Conference napalm::commusic_v1

    > How are people coping with these problems?
    
    With pure joy!!! Part of my sys-ex string sets levels for all the
    SGUs, *within* the SGU. Absolute heaven for me would be to have a board
    with no faders...
    
    I leave all the channels on my board and use MIDI to run the mix.
    
    FX are a slightly different story. MY MVII is only "patch addresable",
    but my DEP-5 has all tweakable parameters.
    
    Edd

    > How are people coping with these problems?
    
    Huh?
    
    What "problems".  What we had before was "problems".  I view what
    you are referring to as a blessed "solution".
    
    What "problems" are you having????
    
    	db

    Most of the high-end PC sequencers these days will put up some kind of
    slider control that sends CC information (or other parameter). 
    'Performer' lets you customize whole banks of these things.  'Vision' gives
    you 32 of them.  No messing with SYSEX is neccesary unless your SGU
    will not map a controller to the parameter you want to tweak.
    
    You drag them up and down with the mouse and the SGUs respond.  If you
    have recording turned on, all your moves are captured in the sequence.
    (And on playback all the little sliders move by themselves.  Spooky.)
    
    You usually can also have external real knobs (foot pedals, mod wheels,
    the 'bend' output from your Midied musical saw, etc) map into these
    on-screen faders.  So you can do things like make the mod wheel on your
    keyboard control the volume of some SGU, and record the adjustments
    as part of the performance.

    pointer to topic 1189, Mixing with MIDI
    

    I think Brian has a valid question, and I suspect people who don't have
    problems with this either don't have multiple synths to coordinate, or
    are doing this is a studio environment. If you have a lot of gear with
    multiple outputs and you have to do a live mix for different rooms it
    can definitely be a problem.
    
    I am using two synths, both of which have mono and stereo audio outs,
    and one of which also has six assignable output ports. I use only the
    stereo outs on both, and feed them into a 4-channel mixer. I set the
    volumes on the synths to fixed levels. It's important that I maintain
    these same levels all the time (at least in a relative sense),
    otherwise the levels that I set on individual parts won't give the mix
    that I want. I set the levels of individual parts using CC sysex
    commands. At that level the mix is "canned".
    
    The control that I have through the mixer is not over individual MIDI
    channels, only over the relative levels and panning of the two synths 
    (again, the panning of individual parts is preset with CC commands).
    I also use eq in the mixer to make adjustments for environmental
    factors.
    
    I don't use effects on my MIDI output at all. (That's not completely
    true - the synths have built in reverb, which I leave on a particular
    setting for all parts all the time).
    
    Ideally it would be nice to have live control over the level, panning
    and eq of each MIDI channel. Using a synth with separate outs for each
    channel, along with a multi-track mixing board would make this
    possible. I know some people who are in fact moving in this direction,
    but I have decided that for me it is not practical, for the following
    reasons:
    
    a) I don't have a sound engineer. I'd have to set things up once and
    leave them that way for the rest of the performance. So a lot of the
    benefit of doing it dynamically would be lost. I'd also have to have
    the same instruments assigned to the same outputs all the time. This
    would be a hassle, if not impossible.
    
    b) I'm lazy. I'd rather go through the effort of getting a good mix
    once and storing it in my sequences. I don't want to have to deal with
    setting up a proper mix before every performance.
    
    c) I'm not that fussy. For the kind of music I'm playing and the level
    of musical sophistication of the audiences, and "canned" mix is good
    enough. I do like having the graphic eq on the mixer, because it allows
    me to get some level of control over thigs without a lot of effort. For
    example, if I get into a room with a lot of echo I can roll off the
    bass some, or if the drums sound a little flat I can add a little punch
    to the high end.
    
    
    I think you have to find some practical compromise between the quality
    of your musical output and the cost of achieving it.
    
