[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference napalm::commusic_v1

Title:* * Computer Music, MIDI, and Related Topics * *
Notice:Conference has been write-locked. Use new version.
Moderator:DYPSS1::SCHAFER
Created:Thu Feb 20 1986
Last Modified:Mon Aug 29 1994
Last Successful Update:Fri Jun 06 1997
Number of topics:2852
Total number of notes:33157

920.0. "Sequencing Basics and Tips" by AKOV75::EATOND (Finally, a piano.) Fri Aug 28 1987 13:44

	I'd like to get into some serious sequencing.  Up until now, I've
used my sequencer (MSQ-100) as little more than a multi-track recorder.  I'd
like to learn how to use one more in an integrated MIDI environment.

	My first question is; how does one load in program change information?
Is it done in step time in an initial measure?  Do you have to rely on a quick
hit of a button just before you begin playing?  Is this something that differs
widely from sequencer to sequencer?

	I know I'll have more questions as time goes on.  Perhaps this note can
be a base for tips and warnings on sequencing in general.

	Dan
T.RTitleUserPersonal
Name
DateLines
920.1Cheap sequencer expert....JAWS::COTEPractice Safe SysexFri Aug 28 1987 14:4226
    Loading a program change is just like loading any other MIDI data,
    you press the button and it gets 'recorded'...
    
    A few caveats - Make sure your sequencer is equipped to pay attention
    to program changes. I can't think of any that aren't, but it may
    be a software switch "program change enable" or something.
    
    Also make sure your sunth is sending them. On my DX, there are 2
    parameters I have to toggle. "Channel Message" must be ON and
    "System Message" must be OFF. No other combos work.
    
    Remember, your telling the synth to go to patch #nn. You're NOT
    telling it "Go to Funky P-BASS II". If you put some other patch
    in the place where Funky P-Bass II used to live, you'll get the
    new patch.
    
    Try to insert patch changes during rests. Aftertouch recommends
    finding at least an 1/8 rest to put the change in. I've successfully
    used 16th rests.
    
    Be careful, if the note hasn't run out it's decay, you'll get weird
    results when you change patches....
    
    It's easy! Experiment a bit!!
    
    Edd
920.2Oh Yes, I Remember That Problem...DRUMS::FEHSKENSFri Aug 28 1987 15:1347
    The MSQ-100 (I still have mine!) will not record patch changes in
    step time programming mode.  I guess Roland figured this was a feature
    (my motto - Roland gets it 95% right all the time).
    
    So, how do you get a program change in at a crucial point?
    
    Everything Edd said is relevant, but here's the MSQ-100-specific
    stuff.
    
    This is how I used to do it:  first, step time program the notes,
    regardless of patch/program.  Then, figure out where all the program
    changes need to go.  They will have to be on bar boundaries (ah,
    the tyranny of the bar line...)!  Put the MSQ in real time overdub
    mode.  Goto the desired bar and start recording. Send the patch change
    to the MSQ while it's counting into the bar (it counts 8 beats,
    displaying -8, -7, -6, ... -1 in the display while doing so).  If
    you have any worries about getting the patch change sent in this
    time (like if you have to run across the room), just crank the tempo
    all the way down and you'll have up to 16 seconds (8 beats at 30
    bpm), which ought to be enough to run a few feet and punch a few
    buttons.  The patch change will be recorded as the very first thing
    in the bar where you started the "overdub".  This also works for
    stuff recorded in realtime mode if you don't have enough hands to
    play and change patches at the same time.
    
    Warning - get it right the first time.  Otherwise you'll have to
    delete the bar containing the offending patch change and reenter
    it (the notes as well as the patch change!).  If you have any ties
    from that bar into the next, my condolences to you.  The MSQ doesn't
    deal with this situation very well, i.e., it can't be done.  Your
    notes get truncated.  You may have to delete everything up until the
    first bar that *isn't* tied (in any way) to a preceeding bar, and
    reenter it all.  Hey, what'd'ya want from a "one of the first ones
    designed" sequencer?
    
    Also, if your recording multi(MIDI)channel stuff, make sure your
    transmitting on the right channel and the MSQ's channel offset is
    0 or whatever it should be to get the data on the right channel.
    You *can* delete only one channel's worth of data from the MSQ,
    but you don't really want to delete anything unless you absolutely
    have to.
    
    My MSQ-100 taught me how to appreciate my MC500.  Now I tend to think
    of it (the MSQ) as a 6500 note memory cartridge.
    
    len.
    
