Conference napalm::commusic_v1

    Richard, I don't claim to be an expert or even *good* at sampling,
    but I have noticed a few things (potential problems) in attempting
    the sampling of sounds off of other synths and huge layered sounds,
    etc.
    
    One problem is with sampling a synth sound that has a built-in vibrato
     sound, like a flute patch. When sampled and played back, the frequency
    of the vibrato will change with the frequency of the note you're
    playing (the higher you go the faster the vibrato, the lower you
    go the slower it will be). This can be worked around if you can
    leave the vibrato off of the sound you're trying to sample. But
    this isn't always possible and/or desirable. This same problem occurs
    if you're trying to sample and "sample and hold" filter sound off
    of an old analog synth. [The best example of this type of sound
    that comes to mind is on ELP's Brain Salad Surgery album, Karn Evil
    9, 2nd Impression? the sound that's heard before the line "...welcome
    back my friend to the show that never ends..."]
    
    Another problem, which seems to diminish as the quality of your
    sampling machine increases, is that it is harder to loop the big
    textured sounds because the complexity of the waveform makes it
    more difficult to find a smooth splice point. Since higher end machines
    offer longer sampling times and auto loop facilities this type
    of thing is not a problem, but with lower end (like my Mirage) these
    layered sounds can be a bear to loop cleanly. This same problem
    can happen when trying to sample sounds that have effects (like
    reverb etc.) already on them. I think it's easier to sample dry
    and add the effects later (at least on the Mirage it is).
    
    For the most part though, what you're talking about CAN be and IS
    accomplished routinely. I read an article on Keith Emerson in Keyboard Mag
    saying that he was trying to get his Yamaha GX1 (or whatever it
    is) to work just long enough to get a good sample of it (I guess
    it was severely damaged when a runaway tractor crashed into the
    barn [Emerson's studio] in which it was set up). Keith chose this
    option because he wanted his "signature" sound, but he knew he couldn't
    lug his unreliable old keyboard around anymore.
    
    Actually, I've had some limited success sampling 
    other keyboards into my Mirage.
    I've got an old ARP Odyssey, (my 1st synth!) that I bought back in
    '73, that won't stay in tune anymore so I sampled some of it's sounds
    into the Mirage. It works OK but it's not quite the same because,
    for instance, the Mirage doesn't have a Portamento control, so even
    if I've got some of those good old analog sounds I still can't get the
    same effect (if I want the effect of a portamento). But I guess that's
    a limitation of the Mirage itself
    and not of this technique (sampling other synths) in general.
    When I get a higher end sampler, I'm going to rent a bunch of different
    synths (and other assorted keyboards) and sample them all, but it
    is a lot of work for me (or anyone) to do this on the Mirage.
     
    Well I'm probably boring everybody to death so I'll stop.
    
    Mike D

    Sounds that have time dependant variables in them don't translate
    well through sampling.
    
    Edd

    
    I agree with Edd, and Mike to a point. There are plenty of samples
    that you can make of other people's equipment (heck, you can even
    resample someone else's samples of some other machine)... But there
    are definite caveats like no portamento, etc. 
    
    Of course, on the EPS, since you have a selectable filter, you could
    be entirely gross and sample a sawtooth wave or three, and filter
    'em up and down to get that "analog sound"...
    
    I guess that's why I've hung onto my Korg MS-20 (wasn't my first
    synth but I've had it almost forever) so long. 
    
    /pjh

    re .0
    
    Richard brought up a point that I've been wondering about for a
    bit. He mentioned the 'affordable' AKAI sampler. While I've seen
    the Mirage going for some pretty low prices, I've never been able
    to verify some of the low prices I've heard on AKAI's equipment.
    Anyone have some ball park figures on what I should be expecting
    pricewise from the AKAI. And for that matter, who carries them?
    
    dp

    My two synths are currently a Roland D 50 and an Ensonique ESQ M.
    I'd like to get a drum machine, but could I buy a sampler
    (EPS maybe?), record the ESQ M sounds on it, sell the ESQ M,
    not buy a drum machine (use the sampler), and effectively
    have an ESQ M, drum machine, sampler and save $$$?
    
    I know I wouldn't have the Ensonique parameter controls anymore,
    and some sounds might be a problem - i. e. the flute/vibrato
    example, but I'd be saving myself $$$. Would I have a real good
    reason to keep the Ensonique still?
    
    Also - is this legal - (since the EPS is still Ensonique, they
    probably won't care), but what are the legal problems/concerns
    in general, if any?
    
