Korg Poly800/EX800 Users

Hi all, I am new to the group, am expecting my newly bought Poly-800 to arrive, and considering to buy HAWK-800 upgrade.

I've read through the list of HAWK-800 features, and would like to know how this Random Sample & Hold thing works. Any samples to illustrate its usages? Thanks.

--- In korgpolyex@yahoogroups.com, "tim.tashpulatov" <tim.tashpulatov@...> wrote:

>
> Hi all, I am new to the group, am expecting my newly bought Poly-800 to arrive, and considering to buy HAWK-800 upgrade.
>
> I've read through the list of HAWK-800 features, and would like to know how this Random Sample & Hold thing works. Any samples to illustrate its usages? Thanks.
>

It makes random beepy noises. "Sample and Hold" refers to the act of sampling an incoming voltage at a particular instant, and then holding it for a certain length of time. Take a look here:

http://www.super-freq.com/?p=89

In the first circuit diagram the 555 timer is an LFO that generates a very narrow pulse. This switches on the FET in the middle (marked G/D/S) which conducts for an instant. The capacitor connected to "D" charges up to whatever voltage is coming in from the noise source at the top almost instantly. When the FET switches off, the capacitor stays charged. The other FET forms a high-impedance buffer so that the cap doesn't discharge quickly.

The upshot of this is that you have a random voltage from the white noise source, but you hold the output at a particular level for a time depending on the speed of the LFO. These days, of course, it's all done in software - which is how the HAWK-800 can do it ;-)

Mike, thanks for pointing this out. Can the same random effect be applied to notes pitch?

--- In korgpolyex@yahoogroups.com, Michael Hawkins <korgpolyex800@...> wrote:
>
> Hi there,
>
> Check out this video http://www.youtube.com/watch?v=s36VDjadaTc
>
> At 1:05 there is a random sample and hold resonant filter sweep (coming out of the left.
>
> Mike.

--- In korgpolyex@yahoogroups.com, "tim.tashpulatov" <tim.tashpulatov@ ...> wrote:
>
> Hi all, I am new to the group, am expecting my newly bought Poly-800 to arrive, and considering to buy HAWK-800 upgrade.
>
> I've read through the list of HAWK-800 features, and would like to know how this Random Sample & Hold thing works. Any samples to illustrate its usages? Thanks.
>

It makes random beepy noises. "Sample and Hold" refers to the act of sampling an incoming voltage at a particular instant, and then holding it for a certain length of time. Take a look here:

http://www.super- freq.com/ ?p=89

In the first circuit diagram the 555 timer is an LFO that generates a very narrow pulse. This switches on the FET in the middle (marked G/D/S) which conducts for an instant. The capacitor connected to "D" charges up to whatever voltage is coming in from the noise source at the top almost instantly. When the FET switches off, the capacitor stays charged. The other FET forms a high-impedance buffer so that the cap doesn't discharge quickly.

The upshot of this is that you have a random voltage from the white noise source, but you hold the output at a particular level for a time depending on the speed of the LFO. These days, of course, it's all done in software - which is how the HAWK-800 can do it ;-)

--- In korgpolyex@yahoogroups.com, Electrohead <electrohead2000@...> wrote:

>
> So it's sampling from white noise? That's a great idea. Very random.
> Most synths sample a sine wave for S&H.

Mmm, none that I've seen do ;-)

What you'd get is a very aliased sinewave depending on the frequency of te sine and the LFO rate. It's a useful if strange effect, good for cyclic patterns. Feeding a squarewave in will give you a variable pulse width pulse output, and a sawtooth wave will give you a "staircase" wave. A mixture of sinewaves (think Hammond organ waves) would give you very complex patterns.

Right, must go and build an outboard S&H now...

Gordon MM0YEQ

On Monday 04 January 2010, Michael Hawkins wrote:

> A high frequency sine wave sampled at low frequency will resemble a
> random number generator.

No matter how you sample a sinewave you'll always end up with a
sinewave. In your case it will be an alias of the original sinewave
since you are undersampling.


Achim.
--
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk]>+

SD adaptation for Waldorf Blofeld V1.04R2:
http://Synth.Stromeko.net/Downloads.html#WaldorfSDada

So do you use/did Korg use this method to generate pseudo-random noise?

Happy New Year guys, incidentally. All the best for you all for 2010!

On Monday 04 January 2010, Michael Hawkins wrote:

> An alias that was the result of an infinitely low sampling rate
> would "resemble" noise.
>
> No?

No. It would resemble DC. :-)

If you're curious why this happens, the "natural" reconstructed signal
in this case lies between DC and the sampling frequency, which is
infinitesimally close to DC. You could reconstruct the signal alias
in any of the image bands as well, but then it would again be a
perfect sine since the bandwidth of the reconstruction filter has to
be infinitesimally small.

There have been several attempts at creating "true random" signals by
sampling from an ensemble of oscillators. None of them work
particularly well unless they're followed up by considerable digital
processing.


Achim.
--
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk]>+

Wavetables for the Terratec KOMPLEXER:
http://Synth.Stromeko.net/Downloads.html#KomplexerWaves

--- In korgpolyex@yahoogroups.com, ASSI <Stromeko@...> wrote:

> There have been several attempts at creating "true random" signals by
> sampling from an ensemble of oscillators. None of them work
> particularly well unless they're followed up by considerable digital
> processing.