920.3Gee, I forgot about this note...AKOV76::EATONDFinally, a piano.Thu Sep 10 1987 17:2822
	I thought I replied to this already, suprized to find otherwise...

RE .1 (Edd)

	Why does the SYSTEM switch need to be set to off?  What does that do?
Send and recieve system-exclusive data?

RE .2 (Len)

	You said there was a need for bar lines.  Is this just so that you
send a program change only 'between notes'?  I often play grace notes going
into a measure, especially when playing a piano patch (you know, rolling into
a chord...).  Is the single requirement that there be no notes playing at the
time of the change or is the bar boundary *the* only place a patch change can
be recorded?  Or are both points true, the first for the sake of the sound,
the second because that's the only place the MSQ puts those changes?  What if
I sent a patch change in the middle of a bar?  Would the MSQ ignore it, or just
put it at the end of the bar?

Thanks,

	Dan
920.4song position...AKOV76::EATONDFinally, a piano.Thu Sep 10 1987 17:3310
	While you're ansering those ones, I got some more.

	Concerning the interfacing of a sequencer and a drum machine.

	There is talk of song pointers or positioners.  What exactly does that 
mean in practice.  By the sound of it, it would seem that if I pulled up measure
57 on the sequencer and hit 'CONTINUE', it would cause the drum machine to start
playing at the same place, measure 57.  Is this so?

	Dan
920.5??JAWS::COTENote stuck? Try Kawai...Thu Sep 10 1987 17:3714
    I really don't know!!!
    
    ...but after whacking about for a few hours I found that the DX21
    has 2 parameters, chanell msgs, and system msgs. If thy're not set
    up as chanell=on and system=off it just won't send patch changes. This
    is neat; I can control my Mirage disc drive from the DX. I can change
    some Mirage parameters *faster* by using the DX!!
    
    ...and I never muck about in sys-ex data so I can't say much about
    that parameter.
    
    Edd                                              
    
    
920.6Have a cigar!JAWS::COTENote stuck? Try Kawai...Thu Sep 10 1987 17:396
    Re.4 and SPP...
    
    Yep, you got it.
    
    Edd 
    
920.7FWIWECADSR::SHERMANbut I DID simulate...sort of...Fri Sep 11 1987 09:3718
    re: coupla notes back
    
    Ed probably answered this, but with SPP I can hit the Continue on
    the sequencer and the drum machine goes right to the measure I want
    *IF* the drum machine's measures correspond to the sequencer's
    measures.  UNFORTUNATELY, I tend to combine a lot of measures on
    the drum machine to increase the resolution (it's a 505) and that
    tends to screw things up.  I suppose if I really want to, I can
    chain the measures to fix the problem.  But, bein' as lazy as I
    am, I usually program up the drum machine and dump it to the sequencer.
    That way I can diddle with the measures and parameters and use the
    features of the sesquencer on the drum stuff (cut and paste,
    crescendos, macros, etc.).  I find the sequencer's editor easier
    to use for that kind of stuff.  Maybe it's just preference...
    
    
    Steve
    
920.8Point to the spot !MINDER::KENTFri Sep 11 1987 11:4611
    
                                     
    One problem I have sequencing the DDD-1 and the QX5 using song pointers
    is that the DDD seems to take longer to find the position then the
    off-set of the QX5 allows. If I put a delay on the QX5 of .9 of
    a second and then hit a measure position and then continue. When
    the sequence starts the DDD-1 is still tring to catch up. This only
    happens if we are after measure 67. Why 67 ?
                                     
                                     
    				Paul. 
920.9Did This Help?DRUMS::FEHSKENSFri Sep 11 1987 14:0812
    You can insert/overdub patch changes anyplace you want.  What I
    described was a mechanism to force them to occur just before a barline,
    when programming in step time mode (remember, the MSQ-100 will not
    listen to patch changes in step time record mode).
    
    If you want them to occur someplace else, just crank the tempo all
    the way down, go into overdub mode, and push the buttons at exactly
    the right time.  The low tempo gives you more time to be exactly
    right.  
    
    len.
    
920.10AKOV76::EATONDHow bout a rack-mount drum mach?Fri Sep 11 1987 14:179
RE < Note 920.9 by DRUMS::FEHSKENS >

>                              -< Did This Help? >-

	Certainly does.  I just wanted to get some clarification.

