Conference napalm::commusic_v1

    < Note 1615.0 by SMURF::NEWHOUSE >
>I would like to know what rackmount units are available that will support
>4 to 8 simultaneous instrument voices (each a seperate midi channel)
>playing 4 to 16 simultaneous notes in each voice.   
    
    As specified, it's simple.. there is no single unit that has the
    polyphony you require.  Most rack units, synth or sampler, can generate
    8 or 16 simultaneous notes, dynamically allocated to patches or
    voices, receiving the data from multiple MIDI channels.  A reasonable
    low-end (~$500) unit is the Roland MT-32, with lots'o'voices.
    
    A higher-end unit, a rack sample player, is the kurzweil 1000PX
    with true 24-voice polyphony.
    
    karl

RE < Note 1615.2 by SALSA::MOELLER "DECblocks Product Support" >

	Karl,

	Wouldn't the TX816 fit this bill?

	Dan

	Sorry, didn't mean to sound stuffy, I don't even know...
But from what I've heard, it will.

Mike

RE < Note 1615.4 by SALSA::MOELLER "DECblocks Product Support" >

	O.K., I'll tell you about it, then... 8^)

	The TX816 is a rack-mount package Yamaha put out a number of years ago
that was, in essense, 8 TX7's in one.  It had eight TF-1's, which were ganged 
together.  Each TF1 could play its own patch (6-operator FM synthesis) with
16 notes polyphony, or all could play the same monster patch.  I once heard
a demo of a Yamaha rep playing (another) Rhodes patch, but each module in the
TX816 was responsible for a certain component of the overall sound.  It was
very impressive.  

	Dan

>  '4-8' voices EACH playing 4-16 (?) note polyphony

The D110:

A voice is made of 1-4 partials. 1 for the poor ones, 4 for the best ones.

On up to 8 "instruments" you can have up to 32 partials dynamically allocated.

You have to do the math from there.

In practice, I usually use the 3- and 4-partial voices, and I usually use about
3 monophonic instruments and 1 polyphonic instrument.

So assumming the 4-partial voices, that gets me 12 partials for the mono
instruments, leaving 20 for the poly instrument. That is, 1 note for each mono
instrument and 5 notes for the poly instrument.

Then add in the entire 5-octaves worth of drum sounds that seem to be always
available.

See note 1396.* for more D110 info.

/Mitch

>In practice, I usually use the 3- and 4-partial voices, and I usually use about
>3 monophonic instruments and 1 polyphonic instrument.

Actually, I really should say that in practice: because there are usually
3-partial voices involved, and because usually not all 3 mono instruments
are played at the same time, (and of course because of the dynamic allocation)
I have yet to be impacted by the 32-partial limit. I have not found notes
being "cut-off" yet at all. So far, it hasn't limited me in what I do with it.
THAT is a wonderful feeling and the desired result.

Add to that the fact that you can change the voice being played in an
"instrument" at any time during a sequence with a program change message, so
that throughout the course of a song, you can utilize MANY different
instruments, easily.

/Mitch

    Thanks!   I'm going to check out the S10, S110, and the MT-32.
    If you want to list any more units feel free, we better keep the
    discussion on details down - and do that in the notes used for the
    specific units.  Thanks again!
    
    Tim
    

    Bear in mind the comments a couple of notes back on dynamic voice
    allocation when you compare units -- not worth buying without it.
    
    Richard.

    re .10 dynamic allocation. Definitely agree this is a good thing
    to have but may not be necessary. I use a TX802 as a basic workhorse
    that doesnt have DVA (dynamic voice allocation) and I couldnt do
    without it. Certainly worth checking out if you have the bucks.
    

    Aren't the Roland D-family the only ones that have DVA?.

ESQ-1's and ESQ-M's do (have DVA).  I imagine the same holds true for the SQ-80.
If you have an ESQ-M attached to your ESQ-1, with the same patches loaded, it
will use "midi overflow" to automatically do DVA across both synths.

							-Jeff

    Actaully, there are many variations of DVA on the market.  Roland's in
    the D10 series is kind of brain-damamged (from what I gather) and is
    not "truly dynamic". 

    Ensoniq uses a scheme whereby it has a pool of available "notes"
    (note=3osc group).  If a note is not currently sounding, one of these
    is used.  If a note is currently sounding, that same note is used
    again.  If no notes are available, the oldest note is stolen. 

    Korg uses a similar setup in the M1, but note=1osc in this case.