    - Ram

    
    Ram's comments are the sort of thing I meant.  Basically, running a
    couple of monotimbral SGUs into a mixer allows more flexibility
    than a multitimbral SGU with limited on-board mixing (usually just
    level and pan). Some scenarios I have come across:
    
    1.  Two sequencers I have used that allow CC7 adjustments deafult
    to full volume on each track.  This means that you essentially mix by
    leaving the quietest parts on full and turning the louder ones down. 
    I'm sure we have all found how certain patches are inherently louder
    than others with all level controls set equal.  This can cause a
    problem since the mix from the SGU may now be at an *overall* level
    that's too low.  What's the solution, start with all volumes on only
    half-way so you can adjust both up and down?  Otherwise, from song to
    song you will need to adjust master volume on certain SGUs.  What sort
    of approaches do people out there use?  Do you actually tweak mix
    volumes so all songs are at about the same volume?
    
    2. Some SGUs limit polyphony when you use assigned outputs.  The Korg
    O3, for instance, is 16 voices.  In "unassigned" mode, both outouts have
    the same signal, up to 16 voices.  In "assigned" mode, you can assign
    parts to one jack or the other but no jack will produce more than 8
    notes at a time.  So if you are using one jack for drums, say, which
    has low polyphony (typically no more than three voices used at once),
    you effectively lose *5* voices that are available when in the
    "unassigned" mode.  Karl once mentioned a limitation like this on the
    Emax.  So you sometimes have to trade off having individual outs for FX
    processing or stereo panning vs. maximizing polyphony.
    
    3. Units with built in FX can cause a problem of assigning FX to parts.
    Newer SGUs are getting better at this, but there are SGUs where if one
    part gets an effect, *all* parts get an effect.  Similarly, if you have
    only one or two output jacks, you can't separate parts easily for
    processing.
    
    4. No way to EQ individual parts easily. Sure, you may be able to
    adjust the filter on the patch or even use an EQ algorithm on an
    on-board FX unit, but neither is very straightforward.  How many folks
    customize patches for specific mixes (On the ESQ-1, for instance,
    panning is at the *patch* level, not the MIDI/mix level, so if you want
    to create a stereo soundstage on the output, you need to modify the
    patches for that particular  sequence)?  
    
    I'm just curious as to how people are working within these limitations.
    
    
    							Brian

    *In the studio*, I still do well with a combination of CC7 and editing
    velocity.  If there's a part screaming for specific effects and/or EQ,
    I'll bounce it to a track (or 2) on my 8-track.  Why would I edit
    velocity on a part ?  One - got a velocity switched sample that pops
    into the 'hot' sample at inopportune times.  Two - got a keyboard
    spanning drum kit.  Record kick/snare on one track, toms on another,
    and cymbals on another.  Cymbals too loud.  CC7 affects the entire MIDI
    channel.  Lower velocity on cymbal track only.
    
    If I were more bugged with the mixing issue I'd invest ~$300 in a JL
    Cooper MIDIMixerFaderMasterthingy (the 8-channel one).  Unfortunately
    my version of Performer only records from one MIDI channel at a time,
    so I can use the data slider(s) on my KX88 assigned to CC7.  Rarely use
    the volume footpedal while recording.
    
    karl

 >  Do you actually tweak mix
 >  volumes so all songs are at about the same volume?
  
    Yep.
 
>  How many folks customize patches for specific mixes?

   I do! I do!!

  One of the things that I'm so thrilled with since I've "discovered"
  sys-ex is how precise I can be with anything that can be controlled.

  Custom patches are a way of life with me. Once I get it built, I load
  it into the sequencer and it's always there...

  I tweak levels a variety of ways. I always try to ignore my board, unless
  I'm recording. Set all the sliders and leave 'em is my goal. My TZ's can
  be controlled eitehr via CC7 or sys-ex commands that change either the
  overall level or the level of any instrument in a 'performance'. The DX
  responds to CC7, so I leave it's onboard slider wide open and use CC7.
  My Mirage and HR are stoopid, so I have to really use sliders on those.

  Panning isn't a major concern of mine. My submixer and my board both have
  pan controls, though I rarely move them from where they are.

  When I was gigging, I used to pan everything on my submixer to center, send
  1 stereo out to my monitor amp and the other to the house. I'd set up my
  mix using the volume pots on the sub, and when the sound man approved I'd
  leave 'em. After awhile I got to be able to do the monitor->house conversions
  pretty well in real time. (But I'd always keep an eye on the sound man!)