    Bill
    

    
    re: Bill .-1
    
      As far as legality goes, this is an exceptionally gray area. If
    you take a sound off of a record, you are technically committing
    a copyright infringment. If you take a sound from someone else's
    machine, it's not their data you're stealing, it's their *sound*.
    
    The copyrights on software, to my knowledge, only protect physical
    data, i.e. the actual format (stream of numbers) cannot be identical.
    Sounds are another matter. When I was writing video games back in
    college I was told by my lawyer that I could copyright the programs
    but I would have to file a separate copyright on the sound effects.
    
    In the case of Ensoniq, they artificially limit themselves by requiring
    you to buy certain samples (and they copyright their disks). But
    I doubt that they could take you to court if your S-50 took a sample
    of a loaded EPS set of sounds. 
    
    Really - one of my EPS disks (copyrighted by Ensoniq) has the D-50
    "Fantasia" on it - a very good sample I might add.
    
    As far as vibrato, filtering, and other flavors of modulation, your
    sampler should be able to process them locally, i.e. not directly
    within the sample but internal to the machine. The EPS does, at
    any rate, and I believe the S-50 does (right, db?).
    
    /pjh

    
    There're all the pitch/time dependent effects already mentioned, but
    there are also velocity and aftertouch considerations as well.  On my
    Super Jupiter, velocity can affect the envelope times as well as
    levels.  Also, the envelope times can track the keyboard.  The
    envelopes do more than control just a VCA - they can control oscillator
    pulse width modulation, oscillator pitch, oscillator mix, oscillator
    cross modulation, and the usual VCF cutoff frequency.  It's going to
    be quite a trick getting these kinds of effects to work the same way
    on a sampler, even if you had a sample per note.
                                         
    So, it's pretty clear to me that, no, samplers will not make other
    synths redundant or obsolete.  They're just another sound generation
    technology, and quite a useful one at that.  Just like subtractive
    analog, additive analog, digital, phase distortion, FM, and all
    the hybrids possible.  Note that some of the most interesting recent
    new synth designs incorporate sampling at the "oscillator" level,
    e.g., the Roland D-series synths.
                                             
    len.
    

    It occurs to me that the long term trend may be to add all the
    traditional analog processing (like the envelope-controlled factors
    mentioned in .7) to the basic sampler.  Then what a sampler becomes
    is a subtractive analog synth with oscillators capable of arbitrary
    waveforms.  Which wouldn't be a bad place to be...
    
    len.
    

    
    re: .-1
    
      Which is pretty close to an ESQ-1 except that you're stuck with
    the ROM-based waveforms...
    
      The EPS comes very close to this also - with the waveform
    manipulation and crossfading capability, and three EGs, two filters,
    an LFO, selectable output panning/assignment, etc., you come very
    close.
    
    Whatever
    
    /pjh

    Some thoughts...
    
    On the issue of samplers 'obsoleting' synths:
    
    One thing you can do with a synth that is not very easy on a sampler is
    "design" a sound, or customize a sound.  I don't know how other people
    feel, but that capability is very important to me, and for that reason
    I always expect to have a number of good synths around.
    
    One other thing is that samplers seem to have a higher cost/voice
    ratio than synths.
    
    On the issue of copying sounds from synths into samplers:
    
    I have to claim that I don't know much about samplers.  I will soon
    when my tax refund arrives and (moments later) my S-550 is in my hands.
    
    Basically what distinguishes synths from each other, is what I refer
    to as their "architecture" (I don't claim that this term is my own
    invention).  Like computers, most synths have more in common than
    not.  They have oscillators producing various kinds of wave forms, 
    filters, envelopes, etc.  Even FM is not really all that different.
    It's generally the arrangement of how you can use these things together 
    that give a synth it's uniqueness and character.
    
    Now I don't know much about sampling architectures.  But I do know that 
    they ARE different, or that they have basic elements that don't apply
    to normal synthesis.
    
    It occurs to me that all of the problems described about copying synth
    patches are caused by the flexible architectures of the synth. 
    Certainly if you hope to 'lift' a synth patch merely by recording it,
    you are gonna have a hard time doing it without running out of memory
    or spending a lot of time.
    
    What I would like to know is samplers have the same kind of
    architectural things as synths do?  My thought is that if they
    do, you may not be able to just "record" synth patches into samplers,
    *BUT* you can probably duplicate almost any synth patch if you
    copy the the structure of the patch itself (the software), and mimic
    it on your sampler.
    
