Conference napalm::commusic_v1

    Welcome to the club, Len!
    
    	I've been enjoying my D50 for about a year now;  it sounds like
    the D-550 is exaclty alike, minus the keyboard.
    
    	If the factory patches are the same as the ones for the D50,
    I know what you mean -- they sound great - fantastic - but most of
    them are little use to me, especially being in a live performance
    situation.
    
    	Well, enjoy it!
    
    		Ron

    Tell us about the noise floor, len.  A whole world of reviewers has
    slammed the machine for being exceptionally noisy.  True or false, and
    noisy compared to what (eg, Jupiter)? 

-b

	Only bitimbral?  Gee,  that makes the D110 sound like the
	up market machine and the D550 the cheap oh spec.

	Why is the D550 so limited when the D110 is not?

		Curious,
				Barry

    Spent some more time with the D-550 this weekend, I'll try to answer
    these questions.
    
    Noise floor:  the synth itself is absolutely quiet.  Between notes
    there is silence.  The samples (or maybe the onboard reverb), however,
    are a bit noisy.  You hear it most obviously in the onboard reverb's
    tail, and even then it takes headphones and usually it's very well
    masked by the sample itself.  Noise in the D-550 is, in my opinion,
    *not* a problem, not even an issue.  The "breathy" factory patches
    totally mask the noise, but I edited some patches to get rid of the
    breathiness and the noise was still not audible, except through
    headphones while listening hard for it.
    
    Bitimbral D-50/550 vs. multitimbral D-10/110.  You're gonna have
    to ask Roland why.  I haven't looked closely enough at the D-10/110
    yet to tell all the differences, but I believe the architecture has
    been somewhat "simplified".   It is not just a multitimbral D-50/550.
    It is less than that.  I don't think D-50/550 patches are directly
    compatible with the D-10/110, and have to be "ported" like 6 op
    FM patches to a 4 op FM synth.  I suppose Roland figured that
    multitimbrality was more urgently needed by less affluent synthesists.
    
    Be that as it may, the D-550 (and presumably D-50) is just barely
    bitimbral; the implementation is nowhere near as flexible as the
    JX-10 or Super Jupiter implementations.  In addition to the reverb
    cross coupling already mentioned (which is easily defeated by resorting
    to an outboard reverb, although unlike the JX-10 and Super Jupiter,
    which have separate outputs for the upper and lower sections (stereo
    outputs in the case of the JX!), the D-550 has only one stereo output
    that doubles as a pair of mono outputs for the upper and lower sections),
    you cannot independently set the key assign mode for the upper and
    lower sections.  E.g., you can set one or the other to "solo" mode,
    but not both.  This somewhat restricts one's ability to fully exploit
    the D-550 as two distinct synths, something I routinely do with
    the JX-10 and Super Jupiter.
    
    Rather more annoying is the inability to disable envelope retriggering
    in solo mode.  All solo mode gets you is a monophonic last note priority
    synth.  I.e., when playing legato style (with "overlapping" notes)
    in solo mode, you *always* get the attack phase of the envelope. 
    Given the very "attacky" sound of this synth, especially in the
    factory patches, this is a major loss.  Both the JX-10 and the Super
    Jupiter support this feature, with the Super Jupiter giving you
    a choice.  On the JX-10, you don't get a choice;  overlapping notes
    in solo mode *never* retrigger the envelopes.  This feature makes
    it possible to do far more expressive phrasing than otherwise, and
    its absence is a serious limitation for me.
    
    As I said before, a truly gorgeous woman with a permanent zit on
    the tip of her nose.                       
    
    len.
    