  Edd

    Why program pan into the patch?  You could use a CC for that (I think
    there's one defined now for just about everything). 

    One of the things I've tried to avoid in purchasing SGUs are built-in
    FX.  Unless the unit can perform discrete/separate FX with every patch
    (ie, not global only), then it's of limited value to me.  Example: I
    like a nice delay/flange on a string part, but not on a piano. I also
    look for units *with* multiple outs (okay, so I own a 1000PX and am a
    hypocrite).

    The few times I've gone live with MIDI, I've pretty much set levels and
    pan at the mixer and left well enough alone.  I think where the
    controllers really shine is in the studio - you can automate the entire
    synth mix if you want ... unless your machines don't respond to MIDI
    CCs (some of mine don't). 

+b

    Re: .9
    
    > Why program pan into the patch?  
    
    Brad, remember that the ESQ-1 has pan as a parameter of the final DCA on
    a patch and you can modulate the location within the stereo field with
    any modulator.  This is more flexible in some ways than the more
    typical method of assigning a fixed pan location within a multi-timbral
    setup, which is then controlled by CC.  It lets you do things like set
    up stereo pianos, by modulating pan by keyboard location or autopanning
    based on velocity, etc.  Because of this, the ESQ sequencer has no
    panning capabilities built-in.
    
    
    						Brian

    As much as a fan of Ensoniq as I am, I will state that IMO the "pan"
    feature of the ESQ-1/SQ-80 is almost useless unless you are talking
    about a layer.
    
    Basically each voice has a mono output that can be panned.  The kind
    of panning effects that gives you aren't very effective in terms of
    psycho-acoustics.
    
    The main point is that you always get the same signal coming out of
    both sides, you can only vary the volume.  Now if you layer two
    patches, you can get discreet stereo type things and that can be
    very effective, but... of course it burns up polyphony.
    
    Note also, that if you're thinking about using those stereo outputs
    as two separate mono outputs - don't bother.  You can't.  Due to a
    limitation in the hardware, if you pan something permanently hard to
    one side, say left, when you initially hit the note, you will get
    annoying clicks on the right output.
    
    Ensoniq has taken some heat for this, but maintains it's a limitation
    folks will have to live with.  The party line is that usage was not
    one of the intended usages of the pan feature.
    If however, you layer two patches
    
    	db

    re: .6
    
    Regarding problem #1 (balance between sequenced tracks) there are
    a number of solutions.
    
    1) As you mentioned, get into the habit of setting each new track
       to 32.  You can facilitate this by creating empty "template"
       sequences that you just make a copy of whenever you want to
       "create" a new sequence.
    
    2) Don't program the tune from the beginning.  Start with the "busiest"
       section - get the balance you need on that section, and then use
       THAT sequence as the "template" for other sequences in the tune.
    
    3) If for some reason you have one track that is just too loud or
       too soft, you have a couple of choices:
    
    	a) edit the MIDI volume on all the tracks that use it
    	   (which is obviously a bit of a pain, but not really ALL
      	   that bad - I mean if you're at all systematic about
    	   programming your tracks (like in (2) and (3) this
    	   shouldn't happen too often anyway - I often do it this
    	   way).
    
    	b) Most synths allow you to adjust the overall volume of a
    	   patch.
    
    	   So, like on your SQ-80, if you've got a bass track that's
    	   too loud, rather than editing the volume on each sequence
    	   that uses it, you can just edit the volume of the patch
    	   and it will lower the volume on the whole song.
    
    Regarding tweaking patches for specific songs/mixes/etc/eq's.
    
    Well... I think you oughta get used to doing that.  

Here is my approach. First, you should know that
I am currently using a single multi-timbrel SGU,
which feeds stereo into a conventional analogue mixer.
Also inputing into the analogue mixer are vocals (2ch),
Guitar (1 or 2 ch). All effects are in FX loops on the mixer.

On the midi side, I attempt to place each seperate
sound (ie; Bass, Piano, Drums) on a seperate
sequencer track. I attempt to keep the sequencer
tracks the same as a multi-track tape machine, and
essentually use 1 track per sound, unless a song has a
pletheora of sounds.

