[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference napalm::commusic_v1

Title:* * Computer Music, MIDI, and Related Topics * *
Notice:Conference has been write-locked. Use new version.
Moderator:DYPSS1::SCHAFER
Created:Thu Feb 20 1986
Last Modified:Mon Aug 29 1994
Last Successful Update:Fri Jun 06 1997
Number of topics:2852
Total number of notes:33157

1894.0. "Programming Roland MT32 L/A Synth Module" by VOLKS::RYEN (Rick Ryen 285-6248) Tue Feb 07 1989 19:55

I have a MT-32 and a MC300 sequencer. I'd like to try my hand at
crafting(or modifying) a few custom timbre patches to the MT32.
I understand that you can load them with SYSEX data and that there are
32 blank timbre locations for customizing your own. My sequencer 
supports creating SYSEX data messages.

Here's my problem(s).

	I am having great difficulty decyphering Rolands midi implementation
	spec for the MT32. It's very cryptic to me. Also, I don't have much 
	of an understanding of 	the parameters that can be changed
	and their effect on the sound. One of the reasons that I want to
	diddle with patches is so I can learn that type of thing.
	
	But, so far I haven't been able to formulate any valid system
	exclusive message.

Has anybody out there written patches for the MT32?
(they are supposedly Roland D-50 and D110 compatible)

Can anybody suggest a tutorial type book describing how to
read and apply midi-implementation notes?

There is supposed to be after-market MT32 patches available.
Anybody had any experience with them.

Anybody know how a checksum is calculated in a midi message?

Anybody out there read/understand pidgin english?

T.RTitleUserPersonal
Name
DateLines
1894.1Not Quite CompatibleDRUMS::FEHSKENSTue Feb 07 1989 20:077
    MT-32 and D-110 patches are not quite D-50 compatible.  The D-50
    has additional LFOs that the MT-32 and D-110 don't have, and the
    onboard effects are different.  Bearing that in mind, you can port
    patches between the machines.
    
    len.
    
1894.2Time consumming...MASTER::DDREHERTue Feb 07 1989 22:248
    I've done some sysex stuff on the MC-500 to change parameters on
    a Yamaha FB-01.  It is very time consuming.  I was able to change
    FB-01 configurations (not patches) via MIDI and and the LCD display
    to discriptive names for configs. 
    
    You must be formatting the sysex MIDI message wrong. 
    
    Dave
1894.3Upward compatableDREGS::BLICKSTEINYo!Wed Feb 08 1989 12:319
    Perhaps its true that MT-32 patches are compatable with the D-50
    in that any MT-32 patch can be moved to the D-50, but not the other
    way.
    
    I have an MT-32 but I haven't created any patches for it.  Don't
    even have the means to do so (PC).  I'm probably gonna replace it
    with an M-1R eventually anyway.
    
    	db
1894.4huh?NORGE::CHADWed Feb 08 1989 15:077
RE: d50 (somesortof) compatiblity to MT32

Don't think so.  The MT32 uses a different set of PCM samples that are numbered
differently if memory serves me right.  One of many things that are different.

CHad
1894.5ReactionsVOLKS::RYENRick Ryen 285-6248Wed Feb 08 1989 15:2034
re .2 
	I am prepared for it to be time comsuming. I have been sequencing
	complete songs in step mode, which is also very time consuming.
	I think that you are correct that I am formatting the midi
	messages wrong. I'd just like to do something like...
		"I wonder what it would sound like if I changes this
			parameter from x to Y".
	
	I did make some prograss last night. I was trying to do something
	simple first, to learn the message format. Turned out that what I
	was trying went aginst the midi spec. I was trying to do a control
	change from within a system exclusive message. I know better now.

	I have gotten a couple of error messages (checksum error) on
	the MT32. That's progress for me, since I mostly get no
	response at all!

	What is still throwing me a bit is the internal memory layout of
	the MT32. I also need to understand how to calculate a checksum,
	(or maybe the MC300 will do it for me?)
	There was some mention in the spec that the result of the checksum
	is all bits set except for the high order bit. But whats the
	algorithm. (must be MT32 specific)

Re. 3  Compatibility - Yes, understood. I did think that that level of 
			compatibility would make it possible for folks
			who have the more sophisticated gear to answer
			some of my questions.

			Whats a M-1R? (I'm always thinkin about what
			the next upgrade will be, even though I haven't
			had the MT32 long)

1894.6It's a "better MT-32"DREGS::BLICKSTEINSo What? BEEG Deal!Wed Feb 08 1989 15:429
    > What is an M1R?
    
    It's the rack mount version of the Korg M-1 "Music Workstation"
    (i.e. integrated synth/sequencer).
    
    Perhaps it's best to explain how it's different from the MT-32:
    
    	It sounds much better and is much more flexible.
    
1894.7Stay Tuned for MoreVOLKS::RYENRick Ryen 285-6248Wed Mar 15 1989 17:0117
Stay tuned for more.

	As a result of this note, a fellow is sending me some
	"Roland Technical Notes', on how to modify MT32 internal
	parameters.

	I'll update this note when I get the mail, and have tried 
	a few examples.

	Supposedly it is reasonably painless to modify things
	like reverb on individual voices, overall volume etc.
	It should also enable you to craft "custom" voices, although
	that would be a bit more tricky and subject to a trial
	and error method.

/rick
1894.8SYSEX to modify MT32 ReverbVOLKS::RYENRick Ryen 285-6248Thu Mar 23 1989 15:5893
	One of the readers of this note was nice enough to
	send me some notes from Roland on how to write
	system exclusive data to the MT32. It really opens up
	potential versiatility of the box.

	It is a somewhat complex subject to describe all of the
	information in those notes, but here are a few of the simpler,
	 but still very useful system exclusive messages that 
	you can use to modify the operational parameters of the MT32.


System exclusive messages for MT32.

Reset all MT32 Paramaters to power-up state.

    F0 41 10 16 12 7F 00 00 01 F7

    Note: The MT32 display will not be updated
    	  until a new display is selected.
    	  The new display should reflect the
    	  initialized values (eg; volume, tymbre)

	This is very helpful if you just stopped a
	song in the middle, and there were unresolved
	pitch bends. If you don't reinit the MT32,
	you next song may start with an undesirable
	bend, that will trash it. I plan on putting
	this at the head of every song I sequence.

SET REVERB TYPE:

    TYPE:	SYS EXCLUSIVE MESSAGE
    -------	-----------------------------
    Room:	F0 41 16 12 10 00 01 00 6F F7
    Hall:	F0 41 16 12 10 00 01 01 6E F7
    Plate:	F0 41 16 12 10 00 01 02 6D F7
    Delay:	F0 41 16 12 10 00 01 03 6C F7

    Note: Moving the alpha dial will reset the 
	 reverb to the dafault settings.

	I was very suprised at how dramatic the
	delay setting was. I didn't fool with the
	time paramater, bit I suspect it may
	make the delay effect more dramatic.

	I didn't notice a big difference between hall and plate
	and room, but then again I didn't modify
	the time and level parameters. Using the alpha dial to
	reset the level, returns the reverb type to the default.
	
SET REVERB TIME:

    TIME 	SYS EXCLUSIVE MESSAGE
    -----	-----------------------------
    1		F0 41 16 12 10 00 02 00 6E F7
    2		F0 41 16 12 10 00 02 01 6D F7    
    3		F0 41 16 12 10 00 02 02 6C F7
    4		F0 41 16 12 10 00 02 03 6B F7
    5		F0 41 16 12 10 00 02 04 6A F7
    6		F0 41 16 12 10 00 02 05 69 F7
    7		F0 41 16 12 10 00 02 06 68 F7
    8		F0 41 16 12 10 00 02 07 67 F7

SET REVERB LEVEL:

    Level * 	SYS EXCLUSIVE MESSAGE
    -----	-----------------------------
    1		F0 41 16 12 10 00 03 00 6D F7
    2		F0 41 16 12 10 00 03 01 6C F7    
    3		F0 41 16 12 10 00 03 02 6B F7
    4		F0 41 16 12 10 00 03 03 6A F7
    5		F0 41 16 12 10 00 03 04 69 F7
    6		F0 41 16 12 10 00 03 05 68 F7
    7		F0 41 16 12 10 00 03 06 67 F7
    8		F0 41 16 12 10 00 03 07 66 F7

    	* Level relates to level indicated on front panel, when
    	  the alpha dial is used to change reverb lebels. 
    	  I don't believe that the reverb display is affected
    	  when the reverb level is changes with a SYSEX msg.


If anybody finds these useful, let me know and  I'll add some more...
	like... turning on/off reverb effect on a per instrument
	(per rhythm sound) basis.

Thanks again to the noter who sent me the Roland Notes
from across the pond.