    My current understanding of the differences among Roland's D family
    of LA synths is as follows:
    
    			bar	pat				rack/
    		arch	seq	seq	disk	prog	outs	key
    
    MT-32	MT-32	no	no	no	no	2	rack
    D-10	MT-32	no	yes	no	yes	8	key
    D-110	MT-32	no	no	no	yes	8	rack
    D-20	MT-32	yes	yes	yes	yes	?	key
    D-50	D-50	no	no	no	yes	2	key
    D-550	D-50	no	no	no	yes	2	rack
    
    arch:	D-50 architecture has EQ, aftertouch, and 2 additional LFOs
    		(MT-32 cannot LFO modulate TVF, TVA).  D-50/550 chorus
    		and delay said by Roland to be of "higher quality" than
    		in MT-32 synths.  MT-32 synths are 8-timbral.  D-50 synths
    		are 2-timbral.
    
    bar seq:	D-20 has bar oriented sequencer.
    
    pat seq:	D-20, D-10 have pattern sequencer for drum parts.

    disk:	D-20 has onboard 3.5" disk drive.
    
    prog:	front panel form based programming interface with
    		  up/down/left/right buttons.
    
    outs:	number of outputs.
    
    rack/key:	rack mounted or 5 ocatve keyboard.

    

    len.
    

    Re: .5
    
    No individual outs on the D10, Len.

    Hmm, I assumed because the D-110 had 8 outs, the D-10 did too.
    I love Roland's consistency.
    
    Anybody else got any other info about how these synths differ from
    one another?
    
    len.
    

    Has anyone compared the sound quality of the D-110 to the D-50/D-550?
    
    Given the complaints around noise on the MT-32 (and it seems the
    lineage of the D-110 is closer to the MT-32 than the D-50/D-550),
    maybe something else that differentiates the D-550 from the D-110
    is better sound quality.  Perhaps Roland put more $$$ into the analog
    output circuitry on the D-50/D-550 to make them quieter.
    
    		Derek

    Assuming Roland takes a shot at a Super D synth, here's what I think
    it ought to have:
    
    	D-50/550 architecture with the following additions/fixes:
    
    	   envelope retrigger control per partial, controlling all three
    	   envelopes.  (1 additional parameter per partial; 4 all together)
    
    	   TVF applicable to PCM partials       
    
    	   initialize function for patch factors and tone common parameters
                                         
    	Full 8 voice multitimbrality:
    
    	   8 outputs
    
    	   for each output:
    
    		receive channel
    		voice module assignment priority
    		active note limit (maximum number of voice modules assignable)
                relative volume
    	
    len.
    

>    Anybody else got any other info about how these synths differ from
>    one another?
    
The MT-32 has about 32 percussion sounds available on MIDI ch 10.
The D-10, D-110, and D-20 have about 63 percussion sounds available on this ch.

/Mitch

    
    re: .8
    
      The D-10 /110 sound *much* better than the MT-32. I listened to
    the D10 through a pair of large TOA studio monitors at a considerable
    (>90dB) level. Lots-o-presence, no appreciable noise. Some samples
    have that old digital grundge on the high side. side-to-side with
    the MT-32 shows no comparison between sound quality. I suspect the
    D-10 is similar in sound quality to the D-50.
    
      FWIW
    
      /pjh

    See note 1375 for more D-10/110/20 stuff.
    
    len.
                 

Worcester Wurly's now has all of the D-series in stock. Is this a good time
for DECMS to have Wurlygig 2 so that we can all have a chance to listen? If
you are interested in a special clinic for a Roland demo, please send mail
to SUBSYS::ORIN. It would probably be on a friday evening around 7:30 to
8pm and would be open to all DECMS members, friends and family. It would
be at Worcester E.U. Wurlitzer again, with Eddie Fritz providing the demo.
If there is any other special equipment you are interested in, please send
me the info.