    The new Yamaha V80 and the Emu Proteus also support DVA. 

    This feature is a major criteria of mine when it comes to purchasing
    new synths.  If it ain't got it, I don't buy it. 

-b

> < Note 1615.14 by DYO780::SCHAFER "Brad - back in Ohio." >
>     Roland's in
>     the D10 series is kind of brain-damamged (from what I gather) and is
>     not "truly dynamic". 

The D110 has true dynamic voice allocation.  I can't figure out what you
     might be referring to. Any further explanation?

/Mitch

    re .14
    
    Yeah...why is ROLAND's implementation of DVA "brain-damaged"?  DVA
    on my D110 appears to work the same as DVA on my KAWAI K5.  Is it
    also in need of a lobotomy?
    
    Curious (and worried),
    Bill Allen

< Note 1615.16 by MUSKIE::ALLEN >
                         -< More medical expenses...? >-

>    Yeah...why is ROLAND's implementation of DVA "brain-damaged"?  DVA
>    on my D110 appears to work the same as DVA on my KAWAI K5.  Is it
>    also in need of a lobotomy?


Bill,

I agree with you. The D110 DVA is not "brain-damaged". It is, in fact, much
more complicated than most, and therefore since it is so arcane, it appears
"brain-damaged". After all, they thought Einstein was retarded! 8^))
I found the documentation very poor on exactly how to
access all those partials, timbres, parts, voices, instruments, banks,
parameters, etc. etc. etc. That's one of the reasons I sold it. I just don't
have time to futz with all that stuff. I want my replacement for the D110 to
"Gimme an oboe on channel 1, a bassoon on channel 2, a clarinet on channel 3,
a solo violin on channel 4, a cello on channel 5, a viola on channel 6"
and take care of the voice allocation dynamically. I don't want to have to
deal with all those structures and hierarchies. I hope the Proteus does that.

dave

    I'm just repeating what I've heard, guys.  I've not done an indepth
    study of Roland's implementation of DVA (although I intend to be doing
    so soon).  Another fellow I know who had used both the D110 and the ESQ
    said that the ESQ was decidedly better on its DVA scheme. 

    I didn't mean to spin anyone up.  Honest.

-b

< Note 1615.18 by DYO780::SCHAFER "Brad - back in Ohio." >
                  -< I'll check it out when I get a chance. >-

>    I didn't mean to spin anyone up.  Honest.

Brad,

I don't think anyone is spinning. It was a provocative thought (which
we could use more of...it's been dull in here lately) and just gave us
something to blab about finally.

tgif,

dave

Without trying  to sound "defensive", here is some information about the
     Roland D110...

> < Note 1615.17 by SUBSYS::ORIN "Quid, me vexarius?" >
> I agree with you. The D110 DVA is not "brain-damaged". It is, in fact, much
> more complicated than most, and therefore since it is so arcane, it appears
> "brain-damaged".

The D110 DVA is neither brain-damaged nor complicated nor arcane...   it
     is  invisible.    There  is nothing to it.  By default, the factory
     patches are  set  up  to  steal the first note played.  This can be
     modified to do things like reserve partials for specific parts.

Other aspects of the  sound  architecture,  however, can be very arcane,
     but do not need  to  be.  You can read about the D110 in note 1396,
     but here are a few comments...

> I found the documentation very poor on exactly how to
> access all those partials, timbres, parts, voices, instruments, banks,
> parameters, etc. etc. etc. That's one of the reasons I sold it. I just don't
> have time to futz with all that stuff. I want my replacement for the D110 to
> "Gimme an oboe on channel 1, a bassoon on channel 2, a clarinet on channel 3,
> a solo violin on channel 4, a cello on channel 5, a viola on channel 6"
> and take care of the voice allocation dynamically. I don't want to have to
> deal with all those structures and hierarchies. I hope the Proteus does that.

Dave, I agree  with  you 100%.  The documentation stinks.  I also refuse
     to spend any time  dealing  with complex "set-ups", button pushing,
     etc.  I demand "virtual instruments" (just like you described).

I find it unfortunate that you  sold  your D110 without discovering that
     this is exactly how you CAN  use it.  - I use the D110 as 8 virtual
     instruments,  plus  a  percussion  set.    I  sit  down    at    my
     controller...and  say..."gimme  a  horn  on  channel  2, a synth on
     channel  3,  and  a  'seashore'  on  channel  7"...exactly  as  you
     described.   It is easily demonstrated -  to  do  this  you  simply
     ignore the "patch" level of hierarchy, keep the  thing in PLAY mode
     and call up  different  instruments  ("timbres")  on  each  channel
     ("part").  And none of this has to do with DVA - that is invisible.