/rick    

1894.9System Exclusive De-Mystified!NEWVAX::MCLENDONThu Jun 15 1989 02:43289
Welcome to the enlightenment of System Exclusive.  It is TRULY the
power of controlling multiple synthesizers INDEPENDENTLY whether they
be from the same manufacturer or not.  True, in some songs, you can
still use program change messages per channel, but with a Top 40 band,
sometimes having a *blend* of piano or guitar effects is REAL nice!
The *only* way to keep all the synthesizers from getting confused
and selecting the wrong patch (in this example), is by using 

          S Y S T E M   E X C L U S I V E   C O M M A N D S


After reading thru most of these notes, I see there is still a muddy
pond over this System Exclusive thing.  System Exclusive commands are
valid on most synthesizers.  Roland's nasty checksums are not!

What you learn here will apply then (in a generic sense) to *ALL*
synthesizers that have this capability.  What follows at the end of
this editorial, is a quick-reference chart for setting up some things
on the MT-32.  Do they work?  Yes, I've been using the MT-32 for more
than one year in a performing Top 40 band it does WONDERFULLY!  Yes,
the reverb is a bit noisy, but on stage, you just can't tell. Besides,
as you will soon learn, you can tailor it PER SONG and even change it
DURING a song! (provided that you're also using a sequencer or computer 
from which to send the system exclusive commands you've programmed 
BACK into the MT-32!)

I had to struggle as I'm sure many of you had.  Basically, I sat down
with what Roland calls their MIDI Implementation section and stared
LONG AND HARD at what was going on. The "secret" is this (for all you
assembly language programmers): Everything is an offset from some base
address.  When I felt I had the command correct, the only thing that
COULD be wrong is the base address.  Yeah, inputting SYSEX commands
into a sequencer (or sequencing program) can be a pain, but for me,
there was NO choice - and the results have made this box INDESPENSIBLE!

Let's take a quick (?) look at how it's all done.  Got your owner's manual? 
Can you add in base 16 (Hexadecimal)?  Here goes:

Find the last few pages in the owner's manual of the MIDI Implementation
that has a paragraph/section marked, "PARAMETER BASE ADDRESS".

Find   PATCH TEMPORARY AREA (Part 1)
    This is where you make changes to the parameters.
    Note that the start/base address is (hex) 03 00 00
    Note that PART 2 starts 16 bytes later:   03 00 10
    Note that PART 3 starts 16 bytes later:   03 00 20

Find the NEXT box marked  PATCH TEMPORARY AREA
    This box contains the actual offset for the parameter whose value you
    want to change or set.  Stay with me, now:

      00 00   TONE GROUP
      00 01   TONE NUMBER
      00 02   KEY SHIFT
      00 03   FINE TUNE
      00 04   BENDER RANGE

    With just this much, we can assign ANY one of the 128 voices to
    any part by constructing a data stream or just one byte.  And,
    even change a few things, as you're about to see!

*****************************************************************
****  Set  PITCH BEND to 2  for  PART 1  ************************
*****************************************************************

      F0              - REQUIRED Starting System Exclusive byte

      41 10 16 12     - Manufacturer's ID, Unit ID, Model #, COMMAND
                         These four bytes are FIXED for the MT-32 unless
                           you change the Unit ID upon power up.  If you
                           don't know how, then DON'T; it's a lot easier!
                         The "12" is a no-handshake command; i.e., the
                           data stream is sent and no feedback is expected.
			 Therefore, you probably won't need a cable from
                           MIDI OUT jack when you're performing, because
                           by THAT time, you should only be sending data 
                           to the MT-32

      03 00 00        - This is the offset for PART 1
         00 04        - This is the offset for the PITCH BEND
               02     - This is the data
                 77   - This is the checksum (explained later)
                   F7 - This is the end of the SYSEX command


THE ENTIRE COMMAND:

F0  41 10 16 12  03 00 04  02  77  F7 


OK, it's a lot just to change the pitchbend.  BUT IT'S THE ONLY WAY TO
DO IT!  The power in all this, comes from the fact that you can specify
the setting for the ENTIRE VOICE by giving a data stream.  Use the starting
address (03 00 00) and include values for each of the parameters:

F0  41 10 16 12  03 00 00  01 0E 18 32 03  ck  F7

What we've told the MT-32 to do here is this:
 (Remember, this is an ADDRESS followed by values)

    03 00 00 = 01  == Select Bank B (Program change #s 65 to 128)
    03 00 01 = 0E  == Select Voice 14 in Bank B (Really 64+14)
    03 00 02 = 18  == Normal octave (See table example below)
    03 00 03 = 32  == Fine tuning EXACT (no +/- cents adjustment)
    03 00 04 = 03  == Pitch bend set to three half-steps


The answer should be very clear now:  By telling the MT-32 that we're
starting with  03 00 00, the next byte(s) of data will be placed in
the succeeding locations (as described immediately above).  The MT-32
knows when it receives the F7 that the command has ended, so it looks
for the checksum in the preceding byte.  It has NO way of knowing how
long the command stream is until it reads the F7!  Which means you can
change as much (the whole voice) or as little (just the BEND) as you
need for any given part.

Calculating the checksum is nothing more than sequentially adding all the
data bytes to the address (from above):  03 00 00 01 0E 18 32 03


Becomes:  03 + 00 + 00 + 01 + 0E + 18 + 32 + 03  =  5F

The checksum is this result subtracted from 80 (hex!) = 21
Do it in Decimal:   5F = 5x16 + F  =  80 + 15  = 95
                    80 = 8x16 + 0  = 128
                    21 = 2x16 + 1  =  33      <--------+
                                                       |              
	      128 - 95 = 33    <-- WORKS EVERY TIME! --+

The table at the very end of this note makes this much easier, but you
still have to add up the values; either on paper or with a calculator
that does hex math.  Or translate each value to decimal, use the
calculator to add 'em up, and subtract from 128.  Then, convert the
result back to hex.

By examing the following chart, you will learn the correct base
address to use in order to perform changes anywhere inside the MT-32.
I haven't covered all possible options, but by re-reading this section
of the owner's manual, perhaps this tutorial will make more sense.

                            HAVE FUN!

  *************************************************************
  ***********  YOU ARE NOW, NO LONGER, A MIDIOT!  *************
  *************************************************************


                      ROLAND MT-32 SYSTEM EXCLUSIVE SETUPS

     HEX BYTES     < a1 a2 a3 db ck >   EXPLANATION
     -------------- ------------------  ----------------------------------------
     F0 41 10 16 12  10 00 15 0F 4C F7  SET RHYTHM to MIDI CHANNEL 16
                     10 00 0D 04 5F F7  SET PART 1 to MIDI CHANNEL  5
                     10 00 0E 00 62 F7  SET PART 2 to MIDI CHANNEL  1
                     10 00 0F 08 59 F7  SET PART 3 to MIDI CHANNEL  9
                     10 00 10 0n ck F7  SET PART 4 to MIDI CHANNEL n+1
                     10 00 11 0n ck F7  SET PART 5 to MIDI CHANNEL n+1
                     10 00 12 0n ck F7  SET PART 6 to MIDI CHANNEL n+1
                     10 00 13 0n ck F7  SET PART 7 to MIDI CHANNEL n+1
                     10 00 14 0n ck F7  SET PART 8 to MIDI CHANNEL n+1

                     10 00 16 vv ck F7  MASTER VOLUME (0-100: 32>50 46>70 50>80)
                     10 00 01 0m ck F7  REVERB MODE 0-3 (Room,Hall,Plate,Delay)
                     10 00 02 tt ck F7  REVERB TIME
                     10 00 03 ll ck F7  REVERB LEVEL


     F0 41 rc 16 12  01 00 KC dd ck F7  CHANGE RHYTHM PARAMETER 
 key #                                    rc=Rhythm MIDI Channel#  -1   
(BD1=35, for example)                      K=HEX Key value = (Note#-24)x4
                                           C=Command (0-3) added to K
                                              

     F0 41 10 16 12  03 00 00 ** ck F7  SET PART 1 PARAMETERS
                     03 00 10 ** ck F7  SET PART 2 PARAMETERS
                              ** = Data TIMBRE GROUP  0-3   (A,B,Memory,Rhythm)
                                        TIMBRE NUMBER 0-63  (Patches 1 to 64  )
                                        KEY SHIFT     0-48  (-24:+24 18h=norml)
                                        FINE TUNE     0-100 (-50:+50 32h=norml)
                                        BEND RANGE    0-24  (   00h to 18h   )
                                        ASSIGN MODE   0-3   (Poly 1, 2, 3, 4 )
                                        REVERB SWITCH 0-1   (00= Off, 01= On )
                                        ( not used )  00    (Can't skip this!)
                                        OUTPUT LEVEL  0-100 (0>0 32>50 64>100)
                                        PANPOT        0-100 (0=RIGHT, 7F=LEFT)


     mc  CC  7 vv                       SET VOLUME (vv) for any PART
                                         mc = MIDI Channel #
       ALL SYNTHESIZERS ARE              CC = Control Change     
     AFFECTED BY THIS COMMAND            vv = 0>0  32>25  64>50  96>75  127>100 
   THAT ARE LISTENING TO CHANNEL mc      vv = MT-32 value wanted x 1.28

Again, in some cases you can use this.  In others, you will NEED to
use the System Exclusive command to change/set volume for a part.
NOTE:  The  CC 7  command uses DECIMAL values!  SYSEX uses hexadecimal!