rsvp to SUBSYS::ORIN

dave

    My current understanding of the differences among Roland's D family
    of LA synths is as follows:
    
    			bar	pat				rack/
    		arch	seq	seq	disk	prog	outs	key
    
    MT-32	MT-32	no	no	no	no	2	rack
    D-10	MT-32	no	yes	no	yes1	2	key1
    D-110	MT-32	no	no	no	yes1	8	rack
    D-20	MT-32	yes	yes	yes	yes1	2	key1
    D-50	D-50	no	no	no	yes2	2	key2
    D-550	D-50	no	no	no	yes2	2	rack
    
    arch:	D-50 architecture has EQ, aftertouch, chorus, chase play
		and 2 additional LFOs (MT-32 cannot LFO modulate TVF, TVA).
		D-50/550 reverb and delay said by Roland to be of "higher
		quality" than in MT-32 synths.  MT-32 synths are 8-timbral.
		D-50 synths are 2-timbral.  MT-32 synths can play up to 32
		single partial voices, D-50 synths can play up to 16 two-
		partial voices.  Both can play 8 4-partial voices.  MT-32
		synths include 64 drum sounds, separate from partials.
    
    bar seq:	D-20 has 16000 note 8 track bar oriented sequencer.
    
    pat seq:	D-20, D-10 have pattern sequencer for drum parts.

    disk:	D-20 has onboard 3.5" disk drive, stores 76000 notes.
    
    prog:	front panel form based programming interface with
    		up/down/left/right buttons.  Yes1: PG-10 outboard 
		programmer (8 assignable sliders) optional;  Yes2:
		PG-1000 outboard programmer (a gazillion dedicated 
		sliders) optional.
    
    outs:	number of outputs.
    
    rack/key:	rack mounted or 5 octave keyboard.  Key1: velocity;
		Key2: velocity and aftertouch.

    

    I programmed up a D-550 patch from scratch last night (I cheated;
    I got it out of a magazine - I was more interested in what it was
    like to program a whole patch from scratch than the specific sound).
            
    It's quite an experience.  Consider:
    
    	The simplest patch you can program is a single tone patch based
    	on one PCM partials.  This requires 107 parameter settings,
    	which you get at by interacting with 29 forms of 27 different
    	types.
    
    	The hairiest patch you can program is a double tone patch, with
    	each tone based on two synth partials.  This requires 310
    	parameter settings, represented by 97 forms of 31 distinct types.
     
    The forms interface goes 4 deep.  You can define direct access to
    8 of the 97 forms you need to access to program the beast, and then
    call these up with the touch of two buttons.  This is pretty useful,
    as getting to a particular partial parameter requires the following
    sequence of operations:
    
    	select patch
    	EDIT
    	cursor to TONE field
    	ENTER               
    	cursor to PARTIAL field
    	ENTER
    	cursor to PARTIAL PARAMETER GROUP field
    	ENTER
    	NEXT/PREV FORM to PARTIAL PARAMETER form
    	cursor to PARTIAL PARAMETER field
    	set parameter value
    
    Getting out requires a sequence of 4 EXITs (one per ENTER/EDIT).
    
    One really neat thing the D-550 offers is the PARTIAL MUTE feature.
    This allows you to mute any of the 4 partials contributing to a
    tone, so you can listen to only the specific partial you're editing.
             
    You also have to be careful not to lose track of where you are.
    There are, for example, 8 instances each of the Envelope Time and
    Envelope Level forms (one for the TVF, and one for the TVA, for each
    of the 4 partials in a patch).  They're each labeled, but succinctly,
    and otherwise they look identical.
    
    No wonder the thing sounds so great!  310 parameters!
    
    And I got it to dump/load to/from the MRB-500 bulk librarian
    effortlessly!
    
    One interesting deviation from Roland tradition is the fact that
    each patch's tones are its and its alone.  It is not possible, as
    it is on the JX-10 and Super Jupiter, to share tones among patches.
    The D-550 provides a tone copy function to faciliate this sort of
    thing, but the main thing is that, unlike the JX-10 and MKS-80,
    you can have up to 256 tones (with a ROM/RAM card plugged in) in
    the machine!
    
    			MKS-80	  JX-10	    D-550
    		       int cart	 int cart  int card
    
    	user tones      64  128   50  50   128  128
    	preset tones	 0    0	  50   0     0    0
    	patches		64  128   64  64    64   64
    
    
    len.
    	

    I played around some with the D-550's envelopes, and discovered
    a few interesting things.
    
