Conference napalm::commusic_v1

    The following *conjectures* are based on experience with similar
    vintage Roland synths.
    
    	hold - holds whatever notes are playing while in effect.  Does it
    	have a light associated with it?  Maybe it's a push on, push off
    	type switch?  Try playing the keyboard while it's held down.  Or
    	press it, then play the keyboard.  Etc.
    
    	up, down, up & down - these must be arpeggiator modes. 
    	Arpeggiators have just about disappeared from synths.  My guess is
    	that some combination of playing a chord on the keyboard and using
    	one of these buttons will cause the chord to be arpeggiated the
    	specified way in sync with the internal or external clock.
    
    	transpose - just what it says.  Probably, transposes from C to
    	whatever note you play while the button is held down.  From then
    	on all incoming notes (or notes played from the keyboard?) will
    	be transposed that much.
    
    My guess is the clk-in just controls the arpeggiator.  Unless there's
    an onboard sequencer?  If there's no MIDI in, it's not likely you can
    control the unit from an external sequencer.  The clk signal is
    probably some Roland proprietary convention (probably their old "DIN
    sync" used by the now defunct sequencers and drum machines like the
    MSQ-100, MSQ-700, TR-606, TR-808. TR-909, TR-707/727).
    
    Hope this helps,
    
    len.
     

    Thanks for -1, and the mails I've had from various people.  I now have
    it sussed.  I now fancy another analogue synth of some sort.  The guy
    where I bought this was trying to sell me a Juno 60.
    
    
    matty

    Can anyone out there explain the synthesis on this machine to me?
    
    Obviously I have stuble across sounds I like, and can get back to most
    of them, but I don't fully understand some of it;
    
    eg.  Saw tooth, Square wave, Random and Noise.  I have a dial that
    allows me select any on these 4 wave types.  I have figured out that
    Random randomizes some parameter or other in relation to the Modulation
    setting on the VCF, and that Noise adds background noise.
    
    But what parameter is being randomised?
    
    What's the point of all that background noise?  There must be some use.
    
    What do I need to know about Saw tooths and Square waves?
    
    I also have a modulation slider for the VCA (I think it's the VCA). 
    This is great for muddying up sounds if used up to around 15%, but then
    it become useless and just messes up the pitch.  Is there a use for
    this?  On my PSS680 (Yamaha FM) the modulation parameter modulates
    between the 2 ocillators, but the SH101 seems to modulate pitch.
    
    Help!!!
    
    
    matty

    Saw tooth and square waves are rich in frequency content.  You might
    take a look at how you mathematically analyze these waveforms from
    primary (siunsoidal) frequencies using a Fourier series.  What you'll
    discover are interesting patterns that indicate something of what these
    things are made of.  By filtering, you can selectively remove certain
    frequencies to get waveforms that resemble, for example, acoustic
    instruments.  For example, a flute has a frequency domain that is close
    to what a triangle wave looks like.  A trombone or trumpet has a waveform 
    that basically looks like a sine wave whose amplitude is modulated by a
    saw tooth wave.  Powerful stuff.
    
    Steve
         

The dial you are referring to on the SH101 is to choose the shape of the LFO.
The LFO (Low Frequency Oscillator) is used to modulate other parts of the
sound.  For instance, if you set the dial to a triangle wave, adjust the rate to
about the middle of its range, and barely inch up the modulation slider on the
VCO section, you'll begin to hear something akin to vibrato.  Likewise, if you
inched up the modulation slider in the VCF section, you'd get a wah-wah effect.
What you're doing is taking an oscillator that goes very slowly (at least in 
comparison to the main oscillators that produce sound (i.e., VCO) and applying
that shape of that LFO to the pitch of the VCO.  Or in the case of the VCF,
(which is a filter), you're applying the shape of the LFO to the filter and
creating the same effect as if you had a guitarists wah pedal and you were 
moving it up and down.  It would be the same thing as if you could hook a motor
up to a tone knob and had it raise and lower the tone setting in perfect time.

The other settings on the LFO shape knob allow for different kinds of effects.
You have to experiment with them to find what's useful.  Some suggestions:

First of all - pull down all modulation sliders as an initial setting.

	Set the LFO shape to random, LFO rate to medium.  Now pull up the VCO
	section's mudulation slider and you'll hear it create random pitches 
	(similar to what people used to call computer talk many years ago).

Initialize modulation sliders again.

	Set the LFO shape to noise.  Set the rate fairly high.  If you've 
	created a brass-type patch, you can add a touch of "spit" to the
	sound by inching up the VCF modulation slider.

Initialize modulation sliders again.

	Create a fairly simple sound by setting the envelope attack to 0, decay
	to 0, sustain to full up and release to 0.  Open the filter up most of 
	the way (the filter may be called VCF level).  Now set the LFO shape
	to sawtooth wave, medium rate.  Add in a generous amount of modulation
	to the VCO section and you'll have something akin to a siren.

You really have to play with these things to get used to them.  Try the above
out for size.  Better yet, try to pick up a basic synthesis book from a music 
store or library.  It'll help you out.