     F0 41 0p 16 12  00 00 02 dd ck F7  TRANSPOSE PART (0p) ABSOLUTE
                                        dd: 00 (-2 oct)  0C (-1 oct) 18=Normal
                                            30 (+2 oct)  24 (+1 oct)


     F0 41 10 16 12  04 00 00 ** ck F7  SET TIMBRE PARAMETERS for PART 1
                     04 01 76 ** ck F7  SET TIMBRE PARAMETERS for PART 2
                                          ** = Data stream
                                             = Tone name,...partials

You can use this command to write/rename an MT-32 voice, but putting
in hex values for the ASCII letters is a pain!  The various MT-32
librarian/editors for use with a computer are MUCH better for this!

However, if you want to brighten up the string voice, first select the
voice for the PART you want, then execute this command for the same
PART (of course) and just change the TVF value for the partial(s)
used.  Again, each TVF parameter is assigned an offset address *AND* you
will have to add the offset for each partial to be altered:

Again, for PART 1 ('cause it's easier):

      F0  41 10 16 12     (The standard stuff, now that you know it...)

          04 00 00        (PART 1's base address for TIMBRE settings)

..and add to this, the offset for the partial you want to change:

          00 00 0E        Partial 1
          00 00 48        Partial 2
          00 01 02        Partial 3
          00 01 3C        Partial 4

NOW, add the offset for the VALUE you want to set:
             00 17        TVF CUTOFF FREQ

Thus, to set the TVF CUTOFF FREQ=64  for PARTIAL #3 for whatever voice
you've selected to be PART 1,

                       a1 a2 a3    Address
                       --------

                       04 00 00
                     + 00 01 02
                     +    00 17
                       --------
      F0  41 10 16 12  04 01 19  40  ck  F7

Yeah, it's tedious, but the results are worth it.  For one,
the TVF setting will ALWAYS be fixed; because you've embedded this
command in the first measure (probably) of your composition where you
perform the other System Exclusive set-ups.  You *DO* initialize
things in the first measure, DON'T YOU???!!


                               CHECKSUM CALCULATOR

Checksum (ck) is calculated by adding addresses (a1 to a3) plus data bytes (db)
and locating that number by row and column.  The intersection is the checksum.
In the example for RHYTHM channel, you have 10 + 00 + 15 + 0F = 34  (HEX math!)
The checksum for 34 = 4C

In the example for the TVF CUTOFF, the total is 5E   ( = 04 + 01 + 19 + 40)
The checksum for 5E = 22


                0  1  2  3   4  5  6  7   8  9  A  B   C  D  E  F 
             +----------------------------------------------------+
           0 |    7F 7E 7D  7C 7B 7A 79  78 77 76 75  74 73 72 71 | 0
           1 | 70 6F 6E 6D  6C 6B 6A 69  68 67 66 65  64 63 62 61 | 1
           2 | 60 5F 5E 5D  5C 5B 5A 59  58 57 56 55  54 53 52 51 | 2
           3 | 50 4F 4E 4D  4C 4B 4A 49  48 47 46 45  44 43 42 41 | 3

           4 | 40 3F 3E 3D  3C 3B 3A 39  38 37 36 35  34 33 32 31 | 4
           5 | 30 2F 2E 2D  2C 2B 2A 29  28 27 26 25  24 23 22 21 | 5
           6 | 20 1F 1E 1D  1C 1B 1A 19  18 17 16 15  14 13 12 11 | 6
           7 | 10 0F 0E 0D  0C 0B 0A 09  08 07 06 05  04 03 02 01 | 7 

           8 | 00 7F 7E 7D  7C 7B 7A 79  78 77 76 75  74 73 72 71 | 8
           9 | 70 6F 6E 6D  6C 6B 6A 69  68 67 66 65  64 63 62 61 | 9
           A | 60 5F 5E 5D  5C 5B 5A 59  58 57 56 55  54 53 52 51 | A
           B | 50 4F 4E 4D  4C 4B 4A 49  48 47 46 45  44 43 42 41 | B

           C | 40 3F 3E 3D  3C 3B 3A 39  38 37 36 35  34 33 32 31 | C
           D | 30 2F 2E 2D  2C 2B 2A 29  28 27 26 25  24 23 22 21 | D
           E | 20 1F 1E 1D  1C 1B 1A 19  18 17 16 15  14 13 12 11 | E
           F | 10 0F 0E 0D  0C 0B 0A 09  08 07 06 05  04 03 02 01 | F
             +----------------------------------------------------+
                0  1  2  3   4  5  6  7   8  9  A  B   C  D  E  F 
   

1894.10Great stuff!!!!CSC32::MOLLERNightmare on Sesame StreetFri Jun 16 1989 15:1539
	I can't agree with you more about the abilities of the MT32. As
	most people say, there are some issues about noise & some 
	restrictions as compared to a D110. However, I've used both
	the D110 & MT32 in Live situations & they are quite comparable
	in what they can do. My MMT-8 (Alesis) sequencer can build & send
	a system exclusive message. In the next few weeks (when I'm done
	mixing down my duo's new Demo Tape), I'll give this a try. More
	than anything else, I'd like to set the Master Volume Level and
	change the Pitch Bend from 1 octave to 1 or 2 notes & then back to
	an octave (I only use this in 2 songs at the moment, because an
	octave is not always easy to deal with effectively).

	The MT32 has some equalization zits, however, Radio Shack started
	dumping thier Parametric Equalizers for less than $15.00 (they used
	to be about $40.00) & these can be tweeked to correct the problem
	quite nicely. You may have to look around various Radio Shacks to
	find one, since they are discontinuing them, but the look is worth
	it (I own 3 of them now, One is in my Synth Rack, another is going
	into the Vocal/PA Rack & the other is sitting on one of the shelves
	near my 4 track tape deck). - If you are curious, the improved MT32
	equalization involves setting a boost on all 3 ranges & setting
	the low to fully clockwise then the mid and high to fully counter 
	clockwise (well almost fully counter clockwise on the high). The 
	boost is at about 70% to 80% of the range that the Radio Shack 
	Parametric Equalizer allows. This does some amazing things for the
	bass/snare drums & any Bass instruments that you've chosen.
	It looks like Roland put in an RIAA curve (same as used for phono
	input on home sterios) in anticipation that that was what the MT32
	would be used for - Most PA's arent equalized that way.

	There are supposed to be some upgrades to the MT32 that can be added
	that give you the ability to have seperate outputs, Battery Backup,
	Corrected equalization, Improved Reverb & front panel editing, as
	on a D110. I wanted to get some of this added, but they don't do
	it locally & I can't survive without my MT32 for two weeks while
	waiting for someone to enhance it. I wonder what added SYSEX functions
	would be supplied (and to whose standards) for these upgrades.

								Jens 
1894.11? about enhancementsSWAV1::STEWARTThere is no dark side of the moon...Mon Jun 19 1989 14:5214




>>	There are supposed to be some upgrades to the MT32...

	Oh, yeah?  When, where, how, & most importantly, how much?  Where did
	you hear about this?





1894.12This is as far as I looked into itCSC32::MOLLERNightmare on Sesame StreetMon Jun 19 1989 15:1219
	The upgrades are not from ROLAND, but some company on the west
	coast (Oregon??). They were like $90.00 for a battery backup &
	$100.00 for seperate line outs (like the D-110 has). Some software
	upgrade was also available that changed the eq & digital reverb.
	You could also add front panel control.

	From what I read at Prosound (the dealer here in Colorado Springs),
	it looked like you could upgrade in steps & eventually end up with
	something more like the D-110 than the MT-32 when you were done.
	I really wanted seperate outs & wasn't too concerned about the
	other things.

	My nephew has a D-110. I'll lay odds that If I compare the PC board
	of the D-110 to the MT-32 that they have a lot in common.

	Prosound in Colorado Springs is at (719)597-9962 - Ask for Richard
	if you want further info.

							Jens
1894.13Standard parameters, rack mount, reset?VOLKS::RYENRick Ryen 285-6248Tue Jun 20 1989 20:3331
Okay,
	I've now done a few things with SYSEX on the MT32,
	like changing reverbs, layering by making multiple parts
	respond to the same midi channel. All useful and fun.

	Sure wish that I could find the magic bit to tweek to
	reset the whole works back to power on state like you can do 
	from the front panel. >>>anybody know it?<<<
	
	I've also experimented with changing some of the
	existing sounds, detuning, changing the PCM sound etc...
	I've been pretty sucessful in negotiating the addresses
	and have some small inkling of what a few of the parameters
	might mean.

	My approach has been to find a sound similar to what I want, 
	and then do some parameter modifications to effect a change.
	Unfortunately, since I can't READ (examine) the parmeters
	with my MC300, I have no way of knowing the original parameters
	to know how much to change them! Is there any kind of a 
	manual that lists the parameters used in the MT32's standard sounds?