    Let's take the ESQ-1 for example.  One of the things I liked about it
    is that you can use almost any modulator to modulate almost anything
    that can be modulated.  It's very flexible.
    
    I would like to have a sampler that allows me to "record" the basic
    wave forms, and then do with them all the things that I can do with
    waveforms built into the ESQ-1.  If that were the case, I truly
    wouldn't need the ESQ-1 for much of anything.
    
    Suppose I have a sampled wave form.  Can the samplers that are
    available today:
    
    1) Modulate the pitch of the wave form?
    
    2) Combine several separate samples?  
    
    3) Modulate the volume of each sample being played.  (I suspect yes,
       cause this is what is required to do cross-fades.)
    
    4) What modulators do you have (velocity, wheels, x-controllers, LFO's
       Env/adsrs).
    
    5) What kind of filtering can be done?
    
    In short, perhaps it's difficult to 'lift' a synth patch just by
    recording it, but is it possible to do it by "duplicating" the
    patch?
    
    Or can you even make sense of what I've asked?
    
    	db

    I think ``ordinary synths'' have two advantages over samplers, at
    least for my own work.
    
    First, I have to disagree (somewhat) with Edd's thesis.  I admit that I
    have little direct experience with samplers, but there is some
    ``theory'' (<--often a euphemism for ``an incomplete understanding'')
    which is relevent.
    
    Samplers do well at capturing time-domain modulation, but may not
    do well at capturing performance modulation.
    
    Specifically, if you sample a sound, it is just as easy for a sampler
    to capture it if it has huge temporal variation (e.g. the 1812 in
    its entirety) as if it has little temporal variation (e.g. a pure sine
    wave).  Whence the infidelity comes is if you try to ``perform''
    with this sound.
    
    Performance parameters are things like key velocity and number,
    and controller positions, and so forth.
    
    The example of vibrato speed varying with note number is a failure
    of the sampler to fully emulate the performance characteristics
    of the sound source.
    
    Some performance characteristics are easy to emulate with a sampler.
    For example, the velocity response of a synth can be emulated w/i
    velocity multisampling (cross-fades).  The key-position modulation can
    be provided with key-number multi-sampling. 
    
    Some of the LFO-style modulations provided by conventional synths are
    hard for samplers to emulate (e.g. modulating FM carrier amount) and
    some are easy (e.g. modulating output amplitude).
    
    
    Second, I think you lose much of the serendipity of programming a synth
    with a sampler.  In your ``huge stack'' generation example, I don't
    think that I would have the patience to do that.  For me, even when I
    work in the studio, performance is foremost--there must be a
    performance before there is a recording.  I never did have much
    patience with step time entry.
    
    Before anybody flames--I concur that you get a different, but still
    eminently useful type of flexibility with a sampler.
    
    Steph

    When my keyboard player reaches up to his Moog & changes things
    while he is holding down a note, he gets real-time modifications.
    This seems to be something missing from music these days, and at
    the present really can't be sampled, or done on any available digital
    synths (at least that I've seen). While this may seem like a minor
    issue, it is something that certain devices do better than others.
    I'd love to get a rack mount Mirage, or something similar, since
    it expands on what I can already product, but, I like the personality
    of my other toys also.
    
    	Jens_who_likes_Flat_top_Martin_guitars_even_if_they_haven't_come
        _up_with_any_innovations_in_design_since_1850_(referance:George_
        Gruhn_in_Guitar_Player_Magazine)_and_still_likes_his_CZ-101 

    > ... or done on any available digital synths ...
    
    This is patently false.  The DX series provides real-time parameter
    variation, as does the K5, the Korg DS8 (and 707), and I'll bet most
    other digital synths (ESQ?, D50?) do too.
    
    Steph

    The high end Roland synths allow it too, though it's much much easier
    if you have the optional programmer so you can get "random access"
    to all the parameters.  You can also record the parameter changes
    into a sequencer (the synths transmit the parameter variations as
    system exclusive data) so they can be played back.
    
    While I don't use this feature ("real time" parameter variation)
    in "performance" (since I'se just a sequencing guy), I make extensive
    use of it while programming.  Move the slider (or turn the knob),
    hear the change in the sound.
    
    Note that on some synths (e.g., the CZ-101, don't know about the other
    CZ's) changing parameters will "chop up" the sound, so you can't do
    this in performance.

    len.

RE: .13

    ESQ architecture machines allow parameter variations, but in a rather
    bizarre fashion.  Sustained notes are not affected by tweaking
    parameters - only subsequent triggered notes will reflect the change. 

    Given that almost anything can be modulated by almost anything on these
    machines, the point becomes somewhat moot. 

-b

    but,but,but....
    
    To me, these days there is much less control availible in the
    real-time domain. Look. synths used to have scads of sliders
    to farkle in what ever way intended to get a certain effect
    depending on your mood or the performance. SOME synths today
    enable you to get to maybe 1 or 2 parameters (and gee, some
    will even give you lotsa sliders for $500 more as an option)
    in 'real-time', but most require you to program that 'slow
    hi-Q filter sweep', etc....
    
    Unless you want to enable parameter 39 and hold down the
    inc or dec buttons and get lotsa quantization (step) noise...
    
    There IS something about  a 'preset' mentality today that not only
    pervades the synth user-interface, but also is evident 
    
    			IN THE MUSIC PRODUCED 
    
    by the technology/artist partnership.
    
    ron
    
    
    

    re cheap Akai samplers
    
    I saw in (I think) Music Emporium, or maybe West LA Music, that
    they were selling S700s for $795.
    
    S612 with the MD280 drive can be had for around $350 used.
    

    Re: 'Pre-set mentality'...
    
         Couldn't agree more. Seems to relate directly to the proliferation
         of 'point and grunt' computers.
    
    Edd

    Well, I also couldn't agree more, but I don't think the "right way"
    is to go back to the days when synths had all those sliders and knobs.
    
    The "right way", IMO, is to make EVERTHING "modulatable" (another
    new Commusic term?) from EVERY conceivable modulator.  I think that
    the ESQ-1 was a BIG step towards that direction even though it still
    has a lot left to be desired.
    
    So rather than reaching up and finding the right knob in a forest of
    knobs and sliders, I would rather be able to "assign" what I what to
    modulate to my MOD wheel or my CV pedal.
    
    One thing that I hope and pray that Ensoniq will correct is the fact
    that the SPEED of the LFO's is not modulatable.  There are so many
    things that would benefit from that.  I've wasted several hours trying
    to figure out how to do something roughly akin to that using AM mode
    and/or getting "beats" (in the piano tuning sense of the word) by
    using multiple de-tuned oscillators (since you can modulate pitch).
    
    	db

    Another peeve I have on the ESQ-1's routing is that the filter cutoff
    is modulatable- but not the Q.  Maybe modulated Q is not a musically
    useful concept, but I want to prove that to myself.
    	
    Dave: you can approximate (not match, just approximate) the effect of a
    modulated-F LFO by having a slow LFO and a fast one; both with
    near-zero initial and terminal magnitudes, and using positive KYB2
    modulation on one and negative KYB2 modulation on the other.  Thus
    the frequency of the LFO sort-of appears to be note-number modulated.
    
    It aint' quite right, but it might be close enough... (and this
    is synthesis, not sampling, so close enough is good enough). 
                                                       

    Tried that.  Doesn't really work too smoothly.  The swtich from
    one LFO to the other is either very sudden or you get this thing
    where both LFO's are going at once and you get something that
    doesn't modulate uniformly (two independent wave forms is a neat
    effect, but not what I'm looking for).
    
    I'm hoping that the sampler I end up with allows me to do something
    like this.
    
    	db

    The "make everything modulatable (modulable?)" is the Oberheim
    approach.  Guess what guys, making everything modulatable also makes
    everything more expensive.  Not necessarily unaffordable, but certainly
    more expensive.  And making additional routings possible means some
    hardware additions as well as software, which has implications for
    manufacturing cost.  Now, I'm not trying to defend the synth designers,
    just trying to explain their point of view, which is probably strongly
    influence by the "80/20 rule"  or "good enough for most folks"
    mentality.
    
    Sometimes I wonder if the programmable synth is as much a curse
    as a blessing - with "old fashioned" modular synths, the "program"
    was actually an "architecture" (i.e., a selection of modules and
    the connections between them), not just a collection of parameter
    settings for a fixed-for-all-time set of modules wired up in one
    particular way.  Yeah, there're some switches that provide some
    routing options, but only a small fraction of what's actually possible.
    The designers make "optimizations" that are "compatible" with "real
    user requirements" so as to allow "practical, cost effective
    implementations".  "Ivory tower theorists" need not apply.  Sound
    familiar?
    