The U110 is similar, but has fewer levels of hierarchy, has only 6 parts
     - and it has a wart with respect to the use of level envelopes (see
     note 1671).  (don't they all have warts?...:-)

/Mitch

    re : last few
    
    	Using a patch editor like Dr. T's D110 PE for the ibm-pc, also
    makes this "put a kazoo on 2..."-process easier.  Of course you
    have to spend $100.00 or so to get this privelige but you get some
    other bennies with it.  
    
    	I (like the rest of humanity) am anxiously awaiting the coming
    of the great PROTEUS (Proteus...proteus...pro...).  I have sort
    of pre-given it to myself as this year's X-mas gift.  That way I
    can snap it up whenever one is available :-).  Now if only somebody
    had one...
    
    Clusters,
    Bill Allen

> < Note 1615.21 by MUSKIE::ALLEN >
>     	Using a patch editor like Dr. T's D110 PE for the ibm-pc, also
>     makes this "put a kazoo on 2..."-process easier.  Of course you
>     have to spend $100.00 or so to get this privelige 

I don't mean  to  draaagggg  this  out  too  much, but the conference is
     getting reeeall sloooww and this is one of my gadgets soooo....

I can only imagine  a  PE  making  the  "put  a  blip on 2"-process more
     difficult.  The way I use the D110,  it  CANNOT  be made easier.  A
     button press on my controller brings up an instrument  on  whatever
     channel  I'm  transmitting  on.    How  can  that  be made  easier?

(Don't mean to sound harsh, just talkin')

/Mitch

Mitch,

This is an interesting discussion. Please describe your "controller" and
how you use it to "A button press on my controller brings up an instrument
on whatever channel I'm transmitting on." This is what I need, but haven't
spent the time to really dig into the guts of the Kurzweil MIDI board yet
to figure out how to do it. I assume it is done via sysex? Do you define
a "controller macro" or is it done on your Mac?

dave

> < Note 1615.23 by SUBSYS::ORIN "Quid, me vexarius?" >
> Please describe your "controller" and
> how you use it to "A button press on my controller brings up an instrument
> on whatever channel I'm transmitting on."

I use an MKB-200.  I keep only one "virtual instrument" ("part") on each
     channel.  I send program change messages  by pushing the bank/patch
     buttons  on  the  controller.    To get a new instrument on  a  new
     channel, I switch the transmit channel and press another bank/patch
     button.  No  sysex,  no  computer,  no macros.  Each program change
     message calls up a  "timbre"  on  the D110 (each timbre corresponds
     directly to a "tone" -  I ignore the "patch", "tone", and "partial"
     levels of hierarchy).

One of the keys to all of  this  is making sure only one "instrument" is
     on  each channel.  Once this is set-up, you should  never  need  to
     touch  the  front  panel  buttons  of  any of the rack-mount sgu's.
     Everything should be  accessible  through  the controller via patch
     change commands.  I hate using front panels.

32 MIDI channels might help if you have alot of multi-timbral sgu's.

/Mitch

> > < Note 1615.23 by SUBSYS::ORIN "Quid, me vexarius?" >
> > Please describe your "controller" and
> > how you use it to "A button press on my controller brings up an instrument
> > on whatever channel I'm transmitting on."

Ah.  I see where maybe some of the confusion  lies...   I shouldn't have
     said   "...brings  up  an  instrument..."  -  I  should  have  said
     "...brings up a timbre...".

The "instrument" is always there, waiting to be played  or  to receive a
     program change message.

I don't have more "virtual instruments" than I do  MIDI channels, so I'm
     still not  saturated  -  no  more  than one instrument per channel.
     Another key to  this  style  set-up is that your instruments are on
     predetermined channels (eg.  FB01  on  channel  1, D110 on channels
     2-10, U110 on channels 11-16).

/Mitch

    Sheesh - same way the ESQ works.

    This would make a *lot* more sense if mfgrs would standardize their
    terminology.  I had to read your last reply two or three times to make
    sense out of it, Mitch. 

-b (who owns no Roland SGUs)

re .14

The ESQ has a mode where it will reassign a new "note" for every note-on, 
regardless of whether that note is already sounding.  It's storable per patch,
so you can have some voices that do, and some that don't.

					-Jeff