[sdiy] Novