Conference napalm::commusic_v1

    Well there are things out there that have more than 8 voices, but 8
    sure seems to be the standard.  I am somewhat surprised that despite the
    rapid pace of technology, we still have been "stuck" on 8 for so long.  
    In my experience, you really need more.
    
    Roland MT-32 has "up to" 32 (depends on complexity of the patch), Roland
    S-550 (sampler) has 16.
    
    	db

    Humph. The MT32 seems better than straight 8-voice, even if only
    simple patches count as 'single-voice'. 
    
    But to add to .0: surely, more timbres also means more immediately
    comples sounds, sinc voices that can fire simultaneously over MIDI
    can surely be stacked internally. I imagine you could create a knockout
    8-voice performance patch with a 32-voice multitimbral synth in
    this way.
    
    Richard.

    True (about the MT-32) but I think realistically you can expect
    to be able to have more than 8 voices available for most applications.
    Also, the drums seem independent of the partial allocation so they
    also count in a lesser sense.

    The drums on the MT-32 are independant, as far as the other patches
    go. I'd say 16 to 20 is average (at least that seems to be what
    some of my sequences are using). The thing I like about the MT-32
    is it offers similar patches using various quantities of layers,
    so, if you find that you need to tweek a sequence, you select a
    simpler version of a sound & save yourself a layer or two. 8 voices
    are fine if you are playing live. If you are sequencing, you tend
    to run out of resources pretty fast.
    
    							Jens

    
    However, while the MT32 can supply "up to" 32 voices, it still
    only handles eight *simultaneous* timbres.
    
    Do any of the 16 or 24 voice boxes out there support more than eight
    timbres???
    
    The number eight itself is no suprise, after all most synths are
    computer based now and orienting things in byte-sized chunks makes
    sense.
    
    

    Until .5, I was impressed by the MT32.
    
    Definition of terms required. Which of the following are true:
    
    32 note polyphonic + monotimbral  = 32 notes (one patch) sound 
    simultaneously
    
    32 voice poly + multitimbral = 32 notes (up to 32 patches) sound 
    simultaneously
    
    By the way, there are loads of computery numbers -- shame they picked
    8!
    
    
    Richard.

    The EPS sampler is up to 20, or as few as one.  You pays your $2K
    and you takes your choice. 
    
    Dynamic allocation of the 20 voices, of course.  Each patch takes
    as many as it wants.

    8 seems like an entirely reasonable number for multi-timbrality.
    Remember MIDI essentially limits you to 16 timbres anyway, and
    since you can do patch changing via MIDI the 8 timbres limit only
    means that you can't have more than 8 timbres at any one instant,
    but you can have hundreds of timbres during any one song.
    
    Hardly seems limiting to me.
    
    The timbrality seems more limited by 16 MIDI channels than anything
    else.  So with 8 + drums the MT-32 gives you something that can
    fill slightly more than half your timbral space anyway.

    I suppose so. So a good combination would be 32 note poly, 8 voice
    multitimbral.
    
    Richard.

   
   The slick MIDIOT can double the timbral space on some synth modules
   that allow odd and even note seperation and scale  definition.  For
   instance the TX-81Z.  Have one TX-81Z respond to  only  even  notes,
   but change the scale so that your standard twelve notes are  spread
   out across 24,  do the same to a second but have it respond only to
   odd notes.
   
   You decrease your range by half,  which is okay for most orchestral
   intruments,  and can only seperate the note on-off data,   but  you
   CAN have two timbres receive on the same channel this way.
   
   Note also that your score will be somewhat *^&$%# up  but  what  do
   you want for a $350 solution.
   
   _sjz.

    Another solution for more than one timbre on a channel at one time
    is to set up a split point with two separate patches. Notes going
    to one section are played with one patch and those going to the
    other section get played with the other patch.
    
    chad
    

    This has been mentioned before, but with the TZ you can make each
    voice respond only to a certain range.  You can also put a note
    shift on each voice to compensate (+/- 2 octaves, as I recall).  There 
    are 8 voices, so if you split a keyboard into 8 sections and shifted 
    the notes as appropriate you could independently control all 8 voices,
    each over about an octave range, by using different keys and transmit 
    this all on one channel.  If you are sequencing and are limited
    to 16 MIDI channels, you could tie in 16 TZ's on the same bus and
    run an orchestra of 128 mono voices or 512 operators.
    
