Conference napalm::commusic_v1

    Yep, I have NO MIDI-thru. The JX I'm using is a very early production
    model (ser. no. XX000055, I think). While the upgrade solved my
    OMNI-OFF problem, it's no utopia. I now have velocity sensitive
    EVERYTHING. Organ has to be the wierdest... 
    
    I've solved (got around) the problem by always using my DX as a
    controller. When the upgrade is activated, the JX is essentially
    a rack-mount box, as the keyboard is disconnected. The Mirage
    with it's velo-send just sucks up to much sequencer memory.
    
    I think you may run into problems playing scles off of the drum
    machine. Most of the machines I've seen couple one drum voice to
    one, and only one, key on the keyboard. Now if you're lucky, you'll
    have all the right notes. Murphy's MIDI law will naturally take
    precedence at this point...
    
    Edd
    
    P.S. My JX owners manual devotes 1 PARAGRAPH to MIDI....

>    Can you assign the keys to the drum voices in a way so that say
>    the scale of A minor is assigned to the keys in some order and then
>    program the 707 toplay the scale? can I somehow either defeat the
>    note off (I'm assuming that's why the notes were so short) or lengthen
>    the note on? Or something? maybe adjust the decay on the JX?

You can assign different keys for INPUT only.  For output, you are
forced to accept one of the two pre-programmed key configurations.
The second one, I beleive, is the one that covers 15 consecutive notes.

You can't defeat the note-offs.  They come automatically, and each
note will last for a sixteenth note.  You can't lengthen it either.

Try having a long release time set on your patch.  All notes will
still end up the same length, but you determine how hong that is
by how long it takes for the patch to decay.

>    Does the accent on the 707 give different values of velocity? ie
>    by programming in an accent does the midi velocity info change?

Yes.  I believe the numbers are something like no accent = 64, weak
accent = 80, and heavy accent = 96.
    
>    I also could not get the 707 to select the second pre-programmed
>    key configuration (pg. 46 7070 manual figure d-2) is there a mistake
>    in the book that anyone is aware of?

I've done it before.  There didn't seem to be any trick to it.  Just
make sure the machine isn't running, and that your in the correct mode
(Track play?) to switch configurations.

Hope this helps,
-Dave

    Hmm Edd, how do you get the JX into the "velocity sensitive" mode?
    I go the 1 paragraph also, probably the same manual.
                                              
    As far as key cofiguration 2 goes, I tried it in Track play, press
    and hold shift, press midi channel press last step. No change. Is
    this right? That's what the book says. Oh god maybe I need repair......
    
    I'll try adjusting the decay tonight if I get a chance.
    
    I'd like to program the keys and try sequencing. But it probably
    won't work. Oh well, poor man's sequencer. I know I'll change keys
    on the JX and who cares....:-)
    
    dave

    re .2 - the note length (sometimes called "gate time", corresponding
    to how long a key is held down) corresponds to the step length,
    so it will be 16th notes only for the scale (707 term) corresponding
    to 16th note steps.  If you set the scale to eighth note triplets,
    your notes will be an eighth note triplet's worth long.  If you
    set the scale to 32nd notes, your notes will be 32nds.  Etc..
    Given the difficulty of programming melodies (even in the "assign
    drum voices to successive note numbers so you can play a scale"
    mode) and the lack of ties, this feature (using the 707 as a generic
    MIDI sequencer) is of limited use to me.  Especially with a real
    sequencer sitting right next to it.
    
    len.
    

    You activate the upgrade (velo-recieve) by powering the unit up
    while holding voice selector 7 down. All JX on-board sequencer lights
    come on indicating it's 'in the mode'....
    
    Edd

    Thanks to all...I'll be experimenting some more to see what the
    limits are......don't have the money for a real sequencer right
    now.
    
    dave

    Well I fixed most of my problems with the 707, selecting the different
    keyboard patterns.
    