I attempt to mix pan each sound (track) to provide a maximum
seperation and clarity of multiple sounds. For example, 
I don't want the bass drum and the bass guitar colliding 
(at similar frequencies) in the same speaker. So, I'll pan 
them as far apart as possible without sacraficing a 'real' feel. 

What I mean by real is that drums are almost a point source, so you don't 
expect to hear the bass drum from the far left, and the snare hard pan right!
It's just not natural, unless you are simulating a drummer will groillia arms.

Next I set up volume with CC 7 for each track. This may change
slightly within each track to allow for song 'dynamics', but
at this stage of the game I try to get the levels such that all of the
various sounds are about equal as called for by the song. That usually
expresses itself as clarity. Can I still hear the guitar amoungst the
piano and horns?. Is the bass guitar distinct from the bass drum? etc..

More subtile tweeks are done with Note Velocity. Rarely is panning
done after the track is established, but sometimes for an effect
it is interesting. Guitar slides with dynamic pan are fun!

I set pan taking into consideration the mix interaction of frequencies, 
but I also try and visualize physical placement of "virtual band members' or
"virtual band sections". For example, for a Beatles tune, I may have
drums almost centered, the bass off to the left 75%, piano to the right
at 75%, and a string section left at 90%. I attempt to create
the illusion that the audience is standing front center of the stage,
with imagined players in different positions.
If I want stereo piano, it is left playing hand at 75% and right playing hand 
at 80%. The piano is stereo, but still has a virtual position in the virtual
band.  I don't place left piano playing hand at 0% and right playing hand 
at 100% pan. That has no virtual bandmember position.

But, the major criteria for pan is how the sound frequencies are interacting
in the speakers. Frequency collisions are avoided, even if it is at the expense 
of the "placement of players illusion". I'll pan low frequencies farther than high
frequencies, since low frequencies travel farther, and I don;t have to be concerned
about loosing the sound to a person at the far opposite speaker position.

Most of my initial midi mixing is done with headphones. This is convienent, since
I don't have to keep the neighborhood awake while I am doing the initial tweeking.

But, my real goal is to have an optimum 'live' mix, played thru my PA,
as I accompany on guitar and vocals.  So, the next step in preping the mix 
is tweeking while listening thru the PA at near full volume. When I do this,
the whole mix usually needs to be changed drastically. Seems like redundancy,
but it is always worse with no preliminary tweeking, and unmixed and loud
may be unbearable.

Every PA system, no matter how good adds color to the sound. Mine is not so good, 
so it adds lots of color. The bass may be emphasized, or the highs too harsh. 
So, the whole process  of tweeking pan, CC#7 and velocity is refined for the 
live mix. Organization and consistency helps here as somebody mentioned previously.
I attempt to establish pan and volume parameters at the same position
in each of many sequences, so I can quickly locate a parameter that 
needs tweeking. They are always in the first three measures of any sequence
usually at the exact same midi clock.

For EQ, I make every attempt to select and craft sounds from the SGU
as close to the optimal EQ as possible, without resorting to the
need for external mixer EQ. That way, I can play many sequences, without adjusting
EQ for each song. I would do it this way even If I had a
full time soundman. I believe that the best soundmen set things up once,
and then avoid fooling with them unless absolutely necessary. That's my goal
as well.

I didn't mention it before, but I do my 'MIDI Mixing' with my analogue
mixing board being EQ neutral.

The final step is to place all sequences on a single disk, in
the order that they will be played in a set. Then I REPEAT the WHOLE
PROCESS ONE MORE TIME, attempting to equalize volume across all
songs within a given set. Once that is done, I'm ready to play live!

Now, if I were multi-track recording this, I would attempt to put each
midi-sequence channel on a seperate tape track. Each one would get
fine tune tweeked for PAN, EQ and volume, on the analogue board.
The only change I would make in midi might be velocity. Volume and velocity 
should be close already. Seperate tracks is especially important when 
additional tracks will be layed down afterward. New interaction always occurs
as new sounds/tracks are added. I also try and get a sync track for the 
sequences, to allow the midi tracks to be redone if necessary. Of course, 
there aren't always enough tracks for everything, so tradeoffs have to be made.