    
    Steve
    

    Yes, the Kawai K5 and K5m can produce up to 15 timbres simultaneously.
    
    Why 15?  Because 5 lines fits on the editing screen the way they
    have it layed out, and there are three pages for each screen.
    
    Each timbre may produce a fixed or dynamic number of voices.
    
    Steph

    
    As for why it would be nice to have more than 8 timbres simultaneously:
    
    Layering....
    
    The MT32 architecture is ill suited for this (unless you have a
    computer hooked up to it) due to the oddball MIDI channel assignments.  
    
    But by placing more than one timbre on each MIDI channel for layering
    you can easily find use for more than 8 timbres.
    
    
    
    

    
    
    re: Bill and your comment about the EPS
    
      The EPS, while it does 20 'voices', still only allows up to 8
    simultaneous timbres (instruments) at a time, memory permitting.
    The 20 voices are dynamically assigned, as available, to whatever
    instruments need them. You may bias the assignments somewhat by
    setting an instrument to only want a certain number of voices (the
    'set number of voices' prompt)... 
    
      So there's another example of the magic number 8. Of course, the
    8 are all it holds at a time anyway. Not too shabby for a $2K sampler.
    I have no idea why it only deals with 20 voices, but this is a sampler
    with 13-bit i/o, 16 bit storage, a 32 bit processor, 24 bit floating
    point data munging, and a bunch of other crazy numbers that don't
    make sense.
    
      It even allows for 10 outputs...
    
    /pjh

    But you can multisample across the EPS keyboard, so one instrument
    might have more than one timbre (look at the "drum kit" instrument
    :-) )
    	     
    So you can have what sounds like 20 simultaneous timbres...
    
    	-Bill

    
    The latin percussion is better.
    
    Yeah, fair enough, but hacky anyway. You end up splitting your
    instrument into smaller pieces. That's like having a 'sax section'
    made up of alto, tenor and baritone... eats memory, but doable.
    
    /pjh

    The MIDI limit of 16 simultaneous timbres is probably not a serious
    limitation for pop applications, but for "classical" music it's just
    barely adequate.  I'll check some scores tonight to get some numbers
    on what really happens in "representative" orchestral scores (e.g.,
    a Beethoven symphony, a Mahler symphony, a Strauss tone poem,
    Schoenberg's Gurrelieder, a Chopin etude (joke, ha ha!)).
    
    You can get around this limit by running multiple MIDI busses from
    multiple sequencers synched together.
    
    len.
    

    The Kurzweil 1000PX gens 24 voices simultaneously, over 16 timbres
    max (because of the 16 MIDI chan limit, natch).. nice.

    re .18
    
    There are several multiple-out midi interfaces available for software
    sequencers.  C-labs Creator can use 64 midi channels - 1-16 x 4
    busses.  A reviewer claimed he would need this with all his synth
    gear, effects devices and his light show all sequenced.
    

    re. .0 and .13
    
    	I was wondering if anyone was going to mention the K5! Steve
    had one before I did, but in the past 6 months or so I've had a
    ball with it.  There ARE 16 voices available to sequence or play.
    As Steve mentioned you can only STACK 15 of these at a time in what
    Kawai calls "Multi" patches.  But, if I understand the definition
    of the term "multi-timbral", the K5 will give you twice the 8-voice
    limit you now face.
    
    re. .18
    
    	All of the above is great when working on classical music. 
    I have been sequencing a lot of baroque chamber music (Handel's
    Concerti Grossi, etc.).  Having 16 parts available means that with
    just the one K5, I can cover all the voices and still have a few
    left over for continuo, voices, percussion, etc.. At the risk of
    sounding like a Kawai commercial, they have just come out with a
    new synth, the K1, which on initial hearing sounds good and has
    percussion (which the K5 did not). (I am thinking of getting the
    K1 in a module which should give me more voices than I know what
    to do with).
    
    Bill