Conference napalm::commusic_v1

    lotsa good questions. ASSUMING you have more than 1 keyboard: 

    Not what Ive DONE, but what I WOULD do:
    
    Equip: MORE than 1 controller (keyboard).  These need to be easily
           assignable to various modules (incl. themselves) so some
    	   sort of midi switcher needed. Now you have network flexibility.
    
    	   NO keyboard controller is the master for patch switching.
           Not only can it take too long, but FB01-like (the future!) boxes
           have setup info needed beyond midi Program Change command. SO,
    	   for patches Id use a micro (guk) OR a sequencer to send out ALL
    	   configuration and voice info. You MIGHT even want to change
    	   something a few times DURING the song! Gasp!
    
    	   "Deselecting" can be moving fader to off, but why not change
           those modules not needed to recieve on a non-used (used for
    	   silence) midi channel.  OR, have 1 SILENT patch on your synth.
    
	   Yes, you do have to work out the connections and voice
           selections beforehand, just like its part of the song.    
      
           what did I leave out that this approach doesnt cover?
    
    	ron
    
        
    

    
    oh yeah, SAMPLES.
    
    er.......
    

    Hmmm....
    
    Let's assume 15 songs * 4 sets per night (60 per night)
    
    I give each song a number, 1-60. Then I put together a step-time
    sequence of nothing but patch changes. Song number one, measure
    number one, song 28, measure 28. Even my QX7 will FLY through
    measures. Pick a song, advance/'rewind' to the appropriate measure
    and press 'Continue'!!! Voila!
    
    I like the idea of mulitple controllers/keyed synths. I can't imagine
    having a 6' rack full of the newest whiz-bang MKS-100,000s, Fb0n's,
    multiple multisamplers, etc. and a bad controller.
    
    Do you find yourselves changing patches often? Or do you have a
    handfull of general purpose sounds?
    
    Do you change patches on *all* your modules or tend to leave 1 or
    2 on 'general' patches (piano, B3) and bob around the others?
    
    Edd

    Good questions. Can I assume you gott the gigg, Edd?
    
    Centralized control of patch/config changes for multiple synths,
    many times during a set.. I used to think that a good keyboard
    controller would deal with any possible situation. One of the 
    problems with the KX88 is that it only xmits on two channels,
    one of which (in my setup) has a volume pedal associated with it.. 
    any other synth (not on these channels) set in OMNI mode (to play at 
    all) is gonna get two notes for every one played, robbing polyphony. 
    On the other hand, in order to receive discrete patch change info, 
    these add'l synths MUST be set to their OWN MIDI channel in POLY mode, 
    shoving the 'xmits on only two channels' problem back in my face.
    
    I begin to see why you want multiple keyboards, although it offends
    my sense of esthetics.
    
    So having one sequence filled with patch change info is interesting..
    but what if you want to actually USE the sequencer to play 'extra'
    parts ? (I know, that's cheating) Now each sequence has a bar in
    front with all patch change info. Fine. Units like the MKS-20 can't HAVE
    a 'silent' patch. So down come the faders OR if you have a MIDI
    switcher, switch it manually to make it 'listen to' a silent controller.
    
    This is why I posted a question about MIDI patch change footswitches
    a while ago. We've got this data transmission medium, which works
    FINE in a quiet, controlled studio environment, but a bit less well
    for live performance with multiple synths where lots must be done
    quickly in less than optimal conditions. A lot needs to be done,
    as we add synths with nonstandard bank/patch/config numbering schemes.

    karl    

    Wow, exactly the questions I've been contemplating lately!  (I've
    been away from notes for a while).
    
    Things used to be so easy!  I had one (1) dx7, and one (1) juno-60.
    No midi connections or anything like that.  I would just organize
    the patches I needed for each song in common banks, on each synth,
    and hit patch changes on each synth individually on the fly.  This
    was pretty easy for most songs, but a real pain on complicated ones.
    