    First, the times for the first stage (t1) are rather different from
    the times for the rest of the stages.  My stopwatched (i.e., not
    accurate) times for the various parameter values are as follows:
       
    	parameter	       time (in secs) for value:
    		  value:  0  60  70  80  90  92  94  96  98 100
    
    	t1                0  .6   1 2.5   5   6   7 8.5  10  12
    
    	t2-t3		  0   1   3   9  25  33  40  48  58  74
    
    Note in particular the nonlinearity of the mapping.
             
    I measured these times by setting up envelopes that had "obvious"
    events at the end of the stage of interest. I used the following
    TVA envelopes for my measurements:
    
    	for t1,  l1  l2  l3  susl endl	  t1  t2  t3  t4  t5
    	grow	100   0   0    0    0      x   0   0   0   0
    
        for t2,
    	         l1  l2  l3  susl endl    t1  t2  t3  t4  t5
        decay:  100   0 100    0    0      0   x   0  50   0
        grow:     0 100   0    0    0      0   x   0   0   0
    
        for t3,     
                 l1  l2  l3  susl endl    t1  t2  t3  t4  t5
        decay:    0 100   0  100    0      0   0   x   0   0
        grow:     0   0 100    0    0      0   0   x   0   0
    
    I got slightly different measurements for the decay vs. grow
    variations, and slightly different values for t2 vs. t3, but they
    were close enough for my purposes.
    
    I did not measure t4 or t5 values.  t5 is hard to measure because
    without some real instrumentation, it's hard to tell when the 
    TVA envelope has really decayed all the way to 0.  Also, there's
    no obvious "event" possible after t4.  And, I did not measure times
    for the pitch envelope or the TVF envelope.    
    
    I confirmed Roland's admonition that if two successive levels are
    "close together", the time to go from the first to the second may
    be much shorter than expected.  Indeed - if the levels are the same,
    the time goes to zero, regardless of the parameter value.  This
    means you can't readily use the early stages of the envelopes as "delays".
    It also means the "time" parameters behave more like "rate" parameters,
    at least for small differences in adjacent levels.  I need to do
    more experiments to see just how pronounced this effect is, specifically
    at what difference in adjacent levels the time becomes difference
    independent (i.e., the time parameter is truly a time rather than a
    rate).
    
    len.
    

    Nonlinear programming of times, frequencies, etc. is pretty common
    on digital synths, Len.
    
    I made a similar table for the ESQ-1; and then found an even 
    simpler mnemonic:
    
    	ESQ-1 LFO and ENV times are logarithms, the log base is
    	approximately 1.15 ...  
    		
    		SO
    
    	To double/halve a time or LFO frequency on the ESQ-1, just add
    	(or subtract) five counts from the relevant parameter.
    
    So, to halve the speed of any LFO, subtract 5 from the LFO's
    current FREQ. parameter.  Likewise, to double the time of any
    envelope segment, add 5 to the Tn parameter for that segment.
                          
    	-Bill

    I'm not surprised by the nonlinearity, I was expecting it, and just
    wanted to call attention to it, as I've only rarely seen actual
    value to time maps of the sort I measured.  I have yet to try to
    derive the sort of rule for the D-550 that you have for the ESQ.
    I really wish the synth manufacturers would provide this sort of
    data (even if only in an appendix) for us technodweebweenies.
    It's not like they don't know.  Uhm, isn't it?  I mean, they do
    know about this stuff, don't they?
    
    Actually, I was really only interested in finding out what the longest
    times possible for the envelopes were.  Usually, I just do things
    by ear.
    
    len.
      

    Actually, the ESQ manuals *do* provide that data, in tabular format no
    less.  If anything has ever kept me from buying Roland, it's the
    documentation (or rather, lack thereof). 

-b

Len -

On the D50, if you press and hold "SHIFT" and then press exit, it will take
you back to the top level menu directly. Does the D550 have this feature?
I'm also curious if the D50 maintenance functions described in the USENET
note I posted works on the D550?

dave

    For those of you who were wondering, logarithms allow digital synths
    to perform the ``multiplication'' required by envelope scaling with
    addition only (no 70 ns TRW chips in these things).
    