	The bottom line is, that I haven't been very sucessful in crafting
	useful new patches. I wonder if this is just beyond what one 
	could expect to do with a sequencer, and not a computer/patch editor.

	Also, has anybody found a rack mount that works? I'm gettin tired
	of finding tables to put it on, and would like to rack it.

Rick

1894.14Here's a cheap but effective wayCSC32::MOLLERNightmare on Sesame StreetTue Jun 20 1989 22:0840
	Rack Mounting the MT-32....

	Here's what I did (it's in my Rack). I took two small angle brackets
	(the type that you use when you build wooden boxes - about $.20 at
	most hardware stores. I took off the original rubber feet, put on
	shorter rubber feet (Radio Shack has them, 12 for $.99). Set the MT-32
	onto the lowest rack position (set it on the bottom), centered it.
	Lay the 2 angle brackets so that the MT-32 fits between them with
	about 1/8 inch on either side. Screw the angle brackets to the rack
	(don't put any screws in the MT-32!!). Pull out the MT-32, then
	place 1/2 foam rubber strips (1 inch X 1 inch) against the
	angle bracket, then place the MT-32 back in. Now the foam is between
	the MT-32 & the angle brackets. Take 3 or 4 long nylon wire ties
	(hardware or electronic stores often have these) & fish one wire tie
	into each of the holes that are left over in the angle bracket.
	You may need to use an extra wire tie or 2 to connect the ties to
	each other, but, you have lowered the top of the MT-32 enough that
	you should have clearance to use the rack space above it. I also took
	a blank panel & cut it to fit on either side of the MT-32, so it
	looks like it's rack mounted but it really isn't. It's still easy
	to remove (the on/off in the back is no fun) & now fits in one space.

	I couldn't see wasting 2 spaces when I only needed one.

	The front view sort of looks like this:


			  +---------------------------+	
			| | ########### = = =     ( ) | |
			| | ########### = = =         | |
		    ----+ +--v---------------------v--+ +----

		     ^   ^             ^               ^   ^
                 Angle   Foam        MT-32          Foam   Angle
	       Bracket   Rubber                   Rubber   Bracket

	It's been traveling for the last 6 months this way & hasn't had any
	problems (you could use 2 sets of angle brackets if you wanted).

							Jens
1894.15Is $40 and 2 rack spaces too much for you?WOOFY::DIORIOCellulite Heroes never really dietWed Jun 21 1989 15:0011
	Rack Mounting the MT-32....

RE -2  Hey Rick, I saw a rackmount kit for the MT32 in one of the magazines I
subscribe to (I can't remember which one but it is either Keyboard, Electronic
Musician or the Transoniq Hacker--those are the only mags that I subscribe to).
I remember it costing $40 I believe, and takes up 2 rack spaces. Have you seen
that one? If you're interested, I'll find out all the necessary info for you
to get one. 

Mike D
1894.16HPSMEG::LEITZbutch leitzWed Jun 21 1989 15:102
...they have 'em at Wurly's...

1894.17I'm inspired to built a rackVOLKS::RYENRick Ryen 285-6248Wed Jun 21 1989 15:599
et.al. re; Rack mounting..

	Hmmm $40...,, I think I've been inspired to build my own. I like
	the idea of putting it on the bottom, and only using a single
	rack space. Gotta save room for other gadgets!

	Thanks for the suggestions.

Rick
1894.18System Exclusive Editing - Part 2MDCRAB::MCLENDONThu Jun 22 1989 02:2495
      RE:  .13

      (Portions of .9  may be a pre-requisite for this discussion)
      (Also, I have errored in that when you tell the MT-32  "07")
      (the result is an "8" -- MIDI is zero-based.  Gotta add a 1)
      (to just about everything you do.  Sorry for any confusion.)

      Rick,
            Glad to hear that things have worked out well for you!
      I don't think that "magic bit" exists.  If it *DID*, it would
      appear in the MIDI IMPLEMENTATION section... and beLIEVE me,
      I would have found it!  Sorry, but it sounds like some of the
      Roland engineers didn't talk to the rest.  Oh, well, looks like
      you've got a LOT of system exclusive programming ahead of you!
      Let's see how much of it is REALLY worth doing from a sequencer!

      RE: "Is there any kind of a manual that lists the MT-32 sounds?"
      Yeah, once again, it's (sorta) in the MIDI IMPLEMENTATION guide.
      *AND*, once again, it is faaar more efficient to design a sound
      using dedicated s/w!!  You can really use BOTH!  Once you know
      which element of which partial to change, it's back to SYSEX to
      effect that change (see example in .9).  Yes, you CAN examine the
      existing patches and the MIDI IMPLEMENTATION guide will tell you
      what you're looking at.

      Let's expand on what's needed here:

      First, make sure you've got MIDI-IN and MIDI-OUT connected to your
      sequencer (and MT-32, of course) then enter this command onto a 
      track.  Then just "playback" your sequencer in record mode to
      send this command and capture the returned data:

            F0   41 10 16 11   04 01 76   00 01 76   0E  F7

      This tells the MT-32 to send the  00 01 76  bytes of data at
      address  04 01 76    (PART 2   Timbre Temporary Area).
      We'll get into the math in a moment, but the 0176 translates to
      246 decimal and *not* to 374 decimal!  Note that delta = 128!

      So, what happened here?  Where does the data go?  Where is it?

      ANSWER:  The "11" command is a request to dump and that's EXACTLY
      what happened!  The data requested goes OUT the MT-32.  So, if
      you were recording the data as you were playing back this command,
      (( any sequencer MUST be able to do this, otherwise you'd never ))
      (( be able to layer your sessions and thus build a song! ))
      you will find the returned data on whatever track you were recording.

      And, that data you've just recorded begins:

            F0 41 10 16 12 04 01 76 ...

      (( Note that the "11" is now a "12" so you can effectively send ))
      (( this command back to the MT-32... like after you've altered  ))
      (( some data... but you really wouldn't WANT to do it this way. ))

      If you now examine this data, and your MT-32 manual is handy, you'll
      see the first 10 bytes are characters in hex that spell the name of 
      the patch at PART 2.  You'll also note that the name is the first 
      element of the common parameters for a given tone.  This is then 
      followed by a few bytes for structure data and envelope mode.  Each
      partial requires 3A bytes of data, starting with WG PITCH COARSE.
      Multiply this out:

          3*16 + "A"   =   48 + 10   = 58 bytes
                                      x 4 partials
                                     ----
                                      232 bytes
                                     + 14 bytes for COMMON data
                                     ----
                                      246 bytes for ONE TONE!

      Here's how the  01 76  gets converted to the 246:

            01x80  + 76   = F6    = 15x16 + 6    = 246

      REMEMBER:  In MIDI, any value above 7F is a command, so you CAN'T
      read the 0176 as   256 * 1   + 76     This is why the hex 80 is used
      in the above calculation.

      Yeah, this becomes REAL cumbersome!  But, the MIDI IMPLEMENTATION
      is your map to what's where... it's just that a software editor is
      sooooo much easier to make these kinds of changes.  And, how do they
      do it?  Why, they just read the MT-32's "reply" into a buffer and
      then write it out to the locations beginning at  05 00 00   or
      08 00 00   depending on what you want to do.  This is DEFINITELY a
      job for a software editor, to be sure.  But, for a few small and
      effective changes, the system exclusive will work quite well, now
      that you know what you're doing!

      A good (Dr T's) MT-32 editor is about $80.  Be SURE to get an editor
      that can send output to a printer!  It's faster than a file dump that
      makes you invoke a WP just to print!

      Good Luck & have fun!  The gain is -almost- worth the pain... sometimes!
1894.19Thanks for the solutionVOLKS::RYENRick Ryen 285-6248Tue Jun 27 1989 14:4719
re: .19

	Wow! I think that it is really slick that when you request
	the MT32 to send data, that it's formatted so that you can
	send it back again.

	I was aware of the ability to request data, but I had no
	idea of the format that it would get sent back. With this
	new knowledge, I should be able to do exactally what I originally
	wanted to do. That is, read parameters for a sound, tweek them a bit,
	and observe how the changes effects the sound. 

	I have a pretty good map that identifies the parameters,
	(basically just pictorals of the implementatiopn spec)
	Can't wait to find the time to play around with this.

Thanks much,
Rick

1894.20Roland Gets This RightDRUMS::FEHSKENSFri Jun 30 1989 15:489
    re .19 - That's one thing that Roland usually gets right - their
    bulk dump and load formats are the same, so if you have a sequencer
    that can record SysEx data, all you have to do is play it back to
    the synth to do a bulk load.  Some of the older synths require a
    "handshake", and the send and receive headers differ in one byte,
    but the more recent synths are a cinch for bulk load/dump.
    
    len.
    
1894.21MS-DOS patch editor for MT-32SWAV1::STEWARTThere is no dark side of the moon...Wed Jul 05 1989 23:4014



	I have a public domain patch editor for the MT-32 that I pulled
	down off of IEMUG.  I haven't loaded it into my system yet, but
	if anyone's interested I'd be glad to share my opinions...  and
	yes, I know about viruses; my music computer doesn't have a hard
	disk anyway... 





