The traditional synth signal path, tweaked ...

[girevik]

I'm guessing somebody coughed up a huge tax bill recently. Actually, several - judging by the grouchiness of several individuals.

[mildbill]

Keyboards and synthesizers are two distinct things. You can remove the sound generating synthesizer, and still have the physical keyboard left in front of you.
Use a modular synth without a keyboard.
Or use a keyboard strictly as a controller, with no sound generation capabilities.

But they can also be combined. And when they are combined, the normal, usual, or traditional signal path is vco>vcf>vca (with modulation possibilities).


EDIT: I think even people who work exclusively with modulars probably start patching with vco>vca, for the simple fact that it can be heard.

[robotsquid]

Leaving aside all the arguing for a moment, if you used a VCO->VCA->VCF signal path with a self oscillating filter you wouldn't be able to gate its output when it was resonating, which would be a problem sometimes.

[setAI]

I'm guessing somebody coughed up a huge tax bill recently. Actually, several - judging by the grouchiness of several individuals.

 

 

 

my wife and I owe $9000 - just started making $1500 dollar a month payments- so- yeah- you are teh psychic

[Tusks]

The real reason is historical. Minimoog was the first synth with a prepatched architecture. The crowd went wild. Everybody copied. Simple as that.

 

 

This is how I view it also. There were some analogies to acoustic instruments in Bob's mind even while working the modulars (the dynamic filter being a key development in that process) from what I have read.

 

Parallel resonant filters or a filter bank at the last stage ... is probably the closest analogy to acoustic. Not that that's better than a Freqbox with no filters. But we spend so much time going on an on about how something needs to be warm and organic ... and we choose to have only one ADSR shaping the filter response when we could have two. (It occurs to me a pre-filter mixer with ADSR/motion control would be great. Especially if you could do it vector fashion like the Wavestation.)

 

Jerry

[girevik]

my wife and I owe $9000 - just started making $1500 dollar a month payments- so- yeah- you are teh psychic

 

 

I merely spoke in jest. Sorry to hear that, man. Sounds brutal.

[raffor]

[urbanscallywag]

Leaving aside all the arguing for a moment, if you used a VCO->VCA->VCF signal path with a self oscillating filter you wouldn't be able to gate its output when it was resonating, which would be a problem sometimes.

Finally some real logic in this thread!

[Metrosonus]

I dunno. my head hurts and i think bob moog invented the internet?

[ElectricPuppy]

The internet is a series of filters with the VCA at the front.

[wetwareinterface]

The internet is a series of filters with the VCA at the front.

 

 

more like a large batch of noise generators with no filtering at all and random value generators running the individual vca's for each one.

 

but anyway the whole resonant vcf after the vca is quite possible as long as your vcf isn't self resonating. if the filter won't acheive self resonance you can do a vca before it and the gating will come from the audio signal tapering at the vcf's input stage. besides self resonance isn't necesary at all to emulate anything other than a sine wave and if you have a sine wave osc becomes redundant and useless.

 

and no there isn't any technical reason why you can't put the vca in front of a vcf as far as the audio signal is concerned. the vca is not an amplifier but rather an attenuator. it takes the raw max signal and SUBTRACTS from it so therefore the filter wouldn't be overdriven at all. they call it subtractive synthesis for a reason other than just subtracting from the frequency they also subtract from the amplitude.

 

and the esq-1 dca's aren't digital, they are analog vca's controlled by a digital controller signal. the signal path in the esq-1 is very hybrid. the controller signals are digital and the osc is digital. at the osc output it all goes analog as far as the signal path of the audio chain. the filters and dca 1-3 and final dca 4 are all analog signal paths. they just take a digital control feed for modulation. which is why you can change patches on an esq-1 without interupting the currently played notes, until the next key strike the osc won't resend a new signal and the onboard processor won't send new control data for the digital pathway, the processor just accepts the next digital control parameters to send on key strike and stores them in the buffer.

[Raymar]

Isn't putting the VCA before the filter simular to plugging a synth, guitar, or any electronically produced or enhanced sound into an effect? I have a Roland U20(oh crap, its not even analog!!!) that has no filters but it does have built in effects after its amp(I believe), and I could run it out through something - like say a Moogerfooger lowpass filter.

To me if you can tweek it to sound unlike a musical instrument, or a hybrid musical instrument, its a synth, maybe somewhat limited but a synth still.
The point is that's its alot of fun!

Steve

[mildbill]

... if you used a VCO->VCA->VCF signal path with a self oscillating filter you wouldn't be able to gate its output when it was resonating, ..

 

 

 

 

Why not?

 

 

Just use an envelope.

[urbanscallywag]

Why not? Just use an envelope.

An envelope to control...?

[wetwareinterface]

An envelope to control...?

 

the output level of the vca before the vcf.

 

feed a resonant vcf no signal and you end up with? that's right no signal. if you feed a self oscillating vcf no signal depending on the design of the vcf you end up with no signal or a sine wave. simply limit the vcf to not self oscillate and you have no problems at all cutting it's volume down with a vca in front of it. run your audio through a filter input with the onboard vca's envelope set to wide open and it's the same thing as a vca before a vcf that doesn't self oscillate.