    Steph

    re .20 - I haven't tried that yet.  Will do.  Re the maintenance
    features from the Usenet article, it references a "space" key,
    which the D-550 doesn't have, so I didn't know what else to try.
    
    len.
     

    How useful do you find the ability to program a different set of FX
    with each patch, Len?  Is there really a benefit to this, or are most
    of the patches simply chorus/reverbed to fatten 'em up? 

-b

    Yeah, it also means for live situations that you don't have to
    coordinate your D-50 patches with external sound processing.
    
    I wish the SQ-80 had that. 
    
    	db

RE: .24

    Of course, you're talking multiple FX per patch, right?  You're not
    really limited to just verb or just delay or ... right?  Howzit work?

-b

    Just enable patch-change output, and plug a DSP-128 just downstream
    on both audio and MIDI....
    
    The overall cost will be about what the D-50 goes for, you'll have
    poly-pressure, and the disk drive too.
    
    	-Bill (cursing_his_single_mono_effects_channel_after_the_fact...)
    

RE: .26

>    Of course, you're talking multiple FX per patch, right?  You're not
>    really limited to just verb or just delay or ... right?  Howzit work?

On the D50, each tone has a set of common parameters. Two tones combined
make a patch. The tones are called upper tone and lower tone. The common
parameters include 8 different chorus/flanging fx, EQ, modulation, pitch
envelope, and structure. The D50 manual calls the chorus types...

Chorus 1
Chorus 2
Flanger 1
Flanger 2
Feed Back Chorus
Tremolo
Chorus Tremolo
Dimension

This means that each of the tones in a patch can have a different chorus
type, EQ, mod, pitch envelope, and structure. For patches, the relative
tone volumes can be variable or fixed. Finally, at the output mode stage
of the patch, one of 32 reverb types can be selected and mixed. The
reverb types are...

Small, medium, or large hall
chapel
box
small metal room
small, medium, medium large, or large room
single delay 100ms
cross delay 160ms, 200ms, or 132-264ms
short, or long gate
bright hall
large cave
steel pan
delay 210ms, 300ms
cross delay 140ms, 225ms, or 224-122ms
gate reverb
reverse gate short, or long
one of 3 slap backs
twisted space
space

There are four reverb output modes...

stereo reverb, works on the mixed sound of both tones and is sent out in stereo
mixture of tones takes on stereo reverb, direct sound sent out separately
only the upper tone takes on reverb, upper and lower tones sent out separately
only the lower tone takes on reverb, upper and lower tones sent out separately

dave

  The various ROM cards from Roland with new patches also provide new reverb
  types which can also be saved off the card, so one can collect his favorite
  reverb programs.

Chad

    I usually program with all the effects disabled (something Roland
    has made relatively easy to do - there's a "4 position" global
    reverb level, and the chorus level can be set to 0 temporarily).
                 
    If you'll pardon my male chauvinist pigginess, this is sort of like
    checking out your new honey first thing in the morning before she
    gets a chance to "put her face on".  If she looks good (or the patch
    sounds good), you can be sure she's (it's) gonna look (sound) dynamite
    with the special effects.
    
    The effects in the D-50/550 are nice, no question.  The chorus is
    excellent, I wish the reverb had the same kind of EQ capability
    that the SRV-2000 has, but what do you want from a "feature" of
    a box that costs about the same as an SRV itself? 
                         
    There must be some user-accesible way to program the reverb...
    probably via some SysEx magic.
    
    len.
    

    re .16
    I have the Steinberg editor for the D50, and the documentation mentions
    that for the same value of T, rise times are shorter than fall times
    "cos thats what real instruments are like". Also, when you draw envelopes
    using the editor, it displays the real-time value of the time parameter
    you are editing. I've never looked to hard at this, preferring to
    program by ear, so I don't know how accurate it is. Next time I'll
    compare your results with their displayed values and let you know
    of any interesting results.                        
                                                       
    re .20
    I must have missed your posting of the D50 maintenance functions
    on this notesfile. Can anyone give me a pointer to them?
    
    re .23
    I'd really endorse the positive replies to this question. Having
    programmable effects built in really allows you to think of them
    as an integral part of the patch - and it's going to be that way
    whether your synth is linked into your full system or whether you've
    just hauled it out for a quick informal jam session. It's also really
    good for gigs - less setup time, less to go wrong with connections
    etc, and more speed over non-midi effects boxes in changing between
    patches.
    
