Conference napalm::commusic_v1

    Hmm, not a whole lot of interest in this subject?
    
    Well, I'll continue my tale anyway.
    
    Sure enough, the JX-10 transfers bulk SysEx data to and from its
    M-64C cartridge, not its internal memory.  Cute.  This means you
    have to have a working M-64C in order to load patch data from disk.
    And here I was hoping to be totally free of dependence on those
    lithium batteries.  Bad design, very bad design.  The cartridge also
    has to have been loaded with JX-10 data, so you can't just pick up any
    old scratch cartridge.  What possibly could have possessed them
    to do something so dumb?  By the way, JX-10 dumps take about 1
    minute!
    
    However, the Super Jupiter still refuses to play.  I called Roland
    again, and the guy said, "Hmm, I'll have to get back to you...".
    
    I was able to dump and load TR-707 data with no problems, and it
    appears the CZ-101 stuff works too.
    
    Too bad the Octapad and SRV-2000 don't have SysEx dump/load capability.
                                                          
    len.
    

    > ... not a whole lot of interest ...
    
    I was treating this as a review, not a general topic.  I try not
    to comment (esp. flame) on reviews, rather taking them a form of
    ``self-sufficient'' exposition.
    
    If you want an opinion on what I've seen so far, I can supply that,
    but...             
    
    Steph

Thanks for the review, Len! I recently bought an MC500 at the Wurlygig, and
was thinking of buying the MRB software. Now I think not, so you saved me
$140! I have backed up my D50 internal, ROM, and RAM on normal MRC midi
tracks by just recording. What advantages does the MRB have?

dave

    I guess it saves you a lot of scrounging around to get the SysEx
    handshaking sequences for those synths that require such for bulk
    dumps/loads.  For the synths I have, you either can't initiate a
    bulk dump/load from the synth (TR-707, TR-727, TR-909, CZ-101) or
    the dump/load requires the destination/source (i.e., the MC-500)
    to handshake even though you can initiate the transfer from the
    synth (JX-10, MKS-80).  I could probably manually construct the
    appropriate SysEx stuff and insert it into the MC-500 for the former
    case, and "overdub" (play the initiator sequence from one track
    while recording the returned data on another) but it's more trouble
    than I'm willing to go to, and it still won't work for the JX-10 and
    MKS-80 (the MIDI data sheet seems to imply that the MC-500 must
    acknowledge each data packet - I don't know how to make
    the MC-500 do that from an MRC-500 sequencer track; the one idea I
    have is a crock, would be slow, and wouldn't handle protocol
    errors).
    
    I'm curious as to how you managed to do bulk dumps/loads to/from
    a regular sequencer track with the D-50 - I can dump/load single patches
    from/to the "workspace" on the JX-10 and MKS-80, but bulk dumps
    of the entire memory are a different story.  Did they do the D-50
    right?  Are you sure you've saved *everything* and that you can
    get it back?  I don't have my MRB-500 manual with me and I don't
    remember if the D-50 requires a handshake.  I have a D-550 on the way,
    but I think the D-550 and D-50 work differently in this regard.
    
    BTW, still no answer from Roland on why the MKS-80 won't accept
    bulk loads. 
                                                             
    len.

In the D50 manual under "Data Transfer With MIDI", p61 of the advanced manual,
it says...

"There are two methods of data transfer via MIDI: Handshake and One-way"

We used the One-way method. We set up the MC500 to record on a track, waiting
for a MIDI event to trigger the start of recording. We then did the
following...

1. Set the MIDI channel of the receiver to the same number as the transmitter,
   or ALL in the case of the receiver (MC500)

2. Push the Data Transfer button on the D50, the menu displays...

	(B.DUMP) (B.LOAD) (CRD->INT) (INT->CRD)
	   ^
	   |
	press and hold down the button then press the ENTER button

		SENDING

3. Bulk dump complete and recorded on MRC500 track, press STOP on MC500.

4. name the "song"/dump data and save it to MRC500 disk as a "song"

The bulk load is just the reverse, only you press (B.LOAD) after loading
the "song"/dump data from the MRC500 disk.

Dan Arvidson and Ron Barth showed me how to do it, and we were able to
exchange patches from ROM and RAM cards as well as internal memory by
first loading them into the D50 internal RAM, then doing the bulk dump.
After we were done, we did the bulk load and everything checked out ok.
Apparently there is a bug in the D50 firmware, because according to
Ron Barth and Eddie Fritz at Wurly, the handshake B.LOAD doesn't work
right. A new D50 ROM is going to become available soon. If you are
dissatisfied with the MRB s/w Len, perhaps they will exchange it
for something else. Wurly's is pretty good about that.

dave

    Well, that's the way it oughta go on all of them; I don't know why
    Roland felt this handshake stuff was worth the trouble.  Interesting
    that they didn't get the handshake bulk load right on the D-50 -
    maybe they didn't get it right on the MKS-80 either!  Unfortunately
    there's no one way option on the MKS-80 or JX-10.
    
