[sdiy] Soft ADSR ways?

[Tom Wiltshire]

On 14 Feb 2011, at 12:46, Paul Perry wrote:

> No, I hadn't meant to have an analog output - my idea was to
> (somehow!) do some RC exponentiating outside the micro (though started and sensed by it.)
> I'm all for I2C zipping around.

I see - hence why you'd detect the envelope level with an ADC rather than a comparator - makes sense. Interesting approach.

You could make an interestingly barely-analog-but-totally-analog synth by having a processor that generated DCO reset pulses and used it's comparators for envelope control signals. The resulting synth would have analog audio waveforms (although discrete frequency resolution) and analog envelope signals, but be more or less entirely based on a microprocessor. Probably it'd satisfy neither the analogophiles nor the digital gurus, but it might be fairly cheap and straightforward to build.

> And micros are so cheap now, one should think of them as the 555 of today - and spread them around,
> rather than try to shoehorn everything into a single one.

Yeah, but we never had to do inter-processor comms with the 555! Whilst I agree that they're cheap as chips (or "french fries" if you prefer - and here the price of a packet of chips *is* actually about the same as the price of a PIC) I still reckon it's easier to have things more integrated. This is why having designed PIC-based modules that carry out one function (VCADSR, VCLFO) I then moved on to trying to do the same thing on a dsPIC so that I could include all the envelopes, modulators and mod matrix on one chip. It saves me having to generate lots of analog CV signals only to convert them to digital data in the PIC, or having to communicate digitally with several chips when I could communicate with just one.

Tom