	Thanks,

	Dan
920.11It's good, but...AKOV68::EATONDAre PERSONAL_NAMES *that* powerful?!Mon Sep 14 1987 15:2717
	About song-pointers...

	I was sequencing a piece last night and tried out the song positioning.
Lo and behold, it did as you all said.  BUT, I noticed a quirk that could be a
bit of a pain...

	The MSQ has a feature that when you hold down the measure advance or the
one to go back, it will gradually speed up so that you can reach a measure that
is some ways away faster.  Unfortunately, the 707 did not clock with this.  It
would speed up, but not as fast as the MSQ, resulting in a lack of 
synchronization between the two the further you get from the starting point.
This could well be a *real* pain if you wanted to set the two to start somewhere
in the middle of its memory... 8^(.

	Anyone else ever encounter this kind of thing?

	Dan
920.12AKOV68::EATONDAre PERSONAL_NAMES *that* powerful?!Mon Sep 14 1987 15:3915
RE < Note 920.11 by me >

	Now, I realize some of you pragmatists are going to say 'Why don't you 
use the select feature of the 707?'  I haven't tried using the 707 as the
master over the MSQ yet.  I don't know if that presents a change of approach to
my style of song arrangement and sequencing.  As far as selecting the two 
individualy, in order to match them up, I'm thinking mostly about live work
here, so I'd rather not have the extra hassle...

	Now, once all the pragmatists are through, the extravagant among us
will say, 'get yourself an MC500'.

	Me? I'm a pedestrian.

	Dan
920.13Sorry...JAWS::COTE115db, but it&#039;s a DRY thud...Mon Sep 14 1987 15:423
    QX/RX combo has never shown this trait.
    
    Edd
920.14Back to an old note...20981::EATONDWhere is he when the music stops?Thu Jun 02 1988 11:1918
	A few more questions...

	1)  How many of you have used a sequencer for MIDI data dumps?  I
tried using my MIDI DJ for a TX81Z dump a few weeks ago and found something 
surprising:  It seems that the amount of information flooding the DJ caused
its timeing to screw up.  It clicked off at the set BPM, but when it came to 
transmitting the dump, it lost all sense of rythm.  Anyone else experience
this kind of thing?

	2)  How do you all approach sequencing?  Do you map out a song 
beforehand, measure by measure, and then enter it as prescribed?  Do you
free-wheel it as you go?  ('Course this all depends on what kind of sequencing
you're doing, i'd suppose.  I'm referring mainly to the sequencing of either
covers or 'already-written' tunes, not using the sequencer as a compositional 
tool).

	Dan

920.15An example...JAWS::COTEAre you buying this at all??Thu Jun 02 1988 11:5032
    A 'typical' sequencing session in my studio might go something like
    this... 
    
    Assumptions: A known tune, no sheet music
    
    Firstly, record a rough 'sketch' of the first verse using a piano
    or organ type voice. Doesn't have to be perfect.
    
    Play this back using a simple 1 or 2 bar drum pattern. When happy,
    work a bass line against it until the bass line is correct. Now
    fix the drums, especially the bass drum. I'm a stickler for tight
    bass and bass drum.
    
    Dump that piano track. Now I have a back-up band (bass and drums).
    Add the remaining part of the rhythm section, piano, organ, guitar
    whatever.
    
    When the first verse is complete, do a chorus the same way. Then
    do some cut and paste tricks. Do a break the same way. Cut and paste
    as needed.
    
    Eventually I end up with a whole song played by only a rhythm section
    and a drummer who only plays 1 two bar phrase. This is where I
    do the for real drum programming, with all the proper breaks, fills,
    rolls and sneglotals in the the right places.
    
    Next I go back and add any lead instruments followed by any
    orchestration.
    
    Finally I go back and tweak whatever...
              
    Edd
920.16Well, I like it this way...FROST::HARRIMANZero Tolerance = Total NonsenseThu Jun 02 1988 13:1452
    
    In my case, the music is generally original and somewhat thought
    out before I get to the point of sequencing.
    
    I'm much closer to using it as a multi-track deck.
    
    For instance:
    
    I usually put the piano (or piano-like) part in first. However,
    I try to get it clean (not too difficult, since I practice and have
    developed chops).
    
    Then I do the bass track. I don't do a drum track since I use the
    metronome click on the sequencer to keep me synchronized. I'll probably
    quantize the bass on the way in, maybe not, depending on the complexity
    of the bass.
    