    This is ironic, given the possibilities that VLSI parts offer. 
    As "yet another spare time project", I've been sketching out the
    design of a "programmable modular" synth that has all the routing
    flexibility of old style modular synths and all the progamming
    convenience of the contemporary programmable synth.  A handful of
    basic module types (but with lots of instances of them) makes all
    kinds of neat things possible.  Whether they provide useful sound
    generation capabilities is another question.
    
    Love it up here in my ivory tower.
    
    len.
    

    The Emax has nice analog capabilites.. 5-param VCA and VCF envelopes,
    individual LFO per voice, LFO 'variation' mode, plus complete param-to-
    result programmability (velocity to filter cutoff, or aftertouch
    to panning or LFO speed/intensity, etc. Very warm-sounding.
    
    karl

    Len,
    
    My understanding of the ESQ-1 is that making everything modulatable
    is pretty much a software only modification.  I believe that the
    LFO's, envelopes, etc. are all software.
    
    Thus, making everything modulatable is pretty much independent ofd
    "manufacturing costs".  If I'm right, they could introduce it
    with a software upgrade.
    
    One of these days I would like to sit down with an Ensoniq rep and
    get an idea just how much of this machine is software and how much is
    hardware.
    
    	db

    
    db,  does the ESQ have anything like detune.  Most if not all of the
    Yamaha boxes offer a detune parameter on all  operators  to  produce
    just the effect that you are describing.
    
    _sjz

    < Note 1337.25 by DYO780::SCHAFER >
>RE: .23                                             
>I've diddled with an Emax, and have to agree with your opinions on the
>machine ... but don't you find 8 voices limiting?  For applications
>where you need or want to do multi-timbral concurrent polyphonic lines
>(say that 3 times fast), you run out of voices fast.  How do you deal
>with that? 

    At the time I investigated available samplers for eventual purchase,
    the end of 1986, the Emax was a clear winner, with a large library,
    ease of editing, and an internal 16track sequencer. 8 voices was
    the norm. The only close competition at the time was the SC Prophet
    2002 rack (we know what happened to Sequential) and the AKAI S900.
        
    I agree with your agreement.. and yes I *DO* find 8voices limiting.
    Remember that 8 voices CAN mean 16 timbres, in that 'dual' mode
    allows stacking of two timbres on each note, with positional and/or
    velocity crossfading available.
    
    .. but Emax is always used in conjunction with other equipment. Like
    a 24-voice Kurzweil 1000PX rack unit. The two together are simply
    superb. I'm in aural heaven. The 128 high-quality presets (esp.
    the stereo grand piano) in the Kurzweil, in conjunction with the
    (literally) hundreds of sounds (many in stereo) I have for the Emax 
    makes an unbeatable combination. The Emax' sound quality is superb, 
    and due to its architecture, diddling sounds in either the digital 
    or analog domain is extremely easy. 
    
    Check out Commusic V.
    
    karl

    Zach,
    
    If you're talking about simulating the "beats" of a piano, yes, that 
    is how I tried do it.
    
    However there were severe flaws with that approach, mainlly that the
    response is not uniform across the keyboard and the warbling that
    effect you get is really much different than the effect of speeding
    up the rotating horn on a Leslie.
    
    	db

    
    
    IN terms of the original note, Richard has supplied me with some
    cassettes of his multi-moog which I have sampled with varying amounts
    of success into my Akai. I also have the soundworks sample editor
    and librarian which allows me to do quite a bit of work on the sample
    I.E. Envelope, E.Q and Xfade looping adjustment. 
    
    I have to say that I am fairly happy with the results although
    obviously thos really lomg sweep patches are not possible with a
    short sample time. I can however patch the LFO and filter through to 
    Aftertouch on the KX88 so an amount of suitable Wooshyness and movement
    can be simulated with the performance parameters.
    
    I have also sampled other synths E.G. the TR808. With success and
    have found this a fairly food option to having the real thing and
    also getting midi-control over non midi-instruments. 
    
    Plus in the case of Richard's multimoog I have polyphony as well.
    Which is something not available on the original instrument.
    
                                                Paul.
    						
                                                        

 >   Plus in the case of Richard's multimoog I have polyphony as well.
 >    Which is something not available on the original instrument.
  
    I know that, I know that.
    
    Richard.  

    Will 'resynthesizers' make samplers obsolete ?  (soon)
    
    

    Since entering this note, I/ve bought an FZ1 sampler. I find that
    I can sample the moog at low sample rates (say 9k) for short periods
    and get really good results. I multitracked the thing first to get
    the slight detune that 18-oscillators-per-voice moogs have. Then
    diddle the filter envelope, add velocity and aftertouch modulation
    and there's a really good patch that takes up very little of that
    precious sampler memory. You can even squeeze the multisampled factory
    piano and a bunch of synth samples on board in one go.
    