1894.22Is DOS patch editor for MT-32 any good?MOSAIC::HOULEOS/2 - Catch the waveThu Jul 06 1989 16:299
    Re:   <<< Note 1894.21 by SWAV1::STEWART "There is no dark side of the moon..." >>>
    
    I'd be interested in hearing about the patch editor for the MT-32
    and how good it works.  I know other people I work with would be
    interested as well.  Would you please post a note and let us know what
    you find out.
    
    Thanks,
    Paul
1894.23shareware looking for homeSWAV1::STEWARTThere is no dark side of the moon...Fri Jul 07 1989 22:5514
	I'll do better than share my opinions; I'll share the software so
	that you can form your own opinions.  I'm currently seeking an
	existing MIDI repository on the net (VAXmail sent to Brad
	Schafer) where you can get this stuff. 

	Once again, the package is shareware and the author is only
	asking $20.  It includes a Timbre editor (MTTED), a patch mapper
	(MTMAP), documentation, and a sample file of public domain
	patches.  I haven't (and probably will not have time in the near
	future) to load this stuff up and play with it at home, so
	if/when you do, it would be nice to let people know what you
	think. 

	I'll add another reply here as soon as the package is up on the net. 
1894.24worth every penny I paid for itSWAV1::STEWARTThere is no dark side of the moon...Tue Jul 11 1989 00:1214
	I've just spooled the stuff up to DYO780::.  Hopefully Brad will
	tell us when it's available. 

	I unARCed the package and played with it a little on Saturday. 
	First disappointment: the sample timbre file the documentation
	promised was included in the distribution kit was missing.  So I
	should've known better than to believe documentation, I know... 

	Second disappointment: the program ran, allowed me to change all
	the parameters on screen, and downloaded data into the MT32, but
	the patch didn't play.  This could possibly be cockpit error,
	'cause I was in a hurry and just wanted to make a new sound and I
	have no clue as to how LA synthesis works.  Maybe one of you will
	have better luck... 
1894.25MS-DOS MT32 patch editor E-Net locationSWAV1::STEWARTThere is no dark side of the moon...Thu Jul 20 1989 22:4714

	    The MSDOS patch editor for the MT32 is now on the net (Thanks,
	    Brad!)  Try it and let us know how it works for you... 


	    Directory DYO780::DISK$USER05:[MIDILIB.TOOLS.IBM.M32EDLIB]

	    M32EDLIB.ARC;1          116   8-JUL-1989 00:00:00.00  (RWED,RWED,RE,)

	    Total of 1 file, 116 blocks.



1894.26ATPS::GRASMANNI am my Father's SonThu Jan 11 1990 19:1232
    Hi -
    
    Being horribly anxious to be massively creative, I need some help with
    this topic.
    
    I failed to get the previous .ARC software to do anything but crash my
    system, despite the suggestions included.
    
    So, Yesterday, I received some software from a book which includes a
    SEQUENCER and PATCH LIBRARIAN for MIDI.  I can get the Sequencer to
    play music, but the Patch Librarian, from what I've been able to read
    so far, requires a keyboard.  For instance, part of the book reads,
    "Run this program.  Now press the BULK DUMP" on your synthesizer, and
    watch what happens!"
    
    Well, having only an MT-32 for a synthesizer, and no button labeled
    "BULK DUMP", I read the MIDI implementation manual, and user's guide. 
    It says there is a way to do BULK DUMPs and BULK LOADs, but then it
    doesn't tell me how!
    
    Can anyone relay that info to me?  That is, is there a system-exclusive
    message that I overlooked which says, "OK, DUMP-it-to-me,
    DUMP-it-to-me!"
    
    Also, in terms of programming PATCHES for the MT-32, another question:
    
    Is it possible to EDIT patches and hear the effect at the same time,
    WITHOUT a keyboard?
    
    Thanks in advance,
    
    		Steve
1894.27ATPS::GRASMANNI am my Father's SonThu Jan 11 1990 19:267
    ....did I mention that I have an MPU-401 sequencer attached to the
    MT-32?  Well, I do.  And it's plugged into a 386 PC-AT (lookalike)...
    
    After reading more notes, it seems wise to talk about one's "exact"
    setup...
    
    thanks again,  SGr.
1894.28Handshake job required.EICMFG::BURKEJim Burke, @UFCFri Jan 12 1990 07:5710
    As far as I know, bulkdumping to/from the MT32 is a handshake job. You
    have to send it a SYSEX telling it to bulkdump.
    
    I use a Mega-ST with C-Lab's Explorer (handles all Roland LA synths).
    You can tweak patches and hear them by fiddling with the mouse. You
    don't need a keyboard.
    
    Any use ?
    
    Jim Burke
1894.29elucidationDYO780::SCHAFERBrad - boycott hell.Fri Jan 12 1990 12:4619
    To elaborate on .28, the MT32 is a fairly "stupid" beast, as far as
    front panel controls go.  To the best of my recollection, I don't think
    there is a way to initiate a bulk dump from the front panel; it all
    must be done via system exclusive MIDI commands (SYSEX). 

    Like Jim said, dumping patch memory to MIDI via SYSEX requires
    "handshaking" on the MT32; in very simplistic (and somewhat inaccurate)
    terms, the computer says "hey, give me a bulkdump", the MT32 says "you
    ready?", the computer says "yes", the MT32 dumps its brains. 

    As for auditioning patch changes, if your editor is able to send a
    "test sequence" - ie, a short series of notes - to the machine, you
    should be able to.  Most good librarians/editors have this feature. If
    not, you could always have another task that generates MIDI data
    running and context switch to that (you can do that on a 386, right?). 

    Note that all edits made to the MT32 are lost when you power down. 

-b
1894.30ATPS::GRASMANNI am my Father's SonFri Jan 12 1990 17:3921
    Thanks -  I had thought that might be the way it's done.  My problem is
    that I don't know the SYSEX message which says "BULK DUMP/(LOAD)" to
    accomplish this.  All the manuals (Owner's guide and MIDI
    implementation manual) I have don't reveal this info, or if it does,
    does so cryptically and I can't decypher it.
    
    I would like to spend the weekend playing with this, so if anyone
    knows, please write soon!
    
    The software I am using comes from Jim Conger's books, "C Programming
    for MIDI", and "MIDI Sequencing in C".
    
    As for switching contexts, yes, I am running DESQview 386 and can
    switch between various "jobs".  I tried that last night, and it seems
    that the jobs "interfere" with each other.  Somehow, the MT-32 appears
    to get confused when communicating with 2 different programs.
    
    RE: Mega-ST - "any use?"  Yes, it could be.  Is this something that I
    can get "today"?  How much?  Where?  Shareware? (hope, hope...)
    
    	Steve
1894.31a few answers (hopefully)DYO780::SCHAFERBrad - boycott hell.Fri Jan 12 1990 18:0016
    Handshaking is (in my experience) somewhat time-dependent.  In other
    words, if I tell the MT32 to dump its brains, then my program had
    better be ready to get in an "are you sure" message and respond "yes",
    otherwise the MT32 ignores the request.  The SYSEX for this should be
    in the manual somewhere ... but these *are* Roland manuals, aren't
    they?  (Sigh) 

    As for confusion among computer pgms, you've gotta make sure that
    you're not trying to xmit MIDI data from two programs *at the same
    time*. There's a great deal of possibility for brain damage in this
    case.  No reason why it shouldn't work in a serial context-switch
    fashion, though.  Strange.

    Can't answer your other questions - sorry.

-b
1894.32I don't knnow why I can't use 1200 baud...ATPS::GRASMANNI am my Father's SonSat Jan 13 1990 13:5524
> No reason why it shouldn't work in a serial context-switch
fashion, though.

Maybe...  Unfortunately, I think it's "Parallel context" which
would explain the brain-damage...

I'm still looking for the codebook to decypher the ROLAND
manuals...

And I refulse to dial in over the weekend at 300 baud again!

So, if anyone has any technical help...?  I'll be checking
in again Monday morning.

Oh, yeah...  Can you believe this?  The *author* of the books
I mentioned called me up from California last night!  I am
going to call him back today, and I'll bet he can give me
the tech. help I need.  He wrote the assembly level code to
communicate with the MPU-401...  I'll post any info that
might be of use.

ciao,

		Steve
1894.33MT32 breakdown of patches by how many partialsNORGE::CHADMon Jan 15 1990 12:3525
RE: the "fraud" accusation regarding the MT32 and 32 polyphony from the
Proteus note


Anyway, I looked at my photocopy of my MT32 preset patch list over the
weekend and wrote down how many of the 128 used 1 partial, how many 2, etc.

Unfortunately I forgot to bring the list but the breakdown was something like
this:

patches using 1 partial :  16   --  12.5%   (this one I'm sure of)
patches using 2 partials:  58   --  45.3%    (pretty sure here)
patches using 3 partials:  42   --  32.8%    (about that many...)
patches using 4 partials:  12   --  9.4%    (about that many -- it was less than
                                              1 partial patches)

Like I said, this is from memory.  However, the listing of 1 and 2 partial
patches was more than 1/2 of the available patches.