	Dan

    re-1,  Thanks.  I tried all that out which helped me understand how
    the VCF and VCO modulation settings work in conjunction with the LFO
    rate.  By having a low LFO rate I discovered I was able to come up with
    some typical fat analogue sounds and good clear bass noises.  I've
    always left the LFO rate Med to High because it also doubles up as the
    clock rate for the sequencer.
    
    More queries (though I'll keep away from synthesis until I've read a
    couple of explanatory books).
    
    I have 4 sockets; cv in (Control Voltage `in' I pressume), gate in, cv
    out and gate out.  It also has clock rate in (or out I can't remember
    which).  What do all these do?
    
    Finally I have sockets for a modulation grip.  What's this?
    
    
    matty

The CV in/out and Gate In/Out have to do with interfacing pre-midi synths.  For
instance, I have the SH101 and a dilapitated Korg MONO/POLY.  The M/P's keyboard
is nigh unto dead.  But I can breathe new life into the unit by connecting the 
CV's and gates, using the SH101 to control the M/P.  The clock in/out is along
the same lines.  If you've got another unit that sends or receives Roland's
particular clock style signals (like the early drum machines (TR707, TR909, 
etc.) and sequencers), you can have them play in sync.  It's quite possible, if
you know the voltage of the signal that you can control in real time (using a
home built circuit - a power source and a momentary switch, kind of like a morse
code transmitter!) the tempo of the sequencer, the arpeggiator and the LFO.  I 
haven't done this, so don't bet your bankroll on it.

My suggestion to get books on the subject did not mean you should not ask 
questions here - by all means, ask away.  This conference is what got me up and 
running.  There are advantages, though, to some of the books as they can
provide some pictures and charts to help out that are difficult on Charecter 
cell terminals.  I wish I could think of some good books to recommend, as I have
seen some (like one that uses the ARP Oddessy for example settings) that are
very good and would really get you rolling.  Unfortunately I never bought any
of them, as I already had a handle on synthesis (again, thanks to this 
conference) by the time I found them.

Go ahead and ask away - to many of us, its our favorite subject!

Dan

    OK then!  I have a `RES' slider in the VCF section.  It seems to put a
    whistle/high pitch into things if I have either the ENV, or FREQ
    sliders up, but not with both.  What does `RES' stand for and do?
    
    thanks
    
    
    
    matty

    It stands for *resonance* of the filter.  It's used to place a narrow
    peak right at the filter's cutoff frequency, thereby accentuating that
    particular frequency.  At very high settings, the filter should go into
    self-oscillation, this is the whistling you're hearing.  
    
    What's happening with the ENV control is this controls how much the EG
    level will *raise* the filter's cutoff frequency; the FREQ control sets
    the starting cutoff frequency.  If both are set to high levels, the
    filter is probably wide open all the time (i.e the cutoff frequency is
    out of the audio range).
    
    Resonant filters are one oif the cornerstones of the analog sound; some
    newer digital synths (SY77, D70) have resonant *digital* filters, but
    most digital synths have no resonance, many have no filters at all!
    
    						Brian

    What he said.  As mentioned elsewhere, the D70 allows filters with
    resonance and it includes an "analog" parameter that causes the pitch
    to vary.  I'm told (Chad mentioned this) that this variance is random, 
    but the amount of this variance can be diddled with.  Exciting stuff,
    these new synths.
    
    Steve

    re-2,
    
    >Some digital synths have no filters.
    
    Geez it's gunna be a comedown with a TG33! ;-)
    
    
    matty

    I wondered what sort of sounds I should be able to reproduce using the
    SH-101.  I can make a synthetic sounding bass drum and a noise that
    sort of goes `tchew' - Kraftwerk used this sound on `The Robots', `Tour
    De France' and most of the `Computer World' LP.  I can do the wind, and
    jets taking off, and police sirens.  And best of all I can make all
    those subsonic bass noises that were used to death on 90% of the UK
    techno records last year.
    
    Are there any *real* instruments I can copy?
    
    I really love my SH-101.  I notice that Deee Lite are on the cover of
    Keyboard magazine with one!  I'm thinking of getting a MIDI retrofit
    (170pounds from Kenton Electronics), or maybe a MC202 as well.
    
    
    matty

    Some ones that are easy to get:
    
    Clarinets, oboes....try the PWM for these.
    
    Strings, brass....try the sawtooth wave.  Obviously one note at a time,
    it won't sound like a full ensemble the way it would on a polysynth.
    
    The list could go on for a while.   Here in the US, a book claaed
    "Synthesist's Guide to Acoustic Instruments" covers this.  It's
    published by Amsco/Music Sales.  I don't know if they have UK
    distribution.  It goes through various instruments, what gives them
    their sound and how to model them on a synth.
    
    						Brian

    Brian, what's PWM?
    
    matty

    >>> Brian, what's PWM?
    