I now have a SGU with multiple outputs (6). I have been considering
whether it will be worth it to send all six to my analogue mixer for live
performance, in addition to recording.  I don't intend to ever have a sound man. 
It seems that seperating the SGU into multiple just be adding complexity, 
which I certainly want to avoid unless it adds an awful lot. I may experiment 
at some time, but I suspect that I'll try and keep things primarily under
midi control for live, mimizing the need for analog mixer channels,
and for humans to tweek knobs.

I didn't realize that I was this verbose!

Rick

    re: .6
    
    >What's the solution, start with all volumes on only
    >half-way so you can adjust both up and down? 
    
    That would probably be ideal, but I don't do it because my synths are
    fairly noisy. To keep the noise to a minimum I want to have the CC
    levels set as high as I possibly can. So I start with the loudest
    voices (like drums and bass) in the loudest songs set to CC 127, and
    adjust everything else relative to that. This places an artificial
    ceiling on things, so you have to be careful where you place this
    ceiling (by experimenting before you start writing a lot of sequences).
    
    >Otherwise, from song to
    >song you will need to adjust master volume on certain SGUs.
    
    To be avoided at all costs.
    
    >What sort
    >of approaches do people out there use?  Do you actually tweak mix
    >volumes so all songs are at about the same volume?
    
    Yes. I try to keep my setup the same from one mix-down session to
    another (I mix with headphones) so I can judge the relative levels from
    one song to another. There are some obvious difficulties in this, like
    aural fatigue, but as someone said the key to success is to keep things
    as uniform as possible. I also document the settings on every sequence,
    and make notes during rehearsals and performances when anything sounds
    wrong in the mix, either within a song or with levels between songs. I
    go back later and check these things out and tweak them. Over time I'm
    getting a large body of material to the point where everything is
    consistent, and the preset mixes can be pretty much relied on. But I
    still like to have my hands on the controls when I'm airborn.
    
	- Ram    
    

        Re -2
    
    	There would seem to be a convention, to which I subscribe, that low
    frequency content should be panned (or not) to the centre stage. As the
    low frequency stuff is not directional. In fact I was listening to a
    JBL sub-woofer system the other day (designed to go with the little
    control-1s that I use) The system was parked away in the corner of
    the room but the bass and bass drum was still coming from slap between
    the speakers. Now that's psycho-acoustic.
    
    For the gigs I about to do we have decided to stick with a non-stereo
    P.A. system, less money, less hassle. It's only a small place. Shoud I
    rethink.
    
    				Paul.

re: < Note 2328.15 by BAHTAT::KENT "peekay" >

	I think more people run in mono than stereo for
	live. Many people who have stereo set-ups, actually
	run them in mono.

	One of the problems in stereo is that you can't
	be sure that everybody in the house is hearing 
	the same thing. The guy next to one stereo speaker
	cab can't hear the cymbal highs that are comming out 
	of the speaker across the room.

	It can be done, but you have to have most of the sound centered
	to avoid that problem. 

	I would not try to talk you into stereo. I use it, but
	then again I rarely get out of my own basement. Some
	of the stereo mixes that work well there, probably wouldn't 
	work well in a lot of rooms.

Rick

>    For the gigs I about to do we have decided to stick with a non-stereo
>    P.A. system, less money, less hassle. It's only a small place. Shoud I
>    rethink.
    
    No.  Stereo really isn't very practical.
    
    Think about what you would do with the stereo?
    
    Pan instruments - means mix is screwed for anyone not sitting near
    the center.
    
    About the only practical advantage of stereo I can think of are
    subtle stereo effects that don't compromise the sound if you
    happen to be sitting closer to one speaker than the other.
    
    Stuff like mild chorusing, stereo reverb, etc.
    
    I'd love to have my electric piano sound chorused coming through
    the PA, but being that sometimes you want it, and sometimes you don't,
    it means more things to do at the PA board.  If you have a soundman,
    fine - he can do it if you take the time to go over the cues with him,
    otherwise, it's just a pain with very little gain.
    
    	db

	Hmph. I think it's great, I just wish all my SGUs would do it the
	same way. Good thing I don't have too much stuff, eh?

