Moscovitz)

This is a comment on real time patch changes on the JLCooper MSB+.
I have one of these devices and you can do this sometimes.  It is
not repeatable.  The safest think is don't send midi information
on your network for about 1 second after you told the MSB+ to 
change patches.  Maybe they should have called this box the MSB-.
				Howard Moscovitz
				aloft!mosc

>From uucp Tue Jun 19 14:40 EDT 1990
>From billb  Tue Jun 19 13:39:42 1990 remote from mtunj
Received: by mtunj.ATT.COM (smail2.6)
	id AA02897; 19 Jun 90 13:39:42 EDT (Tue)
To: twitch!midi
Subject: Re: Mergers and Acquisitions (of mergers)
Message-Id: <9006191339.AA02897@mtunj.ATT.COM>
Date: 19 Jun 90 13:39:42 EDT (Tue)
From: billb@mtunj.ATT.COM (William Burnette)


>>   This is a comment on real time patch changes on the JLCooper MSB+.
>>   I have one of these devices and you can do this sometimes.  It is
>>   not repeatable.  The safest think is don't send midi information
>>   on your network for about 1 second after you told the MSB+ to 
>>   change patches.  Maybe they should have called this box the MSB-.
>>   				Howard Moscovitz
>>   				aloft!mosc
>>   
There is a bug in the MSB+, probably in the implementation of
running status.  If a program change directed to the MSB+
is followed by sysex to anywhere, the MSB+ changes patches
again, probably to some number found in the sysex data.
I've found that a workaround is to follow the MSB+ patch
change by a patch change or note off to some channel
other than the MSB system channel.  This seems to allow
speedy MSB+ patch changes.  I haven't used this method
in a song sequence, only in a patch editor.

				  Bill Burnette
				  mtunj!billb


>From uucp Wed Aug  9 11:26 EDT 1989
>From srm Wed Aug  9 10:25 CDT 1989 remote from ihlpm
To: twitch!midi
Subject: MIDI switcher revisited



Hi Midi-mailers:

Last week I had said that I had been looking for a cheap MIDI switcher
for some time but had not found anything.  My equipment consists of
a Casio CZ-1 keyboard, a Roland MT-32,  a Roland MKS-100 sampler and
an Atari ST.  Because I need to have 2-way communication between 
each synth and the computer, I need at least 4-in switcher/thru box.
(I won't bore you with all of the possible configurations.)  
I thought that I had a lead on a realtively cheap
Sonus box but that turned out to be only 2-in.  So I was back to the
cheapest alternative being a Kawai 4-in 8-out for about $125.
Keep in mind that this buys you no memory or processing.  I guess
I don't understand why these things are so expensive.

I had toyed with the idea of building my own but really didn't want to
spend time to deal with all the active electronics necessary for the
"thru" part of the box.  The switching part is easy.

Finally I arrived at a compromise.  I realized that I don't have many
delay problems by daisy chaining using the thru ports on my synths. 
The real problem I was trying to solve was that I need to have any of 
the synth's outputs feed the Atari's input.  So, last night I just built a 
simple switcher.  A project box, 4 5-pin DIN connectors,  a double-pole
quadruple-throw switch and some wire.  About $9 at Radio Schlock.  It
took me about 2.5 hours and I'm really not too swift with a soldering
iron.

Here is my current setup (I tried to draw a picture but it was getting
frustrating.)

ST out ---> CZ-1 in
CZ-1 thru ---> MT-32 in
MT-32 thru ---> MKS-100 in

CZ-1 out ---> switch input 1
MT-32 out ---> switch input 2
MKS-100 out ---> switch input 3
Switch output ---> ST in

This allows me to run patch librarians for the Roland things which have
a 2-way protocol for bulk transfers.

If I decide that I can't live with the delays on the thru ports, then I will
buy one of those MIDIMIX 6 thru devices for $35.  I still have one switch
position open if I want to add one more input.

I realize that this is NOT practical for those of you with a much larger
setup but for me it is great.  It did exactly what I wanted and I saved
at least $100. I don't really need merging or a thru capability.... yet :-)

Sam Mullins

>From att!messy.bellcore.com!mo Thu Aug  2 19:01:19 0700 1990
Received: by att.att.com; Thu Aug  2 19:03:29 1990
Received: by messy.bellcore.com (5.61/1.34)
	id AA08681; Thu, 2 Aug 90 19:01:19 -0700
Date: Thu, 2 Aug 90 19:01:19 -0700
From: mo@messy.bellcore.com (Michael O'Dell)
Message-Id: <9008030201.AA08681@messy.bellcore.com>
To: midi@twitch.att.com
Subject: KMX Midi Central... 