    Alex

    re .31 re .16 - my measurements indicate that it's not so much that
    rise times are different from fall times (the times that I measure
    were much the same, regardless of direction, for a given segment),
    as it is the difference between segment 1 and all the rest.
    
           re .20 - the maintenance stuff came off Usenet.  I have aprinted
    copy but didn't save the online version.  It was not of much value
    to me, as the D-550 doesn't have the D-50's numeric keypad, so the
    D-550's maintenance stuff is gotten at in a different, and as yet
    unknown, fashion.
    
    	-------------------------------------------------------
    
    One cute little bit of D-550 (and probably D-50; Dave Orin related
    a similar experience with his D-50 to me) weirdness:
    
    I set up a "fully initialized" patch with all 4 partials playing
    an unmodified synthesized pulse wave.  On certain notes (on my D-550,
    it seems to be Ds), the partials will destructively interfere with
    one another in varying degrees.  The variability seems random.
    It is as if the relative phases of the partials were more or less
    random, at least I can'r detect any pattern.  I'm pretty sure it's
    destructive interference due to phase differences because muting
    one or more of the partials brings the sound back!  With pulse
    waves the interference is blatantly obvious.  I wonder if similar
    things (i.e., random relative phase among partials) happens with
    other (more complex) waveforms (or with the samples), and what effect
    it has on the sound.
    
    Is this a feature or a bug?  Time for another letter to Roland (who
    still hasn't responded to my missives about my MRB-500/MKS-80
    adventures)?
    
    len.
    
    

    It's a feature.  The "muting" is mathematically supposed to be there,
    and some synthesis hacks (like hard sync) won't work if you don't
    allow destructive interference.
    
    If you don't want it, detune the oscillators from each other and don't
    use hard sync (do the D-series even have hard sync?)
     

    I guess I disagree.  The problem is not that it happens, but that
    it happens randomly.  Detuning will not help, as it will just introduce
    a long period modulation to the destructive interference.
    
    The "muting" I mentioned is *not* the destructive interference which
    is "mathematically supposed to be there".  It's a user interface
    feature of the D-50/550 which allows you to disable the output of
    one or more of the partials with a single button press.  I mentioned
    it as proof of the destructive interference - if you mute one of
    the participating partials, it can't destructively interfere with
    the other.  Thus: one partial - sound; two partials - no sound.
    Hence, destructive interference.  The "muting" is of one of the
    contributing waveforms, rather than of the sum waveform.
    
    Sync (hard or soft) does not depend on the opportunity for destructive
    interference.  Sync only means that one waveform's cycle starts
    are determined by another waveform's cycles (usually the zero
    crossings).  I.e., when one waveform (the sync source) has completed
    its cycle (i.e., its phase angle (modulo 2 pi) has gone "all the way"
    around to zero) the other waveform (the slave) is *forced* ("synced")
    to a phase angle of 0.  Note that destructive interference can only
    occur (due to phase relationships) with waveforms that exhibit certain
    symmetry properties (i.e., at some phase angle, the waveform is the
    inverse of the waveform at zero phase angle).  Syncing can be applied
    to *any* pair of periodic waveforms.  And I wouldn't call it a "hack",
    it's a very powerful mechanism. 
    
    And, no, the D series does not have any sync capability.  The JX-10
    and MKS-80 do.  Neither of them exhibits destructive interference
    among their oscillators.
                         
    len.
                                        

    This sounds like one of those interesting implementation-specific
    things (like the Moog filter).  Is it useful?  I tend to like
    instruments that can give you a slightly different per-key behavior for
    the current patch or sample, but usually this is considered a bug,
    and only appears by accident.
    
