January 4, 2009

funny waveshapes

Some people think that a good waveshape is one that looks like what is printed in a textbook. A clean SAW wave is not always the best sounding one. Buchla 259s do not address the waves at the main output by the way they look, but rather by their harmonic content. Sure, there's a square and sine output, but the main output is a blend between even and odd harmonics, between hi and low order harmonics and between lo and high timbre. This always seemed, to me, like the better way for a musician to deal with oscillator output. The 258 claims to have a saw and square wave, but I feel that the spirit has always been about even or odd harmonics. They crossfade from sine to saw on the top oscillator and square on the bottom. The waves never look perfect, but they sound so good that the oscillator has a cult following. People clone it left and right. The above 'scope pics are from the prototype of my clone.

Here you will find some scope shots from an actual vintage Buchla 258C:


Looking good. Sounding good. Feeling good.


wavingpalms said...

Boy, compared with the example you linked, your 'saw' looks dead-on!

Get Loose said...

On the 259, there is saw to triangle converter for 'symetry', a sine to square x-fader for 'order'and a waveforder for 'timbre'. Is this correct? These waveshapers send their output to the 'final' signal output.

What is the order of these circuits I wonder? Is the Saw waveform sent to the Saw-Tri(symetry) convertor and then to the Waveforlder(timbre) before it goes to the Sine-Square x-fader? Or is the X-fader seperate functional block? Are the outputs of these cicuits mixed before being output at the 'final' or are they in series?

It just seems like the Wavefolder wouldn't do much to a square wave, and so the x-fader would make more sense being last in the series of waveshapers.

Could you talk some about the 259 and how the waveshapers function?

I'm also curious about the modulation index. I've wondered what exactly the AM modulation does. I'm guessing it is a VCA for the amount of modulation, rather than an amplitude modulation of the 'final' output waveform.

darkness said...

the 259's AM is just that; the resultant waveform from all other modulation and waveshaping is passed through a VCA controlled by the mod oscillator. this is the same on the 261e, but it doesn't use a vactrol for the VCA's cv input so the effect is different (especially at LFO modulation speeds).

modulating the mod index is on patch cord away (and also much fun at LFO speeds).

on an aside, i really like that the 259's LFO will go _super_ slow if you feed 0V into the CV input of the mod oscillator and turn the scaling knob most of the way to the left. I wish the same thing worked on the 259e/261e/281e!

mark verbos said...

The core of a 259 is a triangle, not a saw. That triangle is sent to a sine shaper as well as a circuit that creates the quasi saw and square shapes and sweeps between them. The sine is sent to the wave folder, which has a vca before it. The timbre control crossfades between the sine and the saw/square wave and at the same time as controlling the gain of the vca before the wave folder. The two outputs of that are then crossfaded between via the order control. AM is achieved through a vactrol based balanced modulator after that.

makenoise said...

Nice explanation Mark. Where/ how does the "symmetry" control fit in?


Get Loose said...

>>That triangle is sent to a sine
>>shaper as well as a circuit that >>creates the quasi saw and square >>shapes and sweeps between them.

Is this the 'Symmetry' shaper then?

>>The sine is sent to the wave >>folder, which has a vca before >>it. The timbre control crossfades >>between the sine and the >>saw/square wave and at the same >>time as controlling the gain of >>the vca before the wave folder. >>The two outputs of that are then >>crossfaded between via the order >>control.

So, if I understand correctly, the tri core is sent to the 'Symmetry' shaper which sweeps between sine-tri-saw-square.

Then it goes to the 'Timbre' vca/wavefolder whose output is cross-faded with square-wave via the'Order'shaper.

Then there is a final VCA for A.M. modulation.

Or something like that. Thanks for the info. I'm trying to make a functional clone using Arp saw/square vco cores(because I have a PCB pattern for them).

Don Solaris said...

Which oscilloscope do you use for monitoring? (looks like a portable one)