    Then I'll switch the MX-8 to 'record drums', and lay down the
    bass/snare. This gets quantized, almost always. Played from the
    Octapad, always. The tricky part is remembering to stop playing
    bass or snare during fills, which aren't played yet, and are therefore
    black holes.
    
    Then do the same for HiHat and othjer cymbals
    
    Then do the same for Toms
    
    Then do the same for any other percussion parts
    
    Then go back and try any other keyboard parts (string background,
    organ, bells, whatever).
    
    By this point I should have anywhere between eight or nine tracks
    filled. I then start the editing process, which on the KCS consists
    of listening for clams, noting the measure number, stopping the
    sequencer, setting the cue for somewhere around the measure number,
    then selectively muting tracks 'til the clam is found, then editing
    the bad note or parameter, trying again, etc. 
    
    This is remarkably close to my normal recording studio protocol.
    The exception is the edit mode, which is one reason I really like
    having a sequencer instead of a lot of physical tracks. You can
    *repair* a clam instead of punching it in all the time, which I'm
    still not good at.
    
    This works for me, both with music on paper (lately 'Jesu, Joy of
    Man's Desiring' got this treatment, but without the drums), and
    with the production work I've been doing for paying customers, and
    the stuff I've been working on trying to get ready for Commusic
    N+1. Whether it works for anybody else is, of course, debatable.
    
    /pjh
920.17You got to put down the ducky!TYFYS::MOLLERVegetation: A way of lifeThu Jun 02 1988 15:0673
    I seem to find that putting together an initial drum track to be
    most useful, then add the rythm instrument (some sort of piano sound).
    I usually section off the parts into logical sections (When I put
    together 'Take Five', this was tougher than it sounded) & do each
    section seperately. Once each section has what it needs, then I'll
    try to merge the sections into verses or choruses (or what-ever).
    Once I have all the parts, I'll usually go back & add the Bass Parts.
    I try to do this, such that the flavor of the Bass style is consistant
    from one part till the next. Since I prefer the feel that an bass
    guitarist might add (and I used to play bass), I spend the time
    it takes to get that type of phrasing. I usually then re-do the
    drum track & try to make it feel better where it needs it (sometimes
    I don't change much at all, short of fills & cymbol hits).
    
    I've been using an Alsesis MMT-8 for sequencing & I also go back
    & fix bad notes thru it's editing functions, & usually adjust most
    of the velocities of various notes. Cymbols & Drum rolls usually
    key me to changes & they have to be promenent enough so that I know
    that a part is about to occur (We use this live).
    
    My keyboard capabilities are not exceptional (I'm a guitar player),
    but, by sectioning things off, I usually can play very difficult
    parts very smoothly. I also like horn sections & try not to limit
    my music to any one style. So far, I find that I like the ability
    to be able to swap from one patch to another while a sonng is playing
    (Usually done on my MT-32 or CZ-101). I also find that I'm using
    my CZ-101 more as sound reinforcement to the MT-32 (layering the
    bass instrument usually) than anything else.
    
    I also set up these standards to help keep things consistant for
    me:
    
    	Drums - MIDI channel 10		(This is an MT-32 limitation)
    	Bass  - MIDI channel 2
        Rythm - MIDI channel 3
        Solo  - MIDI channel 4
        Extra - MIDI channel 5		(for anything extra I need)
        Other - MIDI channel 8		(for use by the keyboard player
    					 if he feels like soloing &
    					 using the MT-32's patches)
    
    On the sequencer: (8 tracks)
    
    Tracks 1-4	Drums
    Track  5    Bass
    Track  6-8  Other instruments.
    
    I also set up 2 versions of each song. One is for when I play with
    a Keyboard player (Duo), the other has the Keyboard players parts
    (for a solo effort).
    
    This is accomplished by defining the lead in parts where:
    
    Track 1     4 clicks (usually the Clave)
    Track 2     Patches & Controller Settings (volume) for Duo
    Track 3     Patches & Controller Settings (volume) for Single Act
    
    Then I copy the Song to another song & modify the first step
    accordingly.
    
    Lastly, the MMT-8 allows you to title your songs. I always add the
    tempo as part of the title (keeps surprizes to a minimum), such
    as:
    
    	REELING  D135		(14 Character display)
    
    This is for 'Reeling in the Years, at a tempo of 135. The D stands
    for Duo, and you guessed it, an S135 would mean for a Single act.
    
    Once I got past the unique (and quite odd) equalization of the MT-32,
    this has worked quite well.
    
    								Jens