    I added the MKS30 to be played through the dx, so that wasn't much
    different -- I just had the dx send patch change info to the module
    and but the correct patches in the appropriate locations on the
    module.
    
    But now I'm replacing the Juno-60 with a D-50, and want to automate
    my whole system (Ideally, hit one button to re-configure all synths
    for a new song, and then step on a footswitch at the correct time
    in the song to get a new sound, on the correct synth.  Hopefully
    the device will be "smart" enough (can be programmed) to change
    synth A's patch without changing synth B, thus eliminating duplication
    of patches in a synth).  
    
    Well, anyway, I think the devices that will approximate this are:
    J.L.Cooper Midilink (i think), Voyce LX4 and LX9, Meico Patch Commander
    and Midi Commander, and Axxess Unlimited Mapper.
    
    Unforunately, not many places in the area carry this stuff (well,
    I didn't check Boston), so I didn't get to check any of it out.
    I do have the J.L.Cooper thing on order, though, so I'll give a
    report after I get to play around with it.
    
    Later,
    
    Ron
    
    
    

    
    I dont see why we are not just using a sequencer to send out
    prog change and sysex to the units. That will change patches.
    
    The other problem is the connection of the performance device(s)
    (keyboard(s)) to the units, and that *seems* to require some
    midi switcher. This TOO could be programmed ffor a configuration
    via the sequencer.
    
    what am I missing?
    
    rr 

    Uhm, I think the problem may be that patch changes are sent by channel,
    ]and if you have multiple synths on the same channel (for layering),
    they will all go to the same patch, requiring you to allocate your
    patches the same way in multiple synths, not difficult but annoying.
    
    Is there some way to send patch changes using sys ex encoding so
    only a specific synth on a specific channel responds?
    
    len.
    

    re .-1
    
    I don't think so.  However, you can bypass the problem on certain
    synths (KORG DW8000, JUNO-106, DX7 from my experience) by actually
    storing the entire patch SYS-EX in the sequencer.  The synth will
    update its current-patch buffer with the received patch if and only
    if the channel and synth-type match the codes in the first part
    of the patch SYS-EX.
    
    If you were willing to dedicate a sequencer to this function, you
    would also save yourself patch memory limit problems.
    
    Guido

    
    WHoa. Program change has channel info in it.
    
    The sequencer 'plays' the program change messages to each channel.
    Boom. 
    The sequencer 'plays' SYSEX messages on whatever channel has something
    (like FB01) that requires other voice and/or performance setup protocol.
    Boom.
    That leaves "network configuration" (if youre REALLY fancy), which
    would be the sequencer 'playing' the change commands to your midi
    switcher (which ins connect to which outs). Or maybe this is first.
    
    Big deal.
    

    There is a little problem with that scheme--SYSEX data is not
    channelised.  Or not standardly.
    
    For example, I know that the Juno-106 puts out a channel number as part
    of its SYSEX message, but the sequencer wouldn't understand that, so if
    you wanted to, later, put your Juneau on another channel, you would
    have to make a new recording of the SYSEX dump which was on the new
    channel.  And I imagine that some equipment doesn't even bother to put
    a channel ID in its SYSEX dump.  Imagine having two identical pieces of
    equipment on the same wire...  The best you could do is set them
    identically, no matter what channel they were set to recieve. 
    
    There's no substitute for enough internal memory.
    
    Steph
    

    This is a little bit outta the flow of this conversation, but so
    what...
    
    I had an idea that would enable people like myself who have sequencers
    that only hold one song to expand that by 16X. Admittedly, it won't
    work when trying to emulate the BSO on stage, but if you're using
    a sequenced synth instead of a bass player, you could fit your whole
    set in one sequence. An MC500 could hold the whole night's bass
    parts... And it's so easy....
    
    Just do the bass line for every song (up to 16) on a different MIDI
    channel! Merge 'em all together and change receive channels according
    to whatever song you want to do.
    
    Good idea, no? 
    
    Send gratuities to me at home. Thanks!!
    
    Edd