I have two and love them.

The KMX only costs a bit more than the others and you get
a completely general cross-bar - one gozoutta and one gozinna
per midithing back to the KMX.  works like a dream....

oh yes - has a merger in 1 & 2, which I route to outputs  14 & 15
so I can run anything into the merger and through as well.

quality stuff!
	-Mike
>From att!media-lab.media.mit.edu!joe Thu Aug  9 16:41:45 EDT 1990
Received: by att.att.com; Thu Aug  9 16:40:58 1990
Received: by media-lab.media.mit.edu (5.57/DA1.0.2)
	id AA28488; Thu, 9 Aug 90 16:41:45 EDT
Date: Thu, 9 Aug 90 16:41:45 EDT
From: Joe Chung <joe@media-lab.media.mit.edu>
Message-Id: <9008092041.AA28488@media-lab.media.mit.edu>
To: midi@twitch.att.com
Subject: Opcode Studio 3


I think someone on this list asked about this box & I replied more or less
in favor of it... I've changed my mind!!

IT STINKS!!! 

I found that it has *massive* problems dealing with slightly flaky
time code when used as a SMPTE to MIDI Time Code converter. I found
that when a SMPTE dropout occurs, the box can end up mis-synched by
over 15 frames, and remain mis-synched after good SMPTE is available.

I called a not-very-well-informed tech person at Opcode who seemed to
think that this behavior was part of the box's algorithm to jam
through dropouts (known as non-continous jam synch). I explained to
him *very* vocally that non-continous jam synch mode is something
which you *have* to be able to turn off, and that in any case, you
rarely want it for Midi Time Code conversion. You almost always want
the real numbers even if there's a drop out or discontinuity, since
good software can theoretically correct for errors. (note that this
stuff has nothing to do with the "Jam" button on the front of the
box... That seems to be for SMPTE copying only.

One factor may have been that I was using 25fps SMPTE... They may have
a bug at this rate which they haven't noticed so far. I'm going to try
to talk to someone who had something to do with building the thing.
(The guy I talked to said "25 FRAMES PER SECOND???" ... I said "yeah,
that's the European standard." He said "no it isn't, it's 24!" Guess
who was right.)

Conclusion: Be extremely careful if you want to use this thing. It
will probably behave OK as long as your SMPTE is flawless, but in my
experience, you always get a few little glitches here and there.
Winding up miss-synched is so incredibly annoying (and in my case
unusable) that the dangers of this box should be carefully considered.

On the other hand, I bought it from Sam Ash for $300, and it is
currently the only rack-mount MIDI interface. It also has a built-in
power supply & detachable cord instead of an external transformer &
flimsy little cable.

The only SMPTE->MTC converter that I would be caught dead with is the
Adam Smith Zeta-3. It is a way-pro device & costs it (~$2500)... It is
also one of the best two-transport tape synchronizers around.

I also messed a bit with the Tascam MIDIizer. I dont' recomend it. We
had a bunch of flakiness, also featuring random miss-synchs. It's a
real pain in the butt to use as a straight SMPTE->MTC converter as
well.

By the way... Anyone mess with the Roland A-880 MIDI patch bay? I had
huge problems with this thing when I used it's merge function. Ever
notice that Roland equipment is very cavalier about sending random
sustain/omni-mode/whatever offs? Why don't they ever let you defeat
these "features?"

>From keyhole!tjt Mon Dec  3 21:05 EST 1990
To: twitch!midi
Subject: re: Midi Cabling

>         I just want to be able to:
>                 1) Drive the Korg from either the Keyboard or
>                    the computer WITHOUT messing with cables.

You need a 'merger', a box with 2 MIDI in's and 1 MIDI out.
Anadek (sp?) makes a 'pocket merger' for < $100.  The November 1990
Electronic Musician magazine has a do-it-yourself project for building
a merger.

>         Can two 3-wire midi cables be "wire-or'ed"?

In some situations it can work (if the 2 inputs are never
simultaneously active), but occasionally still produces garbage.
Anyone know if there are cases when it can actually damage
the equipment?    ...Tim...

>From homxb!gabin Tue Dec  4 08:21 EST 1990
From: homxb!gabin (Jay Gabin +1 201 615 2830)
To: twitch!midi
Subject: re: Midi Cabling


About using merger cables and such...