    I would expect beating to add fullness to some sounds, does it?  On
    an ideal D50-like machine, I'd expect to be able to adjust the phase
    angle between partials--though I suppose that you can already do
    this via fine tuning.  
    
    Is it consistent, always the same beating on the same note, no matter
    what else is happening?  I'm wondering if perhaps the phase delay is
    due to processor busyness.  It strikes me as possible that if the CPU
    is very busy, it might have noticable (phase) lag between the times
    that it 'starts up' the various partials for each note.  I would expect
    the beating to vary if this were the case.
    
    	Eirikur

    Well, actually, they do.  Here's how.
    
    Put the sync source as a square wave (which has all odd harmonics
    only), and the recipient wave as a sine, at 3x frequency of the square
    wave.  Now, subtract them.  Sounds kind of nice, eh, with that missing
    third harmonic.  Very airy and clean.
         
    You need to use hard sync to do this, because analog oscillators
    won't phase-lock and digital oscillators are rarely phase programmable.
                              
    You can get a fun modulation on this by using two analog oscillators
    and detuning them.  Sounds very classy.
    -----
    
    I used the word "muting" to refer both to the button effect and
    the destructive interference effect.  That's probably wrong, it's
    certainly the cause of the confusion. :-(
    
    -----
    
    You can "fix" it as they do in orchestras: route one oscillator
    only per speaker channel.  Then the destructive/constructive phenomenon
    will be at the listener, whose two ears+pinnas+constant head movement
    will hide it pretty well.              

    re .35 - this is not a matter of beating, this is full blown
    destructive interference.  The waveforms are in sync: the phase
    angles, once established, do not shift relative to one another.
    Depending on the (apparently random) initial phases (and only for
    certain notes), you get anything from no interference to partial
    interference ("thinner" tone at lower volume) to almost complete
    interference (extremely "thin" tone at almost no volume - qualifies
    as "noise").  If it was "reliable", I might be willing to consider
    it a feature.  It certainly doesn't make things more interesting.
    
    re .36 - what you say is true, but it still does not imply that
    sync effects *require* the possibility of destructive interference.
    (At least that's what I tought you were arguing.)  In fact, in the
    example you have given it is the other way around; the destructive
    interference *requires* sync for consistent suppression of the third
    harmonic.
    
    I'm not interesting in "fixing" it except correctly.  If I want
    randomness, I want it under my control, not at the synth's whim
    only on certain notes.  That's why I consider this a bug.
                             
    len.
    

    < Note 1499.36 by CTHULU::YERAZUNIS "The light that burns twice as bright burns half as long" >
>    You can "fix" it as they do in orchestras: route one oscillator
>    only per speaker channel.  Then the destructive/constructive phenomenon
>    will be at the listener, whose two ears+pinnas+constant head movement
>    will hide it pretty well.              

    .. unless the recording you're making or the sound system you're
    playing thru somehow ends up mono.. then it's right back again.

    I give len's phenomenon a name.. "Phase Cancellation".
    
    karl

Subject: D-50 PG-1000

(from usenet)

In most synths there are a few hidden functions that are used for repairs.
The keys must be pressed first then turn the synth on after...
 
D-50:   SPACE + DECREMENT : Test mode.  You can test all buttons, foot switch
                            and notes (for velocity + pressure)  Also checks
                            memory and a card...
          "   + INCREMENT : Gives credits and ROM version.....
          "   + WRITE     : Tests L.A. Chip.  You choose exact test by hitting
                            UPPER(Audio test), LOWER(D/A test??), VAL(mute) and
                            LOCAL(exit!).
          "   + DATA TRANS: Initilize global paramaters such as tuning, patch
                            number for startup etc....
                            Does NOT affect Patches (don't quote me on this...)
 
PG-1000: PARA REQ : Button/Led test
         MANUAL   : ROM version
         PARA + MAN : Credits....
 