Chad

PS: I'll be gone for 3 weeks in Europe so any debate here won't get any reaction
from me for a while.

Smiles
1894.34There has to be better things to debateDREGS::BLICKSTEINConliberativeMon Jan 15 1990 13:1719
    re: .33
    
    Two points:
    
    	o Even looking at the "percentage of presets" it's STILL more
    	  than the Proteus and thus it is more fraudulent for the MT-32
    	  to claim 32 voices than the proteus.
    
    	o The "percentage of presets" is misleading.  It shouldn't
    	  come as a surprise to anyone that more of the USEFUL and
    	  better sounding patches fall into the 3 and 4 partial
    	  set.
    
    Chad, the MT-32 is simply NOT as polyphonic (in realistic applications)
    as the Proteus and thus I believe that Brad and I are completely
    correct by claiming that the 32 voice claim is less of a lie for the
    Proteus than it is for the MT-32.
    
    	db
1894.35i agree. end of my arguing.DYO780::SCHAFERBrad - boycott hell.Mon Jan 15 1990 13:487
    I was referring specifically to the D110 in my posting in the Proteus
    topic.  I don't know much about the MT32, except that it sounded lousy
    enough that I didn't buy one. 

    I don't feel that way about my Proteus.  Enough said.

-b
1894.36no lie at allNORGE::CHADMon Jan 15 1990 14:4118
nitpick:  Over half of the presets are in the 1 and 2 partial range and I know
          I used a lot of them.  It is very subjective to say "more of the 
          useful ones" -- it doesn't mean anything

I can get 32 voices at once from an MT32 -- there is no *lie* or *fraud*.
I am going to check the lit. to see exactly what the claim is, not what you 
believe it is.

Chad

PS:  the reason I keep after this is because it really irks me when people go
around claiming someones lying to them when the problem is perceived message
of the person doing the claiming of being lied to.

PPS: A lie is a lie is a lie

See you all in 3 weeks
1894.37ATPS::GRASMANNI am my Father's SonMon Jan 15 1990 14:5822
    Ok, so I guess the last 4 messages weren't for me...
    
    I still need some help.  Jim Conger was more than gracious in answering
    my questions.... Too bad I didn't know the right questions to ask.
    
    My first problem lay in the fact that I didn't have TWO cables running
    between the MPU-IPC and the MT-32.  MT-32 received data, but had no way
    of responding.  OK, so I'm used to DECnet communications with only one
    wire...  At least now I can communicate with the MT-32 from C.
    
    Problem number 2 comes from the fact that I still cannot decypher
    things like "partials", envelopes, etc.  The Patch Librarian I have
    does Roland Alpha Juno syntax, and it APPARENTLY is nothing like MT-32
    syntax.
    
    Can someone *please* shed some light on this, or *please* point me to a
    note that does?  All the other "MT-32" "key"notes don't appear to have
    that info (via "search" command...)
    
    Thanks in advance,
    
    	Steve
1894.38adding to the confusionTOOK::SUDAMALiving is easy with eyes closed...Mon Jan 15 1990 15:4346
    I may be a little out of the context of the discussion going on here,
    but I have also had confusion about the Roland teminology and what I
    can and can't do with my D-10 (which is similar to the MT-32), so let
    me summarize what I've learned through experience.
    
    Partials - A partial is the basic element that makes up a tone. Another
    way to look at it is that it is the smallest element into which sounds
    can be divided (at least in the Roland world). The MT-32/D-110/D-10 all
    have a limited number of partials that can be produced at the same
    time.
    
    Tones - A tone is the next unit up from a partial. From one to four
    partials can be combined to make up a tone. More complex and
    interesting tones tend to use more partials. For example, the
    relatively pure tone of a flute might use only 1 partial, while a fuzz
    guitar might use 4 partials.
    
    Timbres - This is just a kind of "macro" that combines a tone with some
    set of parameters like reverb time, depth, etc.
    
    Voice - A note (how can I say it more simply?)
    
    
    Given this you have a hierarchy of Voices, Tones and Partials. I think
    what Roland claims is that 32-note polyphony means you can play 32
    voices at one time. Since each voice can only be sounding one tone,
    their claims would be true if there were only 32 partials and you
    assigned one to each tone, or 128 partials and you assigned four to
    each tone. I've never figured out from the documentation exactly how
    many partials you can use, but experience shows that it must be less
    than 128, because I often run out. Then I have to chase around trying
    to come up with other tones that use less partials, or rearranging my
    sequence. (There's a list in the back of the documentation for the D-10
    that shows how many partials each rhythm sounds and each preset tone
    uses).
    
    You should also remember that the rhythm parts use partials, and I
    assume these draw from the same pool as the tones do. This can be
    considerable, especially for cymbals and other percussion sounds. So
    maybe there are, in fact, 128 partials, but I'm using them up on rhythm
    stuff. If that's the case, then the MT-32 could indeed have true 32 voice
    polyphony, as long as you aren't using the rhythm channel.
    
    Does that help?
    
    - Ram
1894.40Drums don't cound against the other soundsCSC32::MOLLERNightmare on Sesame StreetMon Jan 15 1990 16:4817
	The MT-32/D-10/D-110/U-110 etc... don't count the drum sounds against
	the other allocations of partials. By this I mean you could play
	all 30 drum sounds at the same time and still have the other 32
	partials be available for allocation. This is nice since drum
	sounds can be more complex (I like lots of percussion in some
	songs) without a penalty against the voices/tibres used.

	My MT-32 has been a real workhorse for me over the last year
	and a few months. It may not be perfect, but it's earned me
	a fair amount of money & is the center of my Duo's versatility.
	Given the choice, I would rather have had a D-110 (hindsight
	is wonderful), since it allows seperate outs, but otherwise
	it's pretty versatile. The 'Noisy' issue has never been one
	of my problems with the MT-32, it's quiet enough for LIVE
	work, which is what I primarily use it for.

							Jens
1894.41I'd love to be the used car salesman who gets youDREGS::BLICKSTEINConliberativeMon Jan 15 1990 20:0528
>          Over half of the presets are in the 1 and 2 partial range and I know
>          I used a lot of them.  It is very subjective to say "more of the 
>          useful ones" -- it doesn't mean anything
    
    To call Hitler a "bad guy" is also subjective.
    
> PS:  the reason I keep after this is because it really irks me when people go
> around claiming someones lying to them when the problem is perceived message
> of the person doing the claiming of being lied to.
    
    Whether or not it is a lie, there seems a fair number of knowledgeable
    people in this conference who fully believe that it is fraudulently 
    deceptive to lead a person to believe that this is a "32 voice".
    
    It is incorrectly setting people's expectations.
    
    > PPS: A lie is a lie is a lie
    
    Let me ask you a question?
    
    If I sell you a car telling you that it'll do 185 MPH, whether it
    is a "lie" or a "deception" or whatever, do you NOT agree that it
    is fraudulent to do that if it will only achieve that speed when going down
    a hill with a 85 degree slope?
    
    It certainly is not a "lie"?
    
    	db
1894.42MIZZOU::SHERMANECADSR::SHERMAN 235-8176, 223-3326Mon Jan 15 1990 23:5215
    re: 32 voices
    
    Look, now, folks.  The MT32 is a 32 voice instrument!  I'll prove it!
    Let's say that before they marketed the MT32 they had an older version
    that only had the presets that one pertial per patch.  Definitely
    a 32-voice instrument.  Then, they came out with the MT32 and announced 
    gobs of new presets.  Unfortunately, some of the new presets used up more 
    voices.  What you have now is a box with more capability.  But, is it
    no longer to be advertised as a 32-voice box?  Of course not.  The
    issue is really that of how useful each voice is.  The argument that
    seems to make the most sense to me concerns which box has the more
    useful voices.  Besides, to somebody like Bon Jovi, *every* synth is a
    one-voice instrument.  ;^)
    
    Steve
1894.43You guys have an unusual concept of fraudSMOP::BLICKSTEINConliberativeTue Jan 16 1990 02:4813
    Steve,
    
    If we took a 32 VUP VAX and added new features which slowed it down
    by 50% (multi-port memory for example) would it NOT be fraud or at
    least unacceptably "deceptive" to continue offering it as a 32 VUP machine?
    
    Normally I love a good debate, I think this discussion is just silly.
    
    Billing the MT-32 as "32 voice synth" leads reasonable people to
    have highly unrealistic expectations.  Misleading consumers is
    fraud in my book.
    
    	db
1894.44Fraud? (Hitler V mt32) no contest!BAHTAT::KENTpeekayTue Jan 16 1990 06:1215
    
    
    re- Before the bickering, (which hasn't changed in context or content
    in my 5 months away from this conference)
    
    To program the midi"syntax" of the MT32 you will need a copy of the
    MT32 manual and a good understanding of midi and its language. I am not
    surprised that the Juno sysex wouldn't work for you. It's a completely
    differnt synth with a completely different voice architecture. You
    really need to get out the Mt32 midi sysex pages and play around. When
    yo have some specific questions come back and ask. There where a couple
    of guys around this conference programming the mt32 via sysex. Perhaps
    they will speak up.
    