    EUW has already offered to take back the MRB-500 software after
    I talked with them about my "adventure".  I guess I'll keep it;
    there's no alternative (I don't expect to see an Amiga-based librarian
    for the MKS-80 and JX-10 unless I write one myself, and I'm really
    not interested in that much work; I suppose I could buy a Mac).
    And the problems I've been having aren't with the MRB-500 software,
    they're with the Roland synths!
    
    len.
    

    >they're with the Roland synths!
    
    Any problem with YAMAHA units?????
    
    ;^)
    
    Edd

    Not having any to try, I wouldn't know...
    
    But of course, Yamaha never screws up anything, right, Edd?
    
    len.
    

    Spoke to Roland again, looks like we're getting closer.  It turns
    out that the MPG-80 programmer for the Super Jupiter, through which
    the Super Jupiter's MIDI input passes (that's right, don't ask me
    why, but you plug the MIDI IN into the programmer and the programmer
    into the Super Jupiter; there's a switch on the back of the Super
    Jupiter that selects the programmer or the Super Jupiter's MIDI
    IN, and in the latter case, the programmer is inoperative!) can't
    handle bulk rate system exclusive data.  Now, I had figured this
    might be a problem, but only on the bulk load side of the situation.
    Of course, the bulk dump requires the Super Jupiter to receive (through
    the programmer) SysEx stuff for the handshake.  So it's probably
    the case that the "data" on disk is "content-free", i.e., the file
    is empty.  Why nobody (Super Jupiter or MRB-500) complained about
    this while the dump was "in progress" is another (separate) mystery.
    Nor is it clear why transmitting an empty file should result in
    the MRB-500 going into the IDLING state.  However, I am familiar
    with utterly bizarre manifestations of software screwups, so I'll
    try this fix tonight anyway (i.e., pull the programmer out of the
    loop for transfers in *both* directions).  The guy at Roland actually
    suggested running MIDI cables to *both* the MKS-80's and MPG-80's
    MIDI INs (from a 1 to 2 THRU box) so as to not have to plug and
    unplug cables; you just have to throw the little switch on the back
    of the Super Jupiter back and forth to program or bulk load/dump.
    Good thing the Super Jupiter rack isn't "up against the wall".
      
    This is really extraordinarily ad hoc design, if you ask me (which
    Roland of course didn't).  If the damn thing didn't sound so gorgeous
    I'd be tempted to blow it up.
                                                                  
    len.
    

    Well, that's not it either.  I'm beginning to suspect a bug in the
    MRB-500's implementation of the MKS-80 handshake protocol.  I pulled
    the programmer and MIDI switch out of the loop and it still hangs
    up in the "idling" state.
    
    So, another call to Roland later today when all those laid back
    California types roll into work.
    
    Also, just for giggles, I added up the number of bytes necessary
    for a bulk dump/load for the JX-10 vs. the Super Jupiter.  One of
    the guys at Roland said the JX takes so much more time (a minute
    vs. a few seconds for the Jupiter) because it stores so much more
    information.  Well, BS.  As it turns out, the Jupiter has to exchange
    more bytes (about 7700 vs 7500 for the JX!).
    
    Sure do wish I could find a customer support type who knew what
    he/she/it was talking about.
    
    len.
    

    In all seriousness, Yamaha units are about the easiest things in
    the world for doing bulk dumps.  The protocols are simple, and the
    units can cope with MIDI bulk data at full speed.
    
    Steph

Len -

There is a Roland clinic at Wurly's in Worcester next Tuesday night May 3,
at 7:30. If you come, you can ask the Roland rep, and perhaps even try it
right in the store. If you can't make it, send me some mail on
SUBSYS::ORIN explaining the problem in detail (a diagram would help)
and I will try to get an answer for you then.

dave

ps. For those who missed the Wurlygig, this is your chance to check out
    the Roland gear and ask the Roland rep some questions. Hope to see
    you there.

    I've been talking directly to the guys in California for a few days
    now, and I hope to have this problem resolved by next Tuesday.
    
    In 25 words or less, the problem is that the MKS-80/MRB-500 combination
    *appears* to work ok when dumping data from the MKS-80 to the MRB-500
    bulk librarian, but hangs up when trying to load the data from the
    MRB-500 to the MKS-80.
        
