All the great filter board options available now makes me kind of wish for an analog oscillator. Don’t get me wrong, the two digital ones that we have are extremely useful and fun, but they have so much character on their own that I sometimes feel that the nuances of an interesting filter can get lost, or perhaps superfluous. As we have it now, I can get lost in wavetables and modulations, and easily forget about the filter.
It would be quite possible, I’m guessing, to make an analog oscillator board to replace the functions of the digital board. It could maybe even use the same layout of knobs, switches, and screen, to make use of the existing cases. Or maybe a simplified layout. It could be two simple saw/pulse oscillators, or maybe a single more advanced oscillator with a sub, and maybe a noise generator.
Or would this be too boring? Do we already have too many machines on the market to fill that sonic niche?
If it was that simple it would have been done
You would still need host processor so you can’t get rid of that. Then you need to find analog oscillator chips or use bulky components to build it up. I think an analog oscillator board would be as big if not bigger than the analog filter boards.
You would need three oscillators to match the current sound architecture, 2 + 1 sub.
You would only really get simple waveforms too, saw, square, sine, tri.
An analog oscillator with good temperature stability and a waveshaper has roughly the same space / part count as a filter + VCA (We’re talking here about 5 DIP14/16 chips surrounded by parts). There won’t be much space to house that + MIDI interface + Controls on a control board. Not to mention that proper DACs would have to be used for accurate pitch reproduction.
A nice idea but there’s plenty of analog stuff out there now. It’s all very fixed, you can’t add new oscillators or tweak them much. Plus if you’re using the same IC chips as other units then you’re not really going to sound all that much different.
Yeah, I think it’s a nice idea, but, but there are other synths out there with analogue oscillators, and personally, I prefer the greater flexibility of the digital oscillators. There’s nothing stopping you only using the basic synth waveforms.
But I would keep the controller board and make a “filter board” with three “discrete” VCOs, a mixer, a VCA and a VCF. I get that it has to be bigger but paired with at Shruthi XT enclosure that would possibly not be a problem. What you get is a DIY synth with memories and VCOs, and those are far between…
The actual Digital board isn’t capable of controlling 3 VCOs so you’ll need to redesign this one as well…
Your “filter board” would then be 2.5 the size of the control board. Those oscillators require space and 2 CVs each (pitch, at least 12-bits or 2x8-bits if you don’t want crappy stepping when pitch-bending ; and PWM). This also means that an extra MCU would be required to receive data from the main MCU and for talking to DACs.
It definitely is a new product almost by the time you’ve done all of that.
Plus if you use discrete VCO VCA chips you can’t really do much in the way of different waves. You would have to mangle them up later down the signal path.
Then there’s the issue of controls and if you bother with modularity, with a synth that is a circuit you have a fixed signal path which means inflexibility. Having patch leads and connectors means you can re-route, but then for complex modulation you need a hell of a lot of wires and such patching can’t be “stored” in the patch (easily).
Even Vince Clarke says he uses soft synths for some sounds because to do complex modulation with modular synths requires a mile of cable.
imo it’s much easier for the user to decide for an existing VCO, Waveshaper, VCF, VCA design and then add what you need in terms of modulation sources and/or, if you like it programmable build a midibox CV. besides Olivier’s approach to the digital FX there are existing designs for filters that work with the chips, with some differences. also there are differences in what you can do with this or that oscillator. but i guess the main reason for wishing something like this for the Shruthi would be the programming / MIDI access to the waveform generation, which i guess is not that easy to do, at least not with just one routine.
When you do signal generation (oscillators + mixing) in the analog domain, it is quite straightforward to stick pots and switches everywhere in the circuit to control things. Want a waveform selector? Bring the signals of the waveshapers to a rotary selector and that’s it. Want a waveform mixer? Bring 3 pots or sliders to an op-amp. Want a glissando control? Add a pot and a cap at the output of the pitch DAC.
Now, if you want to have patch recall or MIDI control over this, you’ll have to replace the waveform selector by a 405x ; the waveform mixer by an array of VCAs ; and the RC filter by a 1-pole VCF.
This is pretty much why all analog synths have a “knob per function” interface. It’s the easy way to do it. Once you want MIDI / CV / patch memory control over all this mess everything gets 2 or 3 times more complex.
I guess that’s why the Minibrute is so much cheaper than any equivalent- no MIDI parameter control or patch memories.
If it were possible to do, it would be neat to have a non-midi, no patch memory, non-digital, 1 knob per function analog oscillator board that would work with the existing analog filters. I would buy it…
Or maybe just midi for keyboard/gate like the Minibrute (having MIDI is awfully nice)?
The ShruthiBrute board?
A relatively easy way to do a somewhat analog VCO would be to use the CEM3396 with a MCU-derived clock, but the downside of that is that it would require lots of necro to get hold of those chips - making a workalike isn’t easy either.
The first synths with “patch” memory were massively stripped down. The memory moog etc.
There’s exceptions, the Prophet 5 and so on. But in 2012 (34 years later) do you really want to keep regurgitating 70s sounds? while bouncing on your space hopper and wearing flared trousers.
It is time to move on
I have a CEM3396. Free to a good home, if you can prove you’re going to be able to use it.
Ah ah, I stopped with the 3379, 2044 and 3109, and it’s certainly not to start again with another member of the gang
A wise choice! It would be too complicated to ask people to once again source some rare old chips