I had the same problem of wanting to drive my set-up from
my main keyboard or my computer without fooling with
cables.  I bought a MIDI patch-bay (I think it's J. L. Cooper
and cost $80 mailorder).  This has 3 inputs and 8 outputs.
With switches on the front, you can make each of the 
outputs receive any of the 3 inputs.  So, with a flick of
these switches I can bring the computer into or out of
the loop.  

Jay

>From tjt Tue Dec  4 10:49 EST 1990
To: midi
Subject: Re:  Midi Cabling

>From att!ifi.uib.no!gunnars Tue Dec  4 08:47:10 +0100 1990
From: gunnars@ifi.uib.no (Stud. Gunnar Sylthe)

James Fischer says:

=>         Can I do this?
=> 
=>                 Keyboard                      M3R       
=>                   Out ---------->----+-------> In
=>                                      ^
=>                 Computer             |
=>                   Out ---------->----+
=> 
=>                 Computer                      M3R
=>                   In <------------------------ Out
=> 
=>         ... by soldering up a custom cable?  Can two 3-wire midi
=>         cables be "wire-or'ed"?
=> 
=>         Or must I buy a "Thru box" to do this?

I think what you need is some kind of MIDI merge box. If I were you
I'd look into the Anatek PocketMerge. I've never used it myself, and
my sole "experience" with it is from reading the ads and a test in a
Norwegian musician's mag, but it is supposedly a cheap, simple and
reliable gadget which does one thing (merging to midi ins to one
midi thru) and does it well.

If you do try it out, why not post some "test results" to the list?
I, for one, would be interested.

--Gunnar Sylthe
gunnars@ifi.uib.no

>From mtuxo!druco!mab Tue Dec  4 10:15 MST 1990
To: mtuxo!twitch!midi
Subject: Anatek PocketMerge

I bought the Anatek PocketMerge a while back and it works as advertised.
It has two INs that are merged into one OUT.  I haven't stress-tested it,
so I don't know how well it would handle lots of simultaneous data from
both INs.  My main use is to minimize cable swapping, so I rarely have
both INs feeding data at the same time.

    Alan Bland
>From mvgpl!mvcrg Tue Dec  4 15:50 EST 1990
From: mvgpl!mvcrg (Christopher R Gayle +1 508 960 2904)
To: twitch!midi
Subject: merge box

Did read in recent MMML items somewhere that there is a build-it-yourself
merge box? I'd like to know where to get the info to build one... has
anyone out there done this? If so, how did it perform? 

How about switchboxes? Are they too complicated for reliable hand-wired
projects? Any experience with these?

- Topher Gayle   x2904   mvcrg@mvgpl.att.com
>From keyhole!tjt Tue Dec  4 19:11 EST 1990
To: twitch!midi
Subject: do-it-yourself merge box

> From: mvgpl!mvcrg (Christopher R Gayle +1 508 960 2904)
> Did read in recent MMML items somewhere that there is a build-it-yourself
> merge box? I'd like to know where to get the info to build one... has

The November 1990 issue of Electronic Musician magazine had an article
that showed how to build one.  You can buy the PC board and programmed
EPROMS from the author of the article, for $28 and $10, respectively.

    ...Tim...

>From tjt Tue Dec  4 19:19 EST 1990
To: midi
Subject: Re: merge box

From: ihlpe!greg (Gregory A Youngdahl +1 708 979 0013)

Topher,

	The build-it-yourself box you are referring to may be the MIDI
switch box that I had mentioned a while back.  This is an electronic
switch circuit described in an article in the Aug. '87 issue of
Electronic Musician.  This circuit may be a solution to James' need
to avoid physically switching cables, but it is not a merger.  I have
a copy of the article that I could send out if you are interested in
it.  I have not built this unit, but I don't see why it wouldn't do
the job it is intended for.  Of course the job could also be accomplished
with a mechanical switch.

--
Greg Youngdahl    AT&T Bell Laboratories    Naperville, IL
att!ihlpe!greg
>From tjt Wed Dec  5 11:48 EST 1990
To: midi
Subject: Re: do-it-yourself merge box

>From ihlpm!srm Wed Dec  5 09:37 CST 1990

>The November 1990 issue of Electronic Musician magazine had an article
>that showed how to build one.  You can buy the PC board and programmed
>EPROMS from the author of the article, for $28 and $10, respectively.
>
>    ...Tim...

Before anyone goes to a lot of trouble....My brother just bought a Pocket
Merge for $59 for DJ's in ChicagoLand.  At that price it may not be worth
building a kit for $38+.  Of course, if you enjoy that kind of thing go ahead.

Sam