    				Paul.
1894.45The finest example of fraud without lying I've seenDREGS::BLICKSTEINConliberativeTue Jan 16 1990 13:0432
    Anyone remember that ad that was running in things like TV Guide
    for a:
    
    			"Satellite Dish Antenna"
    
    This was an ordinary cheapo rabbit ear style antenna with a plastic
    "satellite dish" in front of it.  The dish itself was entirely an
    ornament, it perform no actual reception function or enhancement.
    
    Someone bought 10 thousand of these and ran one of the wildest ads
    I have ever seen.
    
    That ad did everything it could do to lead you to believe that one
    of these babies did the same things that "bigger" satellite dishes
    do, but never actually told an a true "lie".
    
    It said things like:
    
    	"You don't pay cable because you aren't hooked up to cable"
    
    		- which is true, but the dish antenna didn't allow
    	          to RECEIVE cable either.
    
    As if to wink at those that knew better, the summary said that the
    dish was a "marketing breakthrough".
    
    I would like to send a copy to Chad and see if he feels there's any
    element of fraud in that ad even though it tells no lies.
    
    	db
    
    
1894.46let's move this or quit, eh?DYO780::SCHAFERBrad - boycott hell.Tue Jan 16 1990 13:094
    I would like to ask that this diatribe be moved off-line or to the
    banter topic.  No sense in fouling up a good technical topic. 

-b
1894.47Here's a couple specific questions about *programming*ATPS::GRASMANNI am my Father's SonTue Jan 16 1990 15:5636
    RE: .44 (Paul)
    
    Thanks!  Well, OK.  It so happens I have some very specific questions. 
    I just wasn't sure that the "Programming Roland MT-32 L/A Synth Module"
    was the right topic... ;-)
    
    I know about parameter base addresses - I can even get the MT-32 to do
    a dump now, by sending a byte sequence like:
    
    F0 41 16 10 41 03 00 00 00 00 0A F7
    
    This requests a 10-byte dump from 03 00 00 (I forget offhand what that
    address is).  And I can get tones by specifying NOTE ON's followed by
    note/velocity byte parameters.
    
    Now for a couple of the specific questions ... 
    
    1.) To create a "patch change" (is that the right term for creating my
    own "sound" (timbre?)), are there really 58 (if I remember right)
    parameters that I have to pass to the MT-32?  Is that what is described
    on the page that has all those "ENV, TVP, etc." and other cryptic stuff
    in the implementation manual, right after the "command" section?
    
    2.) To what base address do I dump timbre parameters to change the
    voice on Channel 1?  Channel 2?
    
    There are a number of "apparent" addresses according to the
    implementation manual.  like "00 00 00" thru "02 00 00" (which I
    believe are "channel specific" as opposed to "device specific".  I
    can't figure out what they mean, probably due to my lack of
    understanding of the terms used "Temporary area", "Temporary timbre
    setup", etc.
    
    Thanks,
    
    		Steve Gr.
1894.48Bits of midi!BAHTAT::KENTpeekayWed Jan 17 1990 11:4922
    
    
    One of the problems we midi users have is that the standard is more of
    a communication protocol than a data definition statement. Although
    there is , as with all technoligies, some jargon that becomes accepted.
    We are still in our infancy in this area, witness the ongoing
    discussion on what is a ",voice","polypony" et al.
    
    One which has become de-facto is a patch change command. This is not an
    sintructio to the synth to modify the sound it is currently playing,
    but to laod a new sound from it's bank of "patches' even this gets
    complicated by some of the manafacturers who use alterantives sucah as
    ,configurations, performances etc. However pretty much every body
    agrees as to what a patch change is.
    
    The instructions you need to modify, as opposed to load, a sound will
    differ significantly from synth to synth and architecture to
    architecture. So you really need to dig into your  MT32 manual and work
    on it. There may be a way of changing nuance in small instructions.
    There may not be. The manual should tell all.
    
    					Paul.
1894.49ATPS::GRASMANNWhat got your node out of joint?Wed Jan 17 1990 17:3015
    Yeah, that's what I wanna do - I don't care if the note is playing
    *while* I'm changing the sound.  I just want to play a note, flib a
    pit, and play the note again, with a different sound.
    
    I've found some notes elsewhere which might be useful.  And some which
    weren't.  Like the checksum calculator.  Actually, you can't just
    subtract the sum of the data from 0x80 (hex), because there are cases
    where the result is 0xFFFFFE0A.  All you *really* want are the lowest
    7 bits of the result, a very easy programming task.
    
    I'm off to "work the issue".
    
    As for reading the manual, I've tried that.  What I need is a
    Cryptographer versed in Rolandese...  Or another few weeks of no TV, no
    Church, No early-to-beds, and lots of early-to-rises.
1894.50smiley face insert hereSALSA::MOELLERNever trust a Prankster.Wed Jan 17 1990 18:298
    <<< Note 1894.49 by ATPS::GRASMANN "What got your node out of joint?" >>>

>...I just want to play a note, flib a pit, and play the note again
                                ^^^^   ^^^
    
    I'm unfamiliar with this particular technology.  Could you fill me in?
    
    karl
1894.51I've flipped a bit myself...CSMADM::MASHIAGo placidly amid the noise and haste.Wed Jan 17 1990 20:107
    RE .50
    
    Probably meant "flip a bit"...
    
    ...or did you already know that?
    
    Rodney
1894.52ATPS::GRASMANNOn the road... of happy destiny?Thu Jan 18 1990 18:276
    Oh well - I have an extended form of dyslexia...  What can I say?
    
    
    		Dyslexics,
    
    		  UNTIE!
1894.53<I hope this can be done>CACIQE::NUNEZEdgar NunezMon Jan 29 1990 21:1123
    I just got a MT-32 for my son and now I have a big problem in my
    hands. I know I should have read all the notes before buying it
    but such is life. 
    
    The problem is that I didn't know about the MIDI channel assigment
    to the different sound banks, and the keyboard that he has is a CASIO
    CT640 which only transmits in channel one. I tried with my keyboards
    using channel 2 and above and it worked beautiful.
    
    As I can recall from this topic it seem to be posible to reassing
    all banks to one MIDI channel "in my case channel 1" and by using
    the front panel changing the sounds. I don't have a PC to do the
    change via SYS-EX but I have a MMT-8 and it occurs to me that if 
    someone does the SYS-EX portion and dump it into the MMT-8 and 
    sends a cassette to me then I can load it to the MT32 and get my
    son out of my back.
    Are this changes permanent in the MT-32 or do they have to loaded
    everytime is power-up? If not permanent , I'm in deep 
    
    
    Edgar (who_loaned_his_JP-6_until_this_is_resolved)
    
        
1894.54It's actually pretty easyCSC32::MOLLERNightmare on Sesame StreetMon Jan 29 1990 21:528
	You can press a couple of buttons on the front (I think that it's
	the volume and part select button), then press the # 1 key. This
	will shift the whole mess (except for drums) down by 1. So you can
	use MIDI channel # 1 with the keyboard. If you are still having 
	trouble, send me mail & I'll dig it out. I rarely mess with this
	since I use the default assignments for Midi channels.

							Jens
1894.55Roland's book?SWAV1::STEWARTAs a matter of fact, it's all darkMon Apr 23 1990 14:3015



	I was looking through the latest RUG magazine this weekend when I
	noticed that Roland is publishing a book about the MT32.  Anybody
	seen it?  If so, does it explain the L/A voodoo? 

	Side issue: does anybody else get irritated by the thought that a
	manufacturer would charge extra for documentation needed to
	completely use a product? 




1894.56I got a copy of itCSC32::MOLLERHit by a truck, License # RDB31AMon Apr 23 1990 19:0013
>	I was looking through the latest RUG magazine this weekend when I
>	noticed that Roland is publishing a book about the MT32.  

	I got it. It's not much different than the users guide that comes
	with the MT-32, but it is better & more clearly written. It general,
	it doesn't tell you much more than what comes with the MT-32.

								Jens





1894.57Remember, we're talking about Roland - they ha have a rep to maintainDREGS::BLICKSTEINConliberativeMon Apr 23 1990 19:2114
>	Side issue: does anybody else get irritated by the thought that a
>	manufacturer would charge extra for documentation needed to
>	completely use a product? 
    
    In the case of Roland, I'd be overjoyed just to have ANY decent
    documentation at all, whether it's extra or not.
    
    But I do agree with the notion that it's wrong to charge extra for it.
    
    I suspect this arises from Roland-USA being a somewhat separate entity
    from Roland-Japan, and Roland-Japan not being much interested in
    english documentation.
    
    	db
1894.58What is your wishing fix?WEFXEM::COTEA friendly stranger in a black sedanMon Apr 23 1990 19:534
    
    Whating am wrongly for Roland documentation you wish?
    