    I'm not Brian but I'll sneak in here anyway.  PWM stands for Pulse Wave
    (or is it Width?) Modulation.  Many analog synths (like my Korg
    PolySix) had a pulse wave as one of the available choices. With a Pulse
    Wave, you usually can vary the width on another control.  This lets you
    vary from a narrow pulse to a square wave and beyond.  (I think a
    square wave is a special case of a pulse wave.)  Thus:
    
         --            __                      ------          ------
        |  |          |  |                    |      |        |      |
        |  |          |  |                    |      |        |      |
      --   -----------    __          ________        ________        ...
        narrow pulse wave                   wider (square) pulse wave
    
    Another frequent choice for the Pulse Wave is to modulate it with a low
    frequency oscillator (LFO) as another means of altering the wave.
    
    Hope this helps (and hope it's right),
    
    - John -

    John's got it right.  It's pulse *width* modulation, as variations in
    timbre are introduced by modulating (i.e., varying in a controlled
    manner) the width (or "duty cycle") of a pulse wave.  A square wave
    is a special case of a pulse wave, where the "on time" equals the
    "off time", or 1/2 the cycle time.  Many classic analog synths allow
    the pulse width to be determined by a constant (a manually set slider
    or knob type control, an LFO, or an envelope.  "Narrower" pulses ("on
    time" a small fraction of the cycle time) contain relatively more high
    frequencies/harmonics, while square waves (the widest pulse wave possible)
    have a much stronger fundamental.  
    
    len.
    

    Can one of you lot just nip over here and show me what you mean? ;-)
    
    How did you all pick up synthesis?  I mean, you couldn't have all been
    taught via Notes.
    
    matty

    > square wave (widest pulse...possible)
    
    ________  ________   ________   _________   _________
            | |       | |        | |         |  |       |
            | |       | |        | |         |  |       |
            __        __         __          ___        
    
    No? How come?
    
    Edd

    To elucidate a bit (even though I don't seem to know squat about
    exchange rates 8-) ... note that I'm coming from an Arp/Oberheim
    background, and am not familiar with the SH101 proper.
    
    A narrow (or "thin") PW wave is useful for Oboe; thicker (wider) pulse
    waves are more suited to clarinet, bassoon, etc.  My best luck with
    Arps was to be frugal on LFO modulation of the width, while using a bit
    of filter to smooth out the sound.  Depending on how extreme your use
    of LFO might be, you can get PW modulated square waves to really come
    alive.  Synthesized sax and fat leads come to mind as applications.
    
    If you've got a couple OSCs to play with and are able to SYNC them
    somehow, you can actually modulate the width of OSC1 by OSC2.  You can
    get some really nasty harmonic sweeps by syncing OSCS, then modulating
    the secondary OSC by either pitch bend or LFO.  The sound will "sweep
    thru harmonics", so to speak.
    
    One interesting filter-sweep type sound on the OB-Xa uses an envelope
    to modulate the filter AND an LFO to modulate both filter and OSC2 in
    sync mode.  Harmonics going all over the place.
    
    If you can combine noise with a filter sweep and have a couple
    envelopes to play with, you can get a nasty-sounding "swept" electronic
    snare.  Use one envelope to control the filter, the other to control
    the "loudness" (amplitude) of the patch.
    
    A few ideas, for what they're worth.
    
+b

    
    re: .17 - matty, you mean you are NOT a weenie?

    re .18 - the wave you have drawn is the inversion of the "narrowest"
    pulse wave, with a phase offset.  It's a well know fact that neither of
    these transformations changes a signal's harmonic content (or a
    listener's perception of it).
    
    BTW, can any Fourier transform (or other time/frequency domain
    transform) weenie out there comment on how inverting the signal (i.e.,
    negating it) does affect the frequency spectrum?   Does something cute
    happen in the imaginary plane?  Does it affect phase instead of
    amplitude?  Are all the amplitudes just negated?  The latter would seem
    to make the most sense.
    
    len.
    

Let's see...a multiply in the time domain yields a convolution in the frequency
domain.  So multiplying a constant -1 in time (i.e. invert) yields convolve
with a delta at frequency 0.  Which sounds like it does not change the
frequency content any.   Note that the frequency content is the vector sum of
real and imaginary (r**2+i**2)**.5, so that would imply that both imaginary
and real probably change.

Burns

    I'm going to build a synth using a `How to build a synth' book.  I
    reckon this will be fun even if the sounds are simple/boring, but most
    of all it should help me understand what's going off.
    
    Has anyone ever done this?  Any suggestions?
    
    Just imagine, in 10 year's time I might be competing against Roland ;-) 
    
    
    matty

	I've never even considered building a digital-based one, but I
	did my first analog synth back in the dark ages. I used circuit
	diagrams from Electronotes (now long gone), and early Craig
	Anderton/Emu circuits. The second one was a combination of
	refined circuits from the first design, but I went for the
	(then) Aries module kits for the VCOs and VCF. Oh, and the keyboard
	and control voltage generator.

	All in all, it was a good exercise. Nowadays there are chips to
	handle most of the things I built on boards. I think you can
	still get Solid State Music chips, maybe one of the chipheads
	in this conference can tell you. Some of the circuits can be
	real simple if you're doing analog (CV) based circuits, like 
	LFOs and EGs. VCOs are touchy though. 

	Digital circuits are another matter entirely. 

	/pjh