[urbanscallywag]

Avoiding the filter self oscillating is not exactly fixing the problem.

[mildbill]

An envelope to control...?

 

 

 

I don't know. Anyone wanna loan me a modular so I can try it. I'll post the results.

 

 

If you set resonance high, give the filter some input, and then totally disconnect the oscillator and amp, will the filter continue to make sound indefinitely?

[wetwareinterface]

I don't know. Anyone wanna loan me a modular so I can try it. I'll post the results. If you set resonance high, give the filter some input, and then totally disconnect the oscillator and amp, will the filter continue to make sound indefinitely?

 

 

 

depends on the filter design some shut down with not enough input some stay running at self oscillation.

[wetwareinterface]

Avoiding the filter self oscillating is not exactly fixing the problem.

why isn't it?
you don't need self oscillation, in point of fact it doesn't do anything a sine oscillator can't do better, like be pitch modulated and sync'd and used in fm as a carrier or modulator etc...

self oscillation is something you grow up with in a synth and think you need. take it away and replace with an osc that has a sine wave and you quickly learn that it's a useless feature in a filter that's only real purpose is to blow your speaker cones by accident.

there are several synths that don't have self oscillating filters but that get real close to self oscillation that don't suffer for it. having it go self oscillating is just an overkill resonance level. it doesn't need to be that high of a resonance level to get it to do anything you need sound design wise as long as it has a sine wave.

[mildbill]

OK - so an oscillating filter makes sound. That sound has an amplitude, or you wouldn't be able to hear it, right?

So, if you put an envelope on its output, it should control its volume even though there's no vca in the circuit.


I don't know. Anyone want to try it?

[wetwareinterface]

OK - so an oscillating filter makes sound. That sound has an amplitude, or you wouldn't be able to hear it, right? So, if you put an envelope on its output, it should control its volume even though there's no vca in the circuit. I don't know. Anyone want to try it?

the filter is typically only able to be modulated in either its frequency setting or resonance amount. there are damn few filters that have an amplitude modulation input and those that do are simply a filter with an amp stage tacked on the end. so no it won't work.

however if you put the envelope on the vca that is placed in front of the filter and you don't crank the resonance to self oscillation you can control it's volume output just nicely.

[mildbill]

the filter is typically only able to be modulated in either its frequency setting or resonance amount. ....

Well, there you have it. Patch an envelope into the resonance control.

Reduce the resonance and the sound diminishes.


What was the question?

[urbanscallywag]

Not many synths have a sine oscillator either. I wonder if THD is lower for a filter's sine than triangle derived sine. I imagine the tracking is going to be better on most oscillators over a filter though.

 

I disagree self oscillation can be replaced by an oscillator. For example the Korg MS-10/20 have an insane distorted self oscillation (from their Sallen-Key design I think).

[mate_stubb]

You can't just patch an envelope to the output of a filter or oscillator. If you want to control the volume of the output, the envelope must control a VCA.

 

There is a very good reason one would prefer an oscillating filter to the sine output of a VCO - waveform purity. MANY VCOs have impure sine waves generated from conversion circuits, which display audible harmonics. An oscillating filter is purer, and for FM stuff I find it superior.

[wetwareinterface]

i didn't specifically say a vco for a sine wave, i said an oscillator, that can be either a pcm wavetable, or a dco or a vco. 2 of which offer a cleaner sine vs. the self oscillating filter, which by the way suffers from the same charge discharge cycle impurities you attributed to a vco.

the main thing you get with a sine osc over a self oscillating filter is ... filtering of the sine and or other waves and adding resonance. hard to do when you've tied up the filter to generate a sine wave and drowned out any osc waveforms feeding the filter and done it at the final stages of the sound. and there are literally more synth models with sine waves in the osc section than without so that argument is a lame duck urbanscalywag. granted in older analog hardware there are fewer choices in sine wave oscillators, but speaking from the perspective of today... it's the other way around.

the whole reason basically to desire a self oscillating filter is to be assured it will have a high enough resonance peak to handle anything you want to throw at it. however a good design that doesn't self oscillate but gets right up there next to that territory is just as good synth patch creation wise. case in point - esq-1 won't self oscillate. but does have sine waves so you don't need to tie up the filter to get a sine wave to do bass drums or sub bass patches.
another case in point
novation nova/super/novaII et al has a sine wave with the soften osc function applied to a square wave, leaving the filter for filtering duties rather than as a simple sine wave generator.
etc...

now for a synth to get beyond the filter before the amp (attenuator) layout you need a modular or software/dsp based synth or a simple older design that lacked features for the majority of old school designs with the reverse of the common practice. but there are advantages of doing it in reverse, namely more accurate strike to resonant body emulation like a plucked or strumed guitar sound or hammered sounds. as long as the envelope is accompanied by an lfo when modulating the amp that comes before the filter you get a more accurate model of guitar/piano/drum/etc.. type instruments using plain old subtractive synthesis.