    Edd
1894.59JVMS> COPY/WRONG_VERSIONSALSA::MOELLERPanic? I panicked an hour agoMon Apr 23 1990 20:105
   <<< Note 1894.58 by WEFXEM::COTE "A friendly stranger in a black sedan" >>>
>                         -< What is your wishing fix? >-
>    Whating am wrongly for Roland documentation you wish?
    
    Wonderful I am choking so full when being laughing...
1894.60on second thoughtCANYON::XEROXMon Apr 23 1990 21:281
    Never mind....
1894.61let's take lunchSALSA::MOELLERNever trust a Prankster.Tue Apr 24 1990 18:017
>                  <<< Note 1894.60 by CANYON::XEROX >>>
>                             -< on second thought >-
    
    CANYON::XEROX, could you identify yourself ?  It boggles the mind that
    there may be an entity in TFO that participates in this conference..
    
    karl in TUO
1894.62false alarm...CANYON::XEROXTue Apr 24 1990 22:2015
    
    
    
    
    
    No need to get excited...it's just John Stewart from SWAV1 whose home
    system is down at the moment.  You're still alone and out in the
    desert, Karl!  *8')
    
    John
    
    
    
    
    
1894.63Opposite of .53!MACNAS::SALLISONTue Nov 27 1990 06:526
    I have an MT32 which has been changed to the 1-8 MIDI channel
    arrangement. How do I get this back to the original 2-9 default? I have
    tried most combinations of buttons etc to no avail. Any help would be
    appreciated.
    
    ...Sean
1894.64Turn It OffAQUA::ROSTDrink beer: Live 6 times longerTue Nov 27 1990 11:2011
    Re: .63
    
    If shutting the power off doesn't restore it to 2-9, then it's
    broken...
    
    From my experience, that's the only way to do it.
    
    Am I missing something here???
    
    
    							Brian
1894.65DREGS::BLICKSTEINDaveTue Nov 27 1990 13:082
    No, there's an official way to do it, but as I sold my unit, I don't
    have the manual in front of though.
1894.66ALL RESET OK ?... 1COMICS::DSMMGRTue Nov 27 1990 14:365
    I sold mine too, but it was a two key function that brings up the
    message ALL RESET ? and then you press key 1 to signify that it is okay
    to go ahead.
    
    Jonathan
1894.67What should the display indicate?MACNAS::SALLISONTue Nov 27 1990 15:2413
    Maybe I am missing something here. I tried the ALL RESET function. I
    still see the message: CHANNEL (1-8)?  [1]
    
    Should I see CHANNEL (2-9)? after it has been set to 1-8?
    
    If not then maybe I don't have a problem at all. I will try sending to
    CHANNEL 9 and see what happens. This could just be inexperience since I
    only starting playing with the MT32 last night for the first time.
    
    But you must admit that the documentation you get with these things is
    somewhat lacking!
    
    ...Sean
1894.68Just Say NoAQUA::ROSTDrink beer: Live 6 times longerTue Nov 27 1990 15:3510
>    Maybe I am missing something here. I tried the ALL RESET function. I
>    still see the message: CHANNEL (1-8)?  [1]
 
    OK, you're all set...if you hit the "1" button, it shifts the channels,
    otherwise it does not.  But the readout stays there until you do
    *something* else.
       
>    Should I see CHANNEL (2-9)? after it has been set to 1-8?
    
    Nope.
1894.69Ooopps...OK nowMACNAS::SALLISONWed Nov 28 1990 08:4711
    Yuupppp!!!
    
    You are right. There is no problem in fact. The only nuance here is
    that you have to "hold-down" the MASTER VOL button to bring up the
    CHANNEL selection message AND then "let go" of the MASTER VOL to change
    the selection. I had not released the MASTER VOL button and hence the
    selection had not changed.
    
    You live and learn!
    
    
1894.70Programming MT32 on IBM-PCSVCVAX::RIVERAMon Dec 17 1990 19:1911
    I am not too versed musically, but I am a good programmer and I
    have a few simple questions on the MT-32.
    I have a MT32(LAPC-1) card in one of the slots in my IBM PC.  The
    card came with no documentation except for the exclusive messages
    and I am having a hard time accessing the card.  For example, I
    need to know what  pc port to use to communicate with the card,
    and what sequence of commands do I use to init and play sound out
    of the card and into my stereo.
    A point in the right direction would be GREATLY appreciated!
    Pete.
    
1894.71Might be a pointer to IBM/MT32VOLKS::RYENRick Ryen 240-6501 AET1-1/A6Tue Dec 18 1990 14:037
< Note 1894.70 by SVCVAX::RIVERA >
                        -< Programming MT32 on IBM-PC >-

See note 2487.4.

Rick

1894.72rhythm note number assignmentsMAST::GRUNDMANNBillWed Nov 20 1991 21:0785
    In case anyone needs this info (since I typed it in for Brian Rost
    anyway).
    
    MT-32 Rhythm instrument voice note number assignments
    
Default Key Bindings

	note
	number	voice
	
	35	Acou BD
	36	Acou BD		this is C3
	37	Rim Shot
	38	Acou SD
	39	Hand Clap
	40	Elec SD
	41	Acou Low Tom
	42	Clsd Hi Hat
	43	Acou Low Tom
	44	Open Hi Hat 2
	45	Acou Mid Tom
	46	Open Hi Hat 1
	47	Acou Mid Tom
	48	Acou Hi Tom	this is C4
	49	Crash Cym
	50	Acou Hi Tom
	51	Ride Cym
	54	Tambourine
	56	Cowbell
	60	High Bongo	this is C5
	61	Low Bongo
	62	Mt High Conga
	63	High Conga
	64	Low Conga
	65	High Timbale
	66	Low Timbale
	67	High Agogo
	68	Low Agogo
	69	Cabasa
	70	Maracas
	71	Smba Whis S
	72	Smba Whis L	this is C6
	73	Quijada
	75	Claves

Stereo Balance

LEFT	|	|	|	|	|	|    CENTER	|	|	|	|	|	|   RIGHT
7	6	5	4	3	2	1	0	1	2	3	4	5	6	7
|	|	|	|	|	|	|	|	|	|	|	|	|	|	|
|	|	|	|	|	|	|	O Acou BD	|	|	|	|	|	|
|	|	|	|	|	|	|	|	O Rim shot	|	|	|	|	|
|	|	|	|	|	|	|	O Acou SD	|	|	|	|	|	|
|	|	|	|	|	|	O Hand Clap	|	|	|	|	|	|	|
|	|	|	|	|	|	|	|	O Elec SD	|	|	|	|	|
|	|	|	O Acou Low Tom	|	|	|	|	|	|	|	|	|	|
|	|	|	|	|	|	|	|	O Clsd Hi Hat	|	|	|	|	|
|	|	|	|	|	|	|	|	O Open Hi Hat 2	|	|	|	|	|
|	|	|	|	|	|	O Acou Mid Tom	|	|	|	|	|	|	|
|	|	|	|	|	|	|	|	O Open Hi Hat 1	|	|	|	|	|
|	|	|	|	|	|	|	|	|	|	|	O Acou Hi Tom	|	|
|	|	|	|	|	|	|	|	O Crash Cym	|	|	|	|	|
|	|	|	|	|	|	O Ride Cym	|	|	|	|	|	|	|
|	|	|	|	|	O Tambourine	|	|	|	|	|	|	|	|
|	|	|	|	|	|	|	O Cowbell	|	|	|	|	|	|
|	|	|	|	|	|	|	|	|	|	|	|	|	|	|
|	|	|	|	|	|	|	|	|	|	|	|	O High Bongo	|
|	|	|	|	|	|	|	|	|	|	O Low Bongo	|	|	|
|	|	|	|	|	|	O Mt High Conga	|	|	|	|	|	|	|
|	|	|	|	|	O High Conga	|	|	|	|	|	|	|	|
|	|	|	|	O Low Conga	|	|	|	|	|	|	|	|	|
|	|	|	|	|	|	|	O High Timbale	|	|	|	|	|	|
|	|	|	|	|	|	|	|	|	O Low Timbale	|	|	|	|
|	|	|	|	|	|	|	|	|	|	|	|	O High Agogo	|
|	|	|	|	|	|	|	|	|	|	|	|	O Low Agogo	|
|	|	|	|	|	O Cabasa	|	|	|	|	|	|	|	|
|	|	|	|	|	|	|	|	|	|	O Maracas	|	|	|
|	|	|	|	|	O Smba Whis S	|	|	|	|	|	|	|	|
|	|	|	|	|	O Smba Whis L	|	|	|	|	|	|	|	|
|	|	|	|	O Quijada	|	|	|	|	|	|	|	|	|
|	|	O Claves	|	|	|	|	|	|	|	|	|	|	|
|	|	|	|	|	|	|	|	|	|	|	|	|	|	|
7	6	5	4	3	2	1	0	1	2	3	4	5	6	7
LEFT	|	|	|	|	|	|    CENTER	|	|	|	|	|	|   RIGHT