    I'm very pleased with the sampler -- I think the next practical
    breakthrough will be some sort of memory backup that doesn't mean
    you rely on a disk-drive. I haven't played live with the thing,
    but I imagine it's always a bit Russian rouletty when you stick
    your piano software in and hope for the best. "Do I have a piano
    or a very heavy piece of quiet junk?"
    
    Richard.

    
    On just this topic I just spent the best 90 quid ever on a static
    memory upgrade for my akai sampler. It is now in the same state
    when I turn it on as it was when I turned it off which saves about4
    minutes and 15 keystrokes everytime I turn my system on not to mention
    the same every time I turn the system off. Included with the upgrade
    was more memory (now up to 16 onboard samples) plus a cross fade
    looping option. 
    
    I have become used to the Yamaha method of battery memory backup
    which means that that the only button I have to press now  when
    I turn on is the drum machine external clock switch.
    
    Why can't every manafacturer be as thoughtful !
    
    Incidently the new Yamaha TX16w sampler thingy takes 45 seconds
    to load the operating system and another 90 seconds for a full
    complement of samples. Imagine if every synth in the system took
    that long.
    
    Any thoughts ?
    
    					Paul.
     

    That's what I need. Bet I can't have one.
    
    I'm still trying to get hold of high  density diskettes. Every
    salesperson in the country assures me that double density is fine.
    Fine if you don't want high density, I suppose. Mind you, this is
    after asking me 'What model of machine do you have?'. When I reply,
    'It's a sort of piano, really' they think they can sell me anything.
                                                         
    Paul -- is the Akai memeory organised the same as the FZ (each sample
    can be assigned an area on the keyboard, and groups of 1 or more
    areas assigned to a bank)? If so, how much memory do you find you
    need for complex, multisampled sounds? The factory piano sample
    uses most of the 1 meg onboard the FZ. I think I may have to squeeze
    an extra meg in when my wife's not looking.
    
    Richard.
    

    Regarding the original topic to this note...
    
    Samplers can be programmed by computers that provide a familiar
    'programming model' or user interface...
    
    That is to say, a simple model of synthesis, say, subtractive analog,
    can be used to 'program' the sampler. The user enters the parameters
    just as if he were programing a Juno-2, the computer generates the
    equivalent sampler waveform, and downloads the sampler.
    
    Note that this really can't replace an analog synth, because the
    analog synth's parameters can be modified in real time.
    				- HBM

    The last EM had an article on a package that replaces the Mirage's
    MASOS and allows fancy synthesis.. don't remember the details.
    
    karl

    re: .37
    
    A friend with a Mirage got a demo disk of this (called Sound Designer
    I think?). It totally replaces the OS on the Mirage and turns it into
    an ESQ-like synth with replaceable/sampled waves. Some neat sounds
    on the demo disk. They claim that the sound quality is much better
    and brighter than the normal Mirage samples. We thought some of
    the sounds were real good, but a few had obvious noise in them.
    I have the fact sheet at home if anyone's interested. I think the
    software runs around $200-300, which really isn't too bad for turning
    your sampler into a wave-table synth. Also, I've heard that Ensoniq
    was so impressed that they're now including the package with the
    new Mirages.
    
    Rob

    > A friend with a Mirage got a demo disk of this (called Sound Designer
    > I think?). It totally replaces the OS on the Mirage and turns it into
    > an ESQ-like synth with replaceable/sampled waves.
    
    Now THAT is neat.  I've often wished they'd come out with a
    conventional synth (with full synth architecture) with sampled wave 
    forms and I had hoped it would be the ESQ-1 architecture.
    
    Roland seems to be heading in this direction too.  My understanding
    is that there are a few programs coming out for the S-550 to make
    it do non-sampler type things.  They already have a program that
    turns it into a video version of an MC-500 sequencer.
    
    This is the right direction.
    
    	db

    I've found that the FZ-1 allows a great deal of 'synthesizing' to
    be done a la wavesynth. But beyond that, the esasiest way to knock
    up great synthesized patches seems to be to hack up samples and
    use them as sort-of oscillators. Then by x-fading, layering and
    the like you can store a huge number of voices, each comprising
    a variously bizarre recosnstruction of a relatively limited set
    of raw samples.
    
    Richard.