    Set out to do a simple experiment using sync-to-tape. I recorded
    the sync pattern along with a drum pattern. Then placing the 707
    in the MIDI output mode and selceting the proper sync mode, I was
    trying to record a drony bass line in sync to the drums. It started
    fine but after a few bars the sync glitched a little and the 707
    jumped AHEAD for the tape, by the end of the pattern it was approx
    three to four beats ahead. I tried a number of different record
    and playback levels on the sync, the book says record it anywhere
    from -10 to -3 db not very exact. Nothing seemed to change it and
    the jump was not consistantly in the same place, as measured by
    my fine eye and the tape counter on the deck.
    
    So any ideas what's going wrong?

    dave the syncless

    Hmmm - I've synced my 707 to tape numerous times and never had any
    trouble.  The only thing I can suggest is to ignore the admonition
    to keep levels at the -10 to -3 VU level and get as much signal
    on tape as you can.  It's hard to imagine how the 707 could get
    ahead of the tape, usually tape sync fails by the tracking instrument's
    dropping beats.  Are you sure the 707 was listening to the tape
    (I think "s" shows as the tempo)?
    
    Also, you may find it helpful to first erase the existing sync track
    (by recording silence over it) before you rerecord over it.  This
    will guarantee complete eradication of the old sync pulses.  Go
    for a 0 VU to +3 VU indication at record time.  Note also that the
    707 will not sync at high tempos (it can generate tempos up to about
    260 bpm, but will only sync to tempos somewhat less than that).
    
    Also when using tape sync, you should record the sync track BY ITSELF
    with no audio (also you usually want to put the sync on an edge
    track), and then record the audio by syncing it, even if the audio
    is coming from the original source of the sync.  This way any sync
    tracking delays will be shared by all audio tracks, not just those
    layed down after the sync track recording pass.
    
    I have even run the sync from tape through a digital delay and had
    excellent tracking.  I think your levels were too low.

    len.
    

    Len, 
     Thanks I will try out your suggestions. I questioned the levels
    when I read them, but assumed that signal distortion might cause
    mistracking. I have been using channel one for the sync track but
    have been recording the audio at the same time, so I'll try this
    all out. 
     Definately i was running in the sync mode the display had a s and
    some sort of funny looking E (missing the top horizontal bar).
    
     Probably I just needed to spend more time with it, but I used up
    three hours as it was and the wallpaper still needs to be scraped
    (a little at a time). 
    
     What a weekend! my stereo cassette went down again, same problem,
    my 3340 is down, won't record on the "right" channels. Gasp sigh.
    back to the shop and another couple of weeks.....my syncing experiment
    was an attempt to raise my spirits after everthing else failed.....
    Oh well, maybe I should have stuck to the wallpaper :-)
    
    dave

    Well, one ought to practice what one preaches.
    
    Lats night, Dave Dreher I got together to add a sync track to an
    already-been-recorded 8 track tape.  (The tape's by Karl Moeller;
    he asked me to replace the drums.  My original plan had been to
    just play the whole new drum part onto two free tracks, using the
    existing drum track as a click.  Dave suggested that his SBX-10
    could add a sync track so I could program the drums - a definite
    feature, as the piece is about 16 minutes long and punching in to
    fix glitches and getting all the crescendos and fills in the right
    places would be tricky as the waxing and waning on the mixdown
    was down in the mix and isn't on the master, so a lot of cues would
    be missing.  Got that?)
    
    Anyway, Dave further proposed that we first lay down a click track
    using the SBX-10 "tap" input - rather than write an FSK sync directly
    from the tap input, we'd write the click which would be easier to
    punch in to if we had to to fix glitches in my 16 minutes of continuous
    quarter note tapping.  Made sense - you can't punch in an FSK track.
    Since the SBX-10 can sync to a quarter note click track, we could
    restripe the tape with FSK after we had a good click down.
    
    So, we figured it would be good to try this first so we didn't waste
    16 minutes of tapping if something was screwed up.
    
    So we set everything up and I tapped for a few minutes.  We rewound
    the tape, set up teh SBX-10 to listen to the click on tape, set
    up my 707 to listen to the SBX-10's MIDI out, programmed up a
    compatible pattern (a complaint about the 707 - you can only change
    its clock source (sync mode) when in track play mode, but to get
    a pattern to loop you have to be in pattern play mode; switching
    back and forth between these modes all the time tries one's patience)
    and started the tape.  Well, the SBX-10 locked to the click, the
    707 started running, but what we heard was bizarre.  The 707 was
    clearly tracking the tape, but in some totally off the wall way.