    Given the inability of the Wurlitzer staff and Roland's in house
    customer support staff to resolve this issue, I kind of doubt that
    a field rep (probably a sales type) will be able to do more than
    say "read the manual, follow the instructions, I'll get back to
    you".
    
    Regarding Yamaha vs. Roland, I didn't buy my synths based on how
    easy it was to do bulk loads and dumps, I bought them based on how
    they sounded and how easy they were to program.  Whatever problems
    I've had are certainly not a blanket condemnation of Roland's products,
    as I've had no trouble doing dumps from all the other Roland products
    I own.
    
    Regarding the MKS-80 programmer and its inability to pass high data
    rate system exclusive data, please note that Yamaha does not
    provide any equivalent product for their DX line of synths - the
    Super Jupiter has the same "pick a parameter number, set its value
    via up/down buttons" that the DXs offer, built in.  The optional
    programmer is an "old fashioned" knobs and sliders style interface
    that makes programming a cinch.  Roland offers similar optional
    programmers for the JX-10 and the D-50/550, and I have them all.
    If I thought there was something basically wrong with Roland gear,
    I wouldn't buy so much of it.  Yes, they make an occasional blunder,
    but don't we all?  If the DX-7 was so flawless, why is there a new
    one?
    
    And if I thought Yamaha gear had anything to offer me, I'd own some. 
                  
    len.

    Woah, geez.
    
    I didn't mean to imply any of the things that you inferred.
    
    I own as many pieces of (handshaking) Roland gear as I do Yamaha,
    and like both very well.
    
    I was merely stating a highlight of my experience in writing a
    bulkdumper/loader program, for those who might be interested.
    
    Steph

    Sorry Steph, maybe I'm a little too sensitive about this.  I've
    gotten the impression from some folks' reactions to this problem
    that my "honeymoon" with Roland is over.  This is frustrating
    because it's the first and only problem I've ever had with Roland
    equipment.  
    
    So, sorry for beating you up about a perfectly valid comment.
    
    len.
    

Re Roland Clinic -

The clinic will be presented by John Campbell, LA based studio musician
and performer. It's more of a demo than a sales pitch. He is sponsored
by Roland, but not a Roland employee. Good luck Len. I'll ask him anyway
just as a test ;^)

tgif

dave

    The plot thickens.
    
    Roland has basically stopped being helpful.  I have spent a great
    deal of time on the phone to California, mostly being told there's
    nobody around who's familiar with this equipment, or when somebody
    is, that they haven't had a chance to look into it, and finally that
    they *couldn't* look into it because they didn't have a Super Jupiter
    in the lab!
    
    So, I scared up two more Super Jupiters locally, and tried the MRB-500
    software on them.  I haven't gotten a report back yet on one of
    them (owned by one of the guys at EUW), but he mentioned that he
    had seen or heard of similar problems with SysEx bulk transfers
    from Super Jupiters with other software (this is only anecdotal,
    but disturbing nonetheless).  The other one (courtesy of rockin'
    Ron Ross) behaved exactly the way mine does.
    
    So it appears that either:
    
    	All Super Jupiters are broken the same way, and can't do SysEx
    	bulk loads, or
    
    	the MRB-500 software doesn't work with Super Jupiters.
        
    
    Either way, I'm, uh, screwed, but a lot more so in the first case.
    
    Next step is an irate letter to Roland.
    
    Stay tuned.
    
    len.
    

    Well, as it turns out the MRB-500 *does* work with the Super Jupiter,
    but only if you ignore the documentation and whatever the Roland
    product specialists tell you.
    
    When you dump data from the Jupiter to the MRB-500, you hit RECORD
    at the MC-500, then you hit SAVE at the Super Jupiter.  That all
    works fine.  For a load, the MRB-500 manual says to do what the
    synth manual says, then hit PLAY at the MC-500.  The Super Jupiter 
    manual tells you to hit LOAD to initiate a load transfer. WRONG.
    When transferring data from the MRB-500 to the Jupiter, you just
    hit PLAY at the MC-500.  Unlike the Jupiter to MRB-500 transfer,
    no user action is necessary at the Jupiter end for an MRB-500 to
    Jupiter transfer!  In fact you *must not* hit LOAD at the Jupiter;
    it puts it into the wrong state for what the MRB-500 is about to
    do to/with it.
    
    I figured this out after screwing around looking at some MKS-80
    SysEx data (it's fun reading hex dumps), prompted by Ron Ross's
    experiments with his Super Jupiter.  No thanks to Roland, who haven't
    replied to my letter yet.  I'll send them another one now, saying,
    "Thanks for all the help and the excellent documentation.".  
                  
    len.