By the way my ROM showed 1.06 or so.  I think the latest is 2.00
Also the velocity for notes was always from 09 to 127.  

dave

    re .39
    Latest version of D50 control ROM I've seen is Vn 2.2. Major difference
    to Vn 1.6 I know of is that 2.2 allows programmable control of MIDI 
    program change numbers sent on each patch (rather than fixed allocation).
                                                             
    Roland UK say they will supply upgrade ROMs at no charge.

    Alex            

    Anybody know if D-50 and D-550 roms are compatible (i.e., the same)?
    
    len (D-550 owner).
    

Len -

Eddie at Wurly's says the D50 and D550 roms are compatible. He is going to
call Monday to find out when they will be available. They are supposed to
be supplied at no charge.

dav

>Eddie at Wurly's says the D50 and D550 roms are compatible. He is going to
>call Monday to find out when they will be available. They are supposed to
>be supplied at no charge.


Are we talking the ROM sound cards with the extra patches?  If so, they
are compatible but not free.  They list for like $80-90 a piece but mail order
as low as 50-60.  Daddy's is having their MIDI month in August and have
D-50 ROMs at 69.00.

CHad

    No, these are the software roms, in chips.
    
    len.
    

    
    
    Well Mr Yamamoto will be very umpleased with me but I bought my
    first Roland(D50) synth this weekend. I also got an SPX50 on loan
    just to keep the Yamaha rep happy. 
    
    SO far so good. It complements the other stuff well. I wouldn't
    like it to be my only synth though. I also got the Steinberg D50
    editor which makes a better instruction manual than the Roland version.
    And have already one or two usefull patches. 
    
    I have also used it successfully with Steph Bailey's Bulkdump sysex
    storage program. So those of you with Atari's, we could get into
    some  good patch exchanging here.             
    
    I have not bought any of the ROM cards yet? I have never bought
    patches before but I could loan them from my dealer. Is it worth
    the hassle or are they a waste of time.   
                      
    						Paul.

    RE .45
    
    I don't believe you. This is a wind-up, like the KX and the Atari
    and the Akai, isn't it? Your studio consists of a mouth organ, a
    jaws harp and a 4-string banjo with 3 strings, doesn't it?
    
    Richard.

    Two questions:

    1) How is a Roland D550 powered ? External power supply or built in
       power supply ? And if its built-in, is it switchable to 220 Volts (i.e.
       can it be used in Europe?)

    2) How much ($$$) is a D550 today ?
    
    My brother-in-law (lives in LA) will be coming to Europe in Xmas. 
    and I may ask I'll ask him to pick one up for me.

    Zayed

    The D-550 has a builtin power supply (i.e., it has a line cord).
    I believe the power supply can be configured (by opening up the
    box and rerouting some connections), but this is probably considered
    a service operation, and doing it yourself will probably void the
    warranty.
    
    A good deal on a D-550 these days would be about $1200 - $1300,
    maybe less.
    
    len.
    

    Len, thanks for the info.

    I called a mail-order store in Zurich and a D550 is going for
    about $1755 ! Looks tempting to get a US version but you're right
    about the voiding the warranty.

    What are they going for in the UK ? Anyone know ?

    Zayed

    RE: .49
    
    Zayed,
    There is also a board that was advertized over here in Keyboard
    Magazine which will make a D50 (and I assume D550) multi-timbral.
    
    It's put out by Steinberg-Jones and is called the M-EX. I'm not
    sure if this is included in another topic discussion in this conference.
    
    I don't have the ad with me, but it makes the D50 8 channel
    multi-timbral, and gives it 2 new modes - Multi mode and Multi-Dual
    mode. Voice allocation is dynamic.
    
    It also expands the internal memory for an additional 64 (I think)
    patches. There is an option for an additional 64 also on top of that.
    I was told by a sales person at a local shop that it comes with a bank
    of new sounds.
    
    It lists for $425.00. If you have one brought over for you, you might
    want to consider having it installed. You may be able to have it done
    at the shop where it's bought (maybe for less than this).
    
    Regards!
    Bill