    It sounded in sync, but out of sync, if you know what I mean.
    So, simplify the pattern to just quarter notes (more mode switching,
    enabling this, disabling that - it must have taken about 20 button
    pushes to get everything ready to go again) and replay the tape.
    Bizarre!  Count it out - sure enough, it was quarter note triplets!
    The 707 was playing 6 beats to the SBX-10's 4!  What the hey is
    going on here?  Confirm it by changing the 707's "scale" (step size)
    to 32nd note triplets, but with a pattern length of 16.  Sure enough,
    it locked up fine, except it seemed to drift just a little, like
    the two rhythms were "beating" against one another.  Stranger and
    stranger!  But they're both happy!  No obvious problems.  No dropped
    beats, no lurching, all the lights flashing like they should...
    
    Well, it was getting late, and I was getting ready to give up.
    Dave says, "let's try it one more time."  Len says, "OK, but I think
    we need more click level on tape".  So crank the levels up to just
    short of lighting up the OVERLOAD indicator, push a jillion buttons
    (in the right order, of course), and try again.
    
    BINGO, it's a hit.  Perfect sync all around.
    
    On to phase two.  Can we get a good FSK from the 707 onto the tape,
    driving the 707 off the SBX-10 MIDI out?  Oh no, the yoyos at Roland
    designed the 707 so it won't emit FSK when it's a MIDI slave!  Dumb!
    Now what?
    
    Dave conjectures that the MC-500 FSK (no FSK in the SBX-10) is
    compatible with the 707's FSK.  I mean, after all, they're both
    Roland, right?  Right, but that don't make no never mind to them
    Roland guys, the MC-500's FSK is obviously improved.  I was skeptical,
    but we tried it anyway, and sure enough, they're not compatible.
    
    Alright, now what?  Well, I don't have an SBX-10 and Dave was reluctant
    to loan it to me for any length of time (understandable, as he uses
    this little gem extensively, and I'm reluctant to loan even
    nonessential equipment), but my MSQ-700 will write FSK while listening
    to MIDI, so all I had to do was drive the MSQ-700 off the SBX-10's
    MIDI while the SBX-10 was listening to the click track so the MSQ-700
    could write an FSK track.  Got that?
    
    So Dave agreed to loan me the SBX-10 overnight later this week.
    All we had to do now was get 16 minutes of quarter note click down
    on the tape.
    
    So, reconfigure and off we go, tap tap tap for 16 minutes.  The
    SBX-10 has an interesting feature, whereby when you supply the clock
    by tapping, it will compute the instantaneous tempo from the time
    between quarter notes and display it to you, updating the display
    every quarter note.  Quite an interesting thing to watch.  The tempo
    for this piece appeared to be around 186 bpm, and my tapping was
    usually within the range 182 to 190.  Once in a while you'd see
    a 200 or a 170, but I couldn't hear or feel anything to correlate
    with these excursions, and they were only for one beat.  Good thing
    there was something to watch, otherwise I'd have been bored crazy
    tapping for 16 minutes.
    
    OK, stay tuned for the next exciting installment, wherein we write
    an MSQ-700 FSK track, run that track's output through a digital
    delay, start the drum part one bar early with a short bar and use
    the delay to time align the drum audio with the rest of the tape's
    audio to arbitrary precision.
    
    But first this...
    
    len.
    
    
  

    Hmm, interesting....
    
    Now check this out....
    