	My mixing is currently a hybrid split between CC7 automixing and
	"analog". The EPS and ESQ work fine as CC7 listeners, the HR16 and the
	MIDIbass are ignorant of CC7. Soooo, the HR16 gets all 4 outputs to
	separate channels (kick, snare, kit left, kit right). MidiBass takes
	one. I use the EPS left, right, and solo 1-3 generally as separate
	outputs (solves EQ and fx problems). The ESQ just gets a stereo mix,
	which is fine, since there isn't all that much you can squeeze from it,
	it runs out of voices pretty fast. I use the JLCooper Midifader to 
	record (sequence) mix info to a separate track (or tracks). That way I
	can do alternate mixes (I use Dr. T's KCS, making more tracks is pretty
	simple)... I calibrate the analog in at a specific (and arbitrary) level
	from the EPS/ESQ and don't touch the sliders. Tape in, etc., just acts
	like a normal primitive mixing fader. 

	Next drumbox I get will definitely need to hear CC7 and pan. 

	One other note. There is a good use for boxes that don't hear CC7.
	I can double the MIDIbass and the ESQ DIGBAS and fade the DIGBAS
	back without mucking up the MIDIbass. Big deal, huh?

	I think it make a lot more sense to record mix info in your sequences,
	it's one less thing to remember. 

	whatever

	/pjh

    
    The only time I use stereo is in a small room, with a narrow stage, at
    controlled volumes, where it's effects are worthwhile. Otherwise, 90%
    of the time, I run mono. Especially at high SPL's.
    
    CdH
    

    
    My current results from both live mixing and studio (for tape)
    mixing reveals:
    
    	CC7 is NOT USABLE IN REALTIME.
    
    It's ok to set up 'one' initial volume level before the performance
    starts. That's NOT enuf control over mixing.
    
    You need a MIDI controlled board with a separate midi cable (I do NOT
    mean channel. Running CC messages together with performance data will
    cause timing problems in even 'almost simple' sequences).
    
    Unless that mixer is run separately, there is no way you can capture
    mixing changes in realtime.
    
    So, ok, most of you rock with constant um, dynamics. (boy, is that
    an oxymoron: constant dynamics), and the audience expects it...so
    you're probably happy with setting initial volumes. 
    
    I dont call that mixing. Neither would a good sound engineer.
    
    And I guarentee the hit you just heard on the radio was not mixed
    with midi.
    
    ron
    

    Re: .20
    
    The Ensoniq ESQ-family machines have a CV pedal input.  This can be
    used as a volume pdeal or a mod source, depending on the setting of a
    global parameter.  The pedal info can be recorded into the sequencer. 
    A real bug (feature?) is that the definition of volume vs. mod is made at
    *playback*, not recording.....so if you want to use it for mod on one
    track and volume on another, no can do.
    
    OK, so I tried doing a fade-in on a sequence with the pedal and got OK
    results with onboard voices (where the coupling would be more or less
    direct) and also to an outboard SGU, in this particular case, a Korg
    Symphony.
    
    I found that the Korg responded to CC7 via the pedal, either in
    real-time, or from the sequencer, in a jerky manner.  I had to move the
    pedal quite a bit to get *any* signal from the Korg, and when it kicked
    in, it was at a pretty high level, so it was impossible to get a smooth
    fade.  Then, as I increased the volume, it seemed to make rather large
    discrete jumps.  
    
    This is not quite the same thing Ron mentions, of a delay in response;
    I guess all this proves is that the Korg implementation for CC7 doesn't
    allow for very many levels of volume, and that the Ensoniq has more
    levels, therefore, the volume control via pedal *seems* continuous (I
    know that it *has* to be stepped in a digital machine).  
    
    On the other hand, while using the pedal for volume changes may be
    messy with certain SGUs, the ESQ sequencer has a few redeeming features
    when it comes to mixing.  The sequencer works somewhat like a drum
    machine in that you create "sequences" which are chained into "songs".
    Each *sequence* gets its own mix levels (as well as patch assignments,
    tempo setting, MIDI channel assigns, etc.) so that you can essentially
    change mix levels of a *song* on a bar-to-bar basis if need be.  If you
    have a long sequence and need to have volume changes in the middle,
    just copy it to a blank location and use bar delete commands to split
    the sequence up into smaller sequences, each with its own mix level,
    then chain them back up as a song (this is how the factory demos were
    done..some of the sequences are in fact only a bar long).  By putting
    the mix change at the *top* of each sequence, you are guaranteed the
    CC7 message goes out prior to the note-on...any delay in response would
    be due to SGU-specific architecture problems.  
    
