Conference napalm::commusic_v1

    Take patch 0 (or designate "some" MIDI program number) and write
    it into the standard that all conforming equipment will interpret
    that program number as some appropriate interpretation of "do nothing".
    
    Like for effects, it should go into BYPASS.  
    
    For SGU's it should be silence.
    
    If you don't want to use SYSEX stuff (I don't), to do this right now
    you have to create a silent patch, or an effect that still runs
    through the effect, but does not change the sound (which tends to
    be noisier than true "bypass").
    
    Uses:
    
    	1) Convenient for programming configurations
    
    	2) Easy implementation of "mute" 
    
    	3) Providese bypass accessible via program change which most
    	   sequencers, controllers, etc. can generate instead of
    	   SYSEX which many can't generate.
    
    Opinions?
    
    	db

    I think that if someone suddenly came down from heaven, waived a magic
    wand over all my MIDI gear and converted all my MIDI THRU's to MIDI
    MERGE, everything I've ever done would STILL work right and a lot of
    things I CAN'T do now (no, I don't have a DMC MX-8.... yet....)
    would become both "possible" and "easy".
    
    I guess I'm begging a question (or two):  Wouldn't MIDI MERGE be 
    more useful than MIDI THRU?
    
    Are there situations where you REALLY need to have THRU and not
    MERGE.
    
    Which would you rather have?
    
    	db

>.1

    Blasting a MIDI volume of 0 on a channel effectively gives you the same
    functionality.  I can't see where a "patch to null" command would be
    any better. 

>current topic

    I'd like to see mfgrs agree on how things are implemented.  For
    example, MONO mode.  Different implementation on Yamaha gear than on
    Ensoniq gear than on ... 

    I guess I'm more inclined to echo a recent Keyboard article, that there
    should be a ban on new equipment for a period of time, and that the
    mfgrs should use this time to fix bugs or shortcomings (or in Roland's
    case, zits 8-) in their product offerings.

-b

    > I can't see where a "patch to null" command would be any better.
    
    Do you have any MIDI effects receive and interpret MIDI volume?
    
    Part of the rationale is that almost everything with the word
    MIDI on it responds meaningfully to program changes.  Not even
    all synths respond to MIDI volume.
    
    Perhaps your thinking to much in terms of synths.
    
    Right now, a guitar player using a MIDI control pedal (almost all
    which generate nothing other than program change data) can not
    cleanly create a patch in which one audio processor (like his
    delay) can be made to go into bypass without (as I've said)
    programming a "by-pass patch".
    
    The basic problem is that most controllers support only program change,
    not SYSEX.   Certain basic functions of things like effects should
    be accessible via program change and I think the cleanest way to
    do this is by making a universal "convention" of patch 0.
    
    Also, MIDI is going to be used for a whole lot more than synths
    and effects.  There are already lighting systems that can be driven
    from a MIDI sequencer, mixing boards that are driven by MIDI, 
    teleprompters, and god only knows what else they'll come up with.
    
    	db

    OK, having thought about it some more I found a more general way to
    state my real intentions.
    
    I hope we all agree that Sysex is "never safe" and in general a lousy
    way to have to do something, although sometimes it has to be done
    that way.
    
    I feel that "common operations" should get distinguished (non-Sysex)
    treatment in the standard.  The first such common operation is
    "bypass".  There should be a common way to tell MIDI devices to
    "have no effect".
    
    Having thought about it, I can give you another use for the patch
    0 proposal: panic buttons.  It would sure be handy to have a button
    that turns all my effects off.   Note that with my proposal, that
    button could even be hard-wired.
    
    	db

    SYSEX is a perfectly reasonable deal.  GPIB IEEE-488 uses all
    special codes for each instrument.  It's the window to the future,
    the flexible escape for innovation.
    Or maybe it's a rathole, beats me.
    Tom

    Ok, Dave - point taken, but I think that "patch to zero" is the wrong
    way to do it.  I won't argue, though. 

    I'd like to see the 16 channel limitation removed.  It's to the point
    where 16 channels ain't even close to enough.  For example, I have two
    ESQ-Ms (each responds to 9 channels) and one FX box per module.  There
    go 20 channels. 

-b

    > Ok, Dave - point taken, but I think that "patch to zero" is the wrong
    > way to do it.  I won't argue, though. 
    
    
    I encourage disagreement.  I'd like to hear of a better way.  I also
    have certain philosophical problems with the "patch to zero" method
    but I just couldn't think of any better way that would work with
    existing equipment.
    
    	db

    ...seems to me you folks aren't adding anything to the protocol
    per se, but rather are suggesting standardized implementation of
    features.
    
    Yes?
    
    Edd

    re: .9
    
    Well sorta, but not really.  That is, an implementation would not
    be "conforming" at the new level if it did not implement patch 0.
    
    So, it could be done as a "convention", however I would like to
    go beyond that and see the convention "enforced" via validation
    in the standard.
    
    I base this on observation.  MIDI has demonstrated itself to be
    an excellent standard.  There's far more compatability between
    brands than I usually associate with a standard (of course most
    of my standards experience is in the computer software, particularly
    computer languages area). 
    
    Sure there are the occasional glitches (like... say... sequencers that
    UNFORTUNATELY are constantly sending out clock signals that screw
    up step modes in certain popular drum machines  ;-) ), but generallly
    if it says MIDI it will work quite well with anything else that says
    MIDI.
    
    It's the things that aren't (yet) in the standard which have seemed
    to cause problems (the various incarnations of MIDI modes, etc.)
    
    I think in order for these features to work well between brands, they
    have to be IN the standard.
    
    	db

    another vote for the 'patch 0' = channel off trick.  Also, some
    SGU's (like the Kurzweil 1000PX) require a MIDI patch change number that
    is ONE LESS THAN THE ACTUAL PATCH NUMBER.. therefore 'patch 0' is
    actually gonna trigger patch 1, the famed Kurzweil Grand.. big deal,
    I'd like it to be a null patch that tells that SGU's receiving MIDI
    channel to shut up.
    
    And yes, this is not exactly an addition to the protocol per se.
    
    karl

    > Also some SGU's require a MIDI patch change number that is ONE
    > LESS THAN THE ACTUAL PATCH NUMBER.
    
    Another reason to support this change.  Perhaps taking up patch 0
    will discourage implementations from doing that.  ;-)
    
    BTW, the HR-16 does this with pattern #'s.  For those of you without
    HR-16's (if there are any of you w/o HR-16s), MIDI program numbers
    can be used to select patterns (sorta like calling up a different
    drum kit patch among other things).  And it forever confuses me that
    to get HR-16 pattern #12 I have to send MIDI program #11.
    
    Or was that to get pattern #12 I had to send #13?  ;-)
    
    Or maybe it was that...
    
    	db

One feature that I have thought about recently is the ability to send
a "change MIDI recieve channel" message. (I don't think this exists, does it?)

Would this be useful?   I dunno.   Sounds pretty dynamic though.

/Mitch   :-|