    I tried a number of things to get my system up and syncing. First
    I totally erased the tape. Then I recorded the sync track at max
    input level. Then I synced the drum track to it, no problems. Then
    I put the 707 in MIDI out to drive my JX synced it up and before
    the pilot tone was gone the 707 was 5-7 steps into the track. Hmm.
    that's not right. So I started over. Same thing. Hmm turn off the
    dbx (oh no no noise reduction). The 707 syncs fine at the beginning
    of the tape but finishes one beat ahead of the tape. Hmm. try it
    again with a different level on the sync track. The 707 finishes
    one beat ahead of the tape, always in the midi out mode. With the
    dbx in, it finishes 3-7 beats ahead of the tape, always in the midi
    out mode. Conclusion. MIDI out does not sync to tape as well as
    roland thinks it does. Scratch one project.
    
    dave the syncless

    I don't know what's going on up there in Maine, but I've been doing
    exactly what you're trying to do with no trouble.
    
    First, DO NOT USE NOISE REDUCTION ON SYNC TRACKS.  Do not record
    with it, do not play back with it.  It will work, most of the time,
    but to be safe, don't do it.  Same for data saves/restores.
    
    Second - what is MIDI out mode?  The 707 always emits MIDI.  Your
    707 should be in tape sync mode, i.e., the tempo display shows "S"
    (for sync) and "t" for tape.  When syncing off tape the 707 will
    emit both MIDI and DIN clocks.  If you put the 707 in MIDI sync
    mode ("S" for sync "n" for MIDI (yes, "n", it's only a 7 segment
    display)), the 707 will run off of whatever MIDI input it's getting,
    and I don't know where that would be coming from.  The one beat
    difference suggests to me that the 707 was free running at almost
    the same tempo that the FSK on tape corresponds to.  Just tape speed
    variability could account for that - I'll bet you didn't move the
    tempo knob.  When I'm having sync troubles I always crank the tempo
    knob around to some outrageous value so it'll be obvious where the
    tempo is coming from.  In tape sync mode the tempo knob is ignored.
    
    Note the asymmetry - the 707 *will* emit MIDI and DIN sync when
    syncing to FSK; it will *not* emit FSK when syncing to DIN or MIDI.
    
    Once more it goes like this:
    
    	707 is master clock - tempo display shows tempo; put in track
    play mode.  Route 707 tape out to tape edge track.  Set 0 VU level
    from pilot tone.  Start tape in record mode, *no NR*.  Get a few seconds
    of pilot on tape, then start 707.  When 707 stops, stop tape.
    
    	Rewind.
    
    	707 still in track play mode.  Set sync mode to tape ("S t").
    Route edge track output to 707 tape in.  May want to do this through
    something with a volume control. Start 707 - lights on 16 key line
    will go out, and it will not run.  Start tape; when tape gets past
    pilot tone, 707 will start, and will emit MIDI clocks over MIDI out.
    707 will stop when FSK stops.  MIDI will stop when 707 stops.

    	Repeat as necessary for subsequent tracks.
    
    
    I think your conclusion is premature.  This works, I have done it
    reliably several times.  When it didn't work, I wasn't doing it
    right.

    len.
    

    I'll try it again, but I was in Sync t, my 707 does not appear to
    output MIDI when the audio is enabled, ie I can have MIDI out or
    I can have audio not both. But I'll double check that tonight. Since
    my 234 does not have individually selectable noise reduction for
    each channel I'm thinking that sync to tape is not much use to me,
    except for this project.
    
    MIDI out mode:
    
    press and hold shift key, press midi channel release press
    instrument/glide and release, release shift. MIDI data now outputs
    but no audio is output for the track slected in track play mode.
             
    Probably I'm doing something wrong, but I do know if I don't select
    the MIDI out the synth doesn't run.
    
    Try try again.....
    
    dave all confusynced

    MY 707 unquestionably outputs MIDI and audio at the same time.
    
    Oh wait a second - do you want note ons or just clocks?  If the
    latter, you're golden; if the former, right, you can't do that.
    Are you using the 707 as a sequencer?  I've been assuming you just
    wanted clocks to sync up another sequencer with it.  I'm not even
    sure the 707 *can* sync to tape and output note ons at the same time;
    I wouldn't be surprised if Roland never expected it to be used this
    way.
    
    Rereading your reply, yeah, you are trying to use it as a sequencer.
    See if you can find one of those $150 MSQ-100s.
    