Although this approach is used for the hihat/cymbal generator in a lot of analogue drum machines - look at the circuit for the TR606 for example. A single Schmitt trigger is wired up as a rough square-ish wave oscillator and mixed the others to form a complex metallic noise. Two of them are tapped off for the classic 808 cowbell.

a seemingly simple problem that isn't simple at all...

http://en.wikipedia.org/wiki/Pseudorandom_number_generator

there's random, and then there's *RANDOM*. even the definition of random
can be debated among physicists, philosophers, engineers, cryptographers,
mathematicians, computer scientists...

suffice it to say that a lookup table of 256 "randomly shuffled" values
(as currently used in the HAWK-800) should be "random enough" for *most*
musical applications.


--
...atom

________________________
http://atom.smasher.org/
762A 3B98 A3C3 96C9 C6B7 582A B88D 52E4 D9F5 7808
-------------------------------------------------

"There is a theory which states that if ever anyone discovers
exactly what the Universe is for and why it is here, it will
instantly disappear and be replaced by something even more
bizarrely inexplicable. There is another theory which states
that this has already happened."
-- Douglas Adams

> suffice it to say that a lookup table of 256 "randomly shuffled" values
> (as currently used in the HAWK-800) should be "random enough" for *most*
> musical applications.

It also leads to the possibility that you could have a start and a loop parameter to restrict the range within the lookup table that the "random" S&H uses. Couple that with MIDI-synced LFOs and you've got the basis of an interesting little rhythmic pattern generator.

Gordon MM0YEQ

That's a really good point. I have an original (G1) Nord Modular, and
one of the coolest features is a bank of 128 x128 'random' patterns.
In fact, they're not random at all, but there are a lot of them. The
cool thing is you can assign a MIDI controller to pick patterns from a
bank, and control any parameter you like from the resulting output. Of
course, when you save the patch, the patterns get saved too.

a|x

On 4 Jan 2010, at 23:19, gordonjcp wrote:

>
>> suffice it to say that a lookup table of 256 "randomly shuffled"
>> values
>> (as currently used in the HAWK-800) should be "random enough" for
>> *most*
>> musical applications.
>
> It also leads to the possibility that you could have a start and a
> loop parameter to restrict the range within the lookup table that
> the "random" S&H uses. Couple that with MIDI-synced LFOs and you've
> got the basis of an interesting little rhythmic pattern generator.
>
> Gordon MM0YEQ
>
>
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>

> suffice it to say that a lookup table of 256 "randomly shuffled" values
> (as currently used in the HAWK-800) should be "random enough" for *most*
> musical applications.

It also leads to the possibility that you could have a start and a loop parameter to restrict the range within the lookup table that the "random" S&H uses. Couple that with MIDI-synced LFOs and you've got the basis of an interesting little rhythmic pattern generator.

Gordon MM0YEQ

I would love to hear this too...

I've added it into the HAWK-800 "to do list".

And just so as you know how much I appreciate this list too, I want to send out a big thanks to all of you (both HAWK customers and *potential* HAWK customers!).

Without this list, I would not have come up with some of the great features that went into the HAWK. And in fact, the AtomaHawk would NEVER have existed at all (stay tuned for a follow on email about the FM-800 mod').

And it is this kind of discussion that feeds my own development of the HAWK. For example, I have been toying with the MIDI LFO clock sync idea for some time and I haven't been happy with the way the current LFO2 clock MIDI sync operates because it doesn't cycle in perfect time with the MIDI clock (except at the slow speeds it does). But this little exchange got me thinking about that and *bingo!* I struck upon a solution.

So I am going to add another "to do" to the list that will alter the LFO rate values according to whether MIDI sync is enabled or not. When enabled, the clock rate will use rates that ensure that LFO2 stays in cyclic sync with MIDI. I'll put a better explanation together once I implement it.

The random sample and hold controls as suggested can be done but it will take some time because the LFO's are completely software generated and the number of CPU cycles used to drive all of these new controls taxes the Poly's performance. So it's on the list but implementation may take a bit of time.

Cheers,

Mike

--- In korgpolyex@yahoogroups.com, Electrohead <electrohead2000@...> wrote:
>
> Wow! Looping a set of S&H results..
> That's another really great idea!!! I'd love to hear that!!
> I'm on Yahoo groups for every vintage/classic synth I own and this one is by far the best.
> I learn more here in one day than in a year on most of the other groups.
> Bravo to all of you who put in the time and effort to make this such a success!
>
>
>
>
> On Jan 4, 2010, at 6:19 PM, "gordonjcp" <gordon@...> wrote:
>
>
> > suffice it to say that a lookup table of 256 "randomly shuffled" values
> > (as currently used in the HAWK-800) should be "random enough" for *most*
> > musical applications.
>
> It also leads to the possibility that you could have a start and a loop parameter to restrict the range within the lookup table that the "random" S&H uses. Couple that with MIDI-synced LFOs and you've got the basis of an interesting little rhythmic pattern generator.
>
> Gordon MM0YEQ
>