    I know that the MC-500 can overdub volume changes in mid-sequence....is
    this sort of feature not common in PC-based sequencers?  
    
    							Brian
    
    P.S. If God didn't intend rockers to have constant dynamics, he
    wouldn't have given us compressors.  8^)  8^)  8^)

    re: .20
    
    This sounds like it's on the verge of getting religious, but in defense
    of MIDI "rockers who have no dynamics", I'd like to repeat that I think
    preset mixing is highly functional *and* musical. From some of the
    comments I suspect that how well it works has a lot to do with the kind
    of equipment you are using. First of all, my sequencer (an MC-500) is
    quite capable of including CC7 commands in the data stream with no
    timing problems. I can vary the sound of any track at any point in the
    sequence. My synths handle a full range of dynamics from 0-127. In
    addition, I enter sequences using a velocity sensitive keyboard, so in
    addition to the volume range from 0-127 I have control over the
    velocity of individual notes from 0-127. Between the two I have been
    quite satisfied with the dynamic range I have been able to produce in
    sequences, and I think it's as good as any good sound engineer could do
    live (I used to be one).
    
    I'll grant that I can't reproduce the full dynamics of a symphony
    orchestra, but that's a fundamental limitation in the equipment I'm
    using, not MIDI.
    
    If I was the Rolling Stones and had big bucks to spend at every
    performance I'd probably like to have a sound engineer and a live mix,
    especially if I was recording. But I'm not, and I find using CC7 and
    note velocity to produce quite satisfactory results for the typical
    requirements of performing.
    
    - Ram

	re: .20, .21

	Gee, and I have an ESQ-1. Don't have any of the problems you mention,
	and 2/3 of the stuff I do isn't rock, and has plenty of dynamics. 

	My sequencer has NO TROUBLE AT ALL dealing with CC7 strings, and
	can even comb them out to reduce loading. But I've never really
	had to do that, since the JLCooper MidiFader thingy allows you to
	set the granularity of the fader, individually for each fader. I don't
	use the ESQ-1 sequencer for anything. I make "sequences" with no events
	in them so I can get multitimbrality from the machine, and preset
	multitimbral setups. But I don't have much trouble with dynamics, even
	at low levels - of course, it's not my primary SGU, so it rarely is 
	heard by itself.

	And I do funky things like merge timing info too. I haven't had a
	data overrun yet, except in cases where I did something stoopid,
	like made a feedback loop in MIDI with the patcher. Maybe it's the
	configuration, I never thought the MX-8 was more than the MIDI
	equivalent of a repeater, but it does buffer the inputs.

	whatever. I disagree that CC7 doesn't work. I maintain that with
	enough bux and the right equipment it works just fine.

	/pjh

    Re: .23
    
    Check out .21 again, the problem I had wasn't with the Ensoniq synth,
    but the *Korg* box I was driving.  Of course, any box that doesn't know
    what pitch bend is would probably also be dumb enough to have a severely
    limited granularity for volume response.
    
    8^)  8^)  8^)  8^)
    
    							Brian

    1.  I was always under the impression that variation of dynamics
    *during* a song was the responsibility of the performing artiste, not
    the sound engineer.  (Ok, here is where Tom inserts his story about
    the music professor who decided to 'enhance' one of Tom's live
    performances.)
    
    2.  Volume change and velocity change are not the same thing.  Just
    about all acoustic instruments, and many synthesized ones, change their
    tone color the harder you 'hit' them.  Velocity change gives you that.
    Volume control does not.
    
    3.  PC-based sequencers can typically overdub anything, anywhere.

    
    After much frustration, I have finally come up with a method to overdub
    CC7 changes in mid-sequence on ESQ-1 and SQ-80 synths.  See note
    1001.52.
    
    						Brian