    Sorry I misunderstood what you were trying to do.  You're not confused,
    you're just trying to do something it wasn't really meant to do.
    It *ought* to still sync up, but it also *ought* to output FSK when
    it's a MIDI slave and it don't do that either!  

    len.
    

    Accodring to the book, unless I misunderstand it, it is supposed
    to be able to sync to tape or din when in the MIDI trigger out mode.
    But after another severla hours of trying everything I can think
    of and allo f your suggestions again, it doesn't. It's funny though,
    you can hear it speed up now and then, like a micro with a clock
    gone haywire.....but only when trying to output the MIDI note on/note
    offs, it sync fine when all I want is the audio. Oh well another
    approach might work.
    
    dave

    Sounds like the micro may not have enough horsepower to handle 
    sync_to_tape input, MIDI message creation, and reading of sequencer 
    data all at the same time.
    
    Todd.
     

    The Roland SBX-10 exhibits a similar quirk - unless the click in
    is absolutely Grade-A, it *looks* like it's in sync, but the MIDI
    out is off "just a little" bit.  After the session striping Karl's
    tape in Dave Dreher's studio, I borrowed the SBX-10 for a night
    and tried to map the click to FSK.  No dice, my MSQ-700 would drop
    one beat after about 30 bars (not discretely, it would drift off
    until the accumulated drift amounted to a whole beat), no matter
    how high I goosed the click in. (I also considered the possibility
    that the levels were too *high*, and tried some things based on
    that assumption - no dice.)  I finally gave up and restriped the
    tape directly by tapping (*again*) into the SBX-10, using its MIDI out
    to sync the MSQ-100, and taking FSK out of the MSQ to tape.  This
    works, but given the accumulated delays I'm going to have to resort
    to a digital delay and a short initial bar to allow me to align
    the drums to what's already on tape.  Let's see, if I drop 1/16th
    off the initial bar, at 186 bpm, that's 744 16ths per minute or
    12.4 16ths per second or 80.7 milliseconds per 16th, so the low end
    of the 64 to 256 millisecond range should suffice, or if the
    accumulated delay is more than 16 milliseconds I can drop down
    to the 16 to 64 millisecond range.  Maybe I should drop an 1/8
    just to be safe, which would put me right in the middle of the
    64 to 256 millisecond range.
    
    Oops, sorry, just talking to myself...
    
    I think Todd may be right, we're just overstressing those cute little
    microprocessors.
    
    len.

    NO! No, no, no....
    
    MIDI is God. It can do anything.
    
    I find you blasphemous.
    
    Keep trying, Len. It's in there. You must have been tired or something.
    
    Edd

    I've been avoiding saying this, because it will sound too haughty,
    but I can't resist any more.  What you guys need is SMPTE.  It solves
    all your problems, easily and simply.  Get a Roland SBX-80.
    I'm willing to loan mine, if sufficiently persuaded.
        John Sauter

    What I did to solve my problem was this:
    
    1. Recorded the sync track at the same time the 707 was in the MIDI
    trigger out mode and recorded the synth bass line.
    
    2. Placed the 234 in sync input, recycled the 707, synced to tape
    recorded the drum part.
    
    All done with that. A simple solution, I'm surprised I didn't think
    of it earlier.
             
    dave
                   

    re .19
    
    Don't mean to sound haughty either, but how will SMPTE solve what
    basically seem to be level and signal capture problems?  Yes, some
    of these difficulties could be solved by *any* sync box (i.e., one
    that will mix and match all possible sync sources and sinks), but
    nothing we've tried to do required SMPTE encoding.  Note specifically
    that I had problems using an SBX-10, and SMPTE instead of FSK would
    not have helped.

    len.
    

    If you get a minor glitch in your sync audio signal, FSK will add or
    drop a time pulse.  After a while these add up to a noticable
    inaccuracy.  With SMPTE, the same signal problems will be detected and
    corrected by the sync box, because the absolute time is encoded in the
    signal.  As a result glitches cause small local perturbations but they
    don't add over a long piece. 
        John Sauter

    Gee thanks, John, just what I needed - a real good reason to get
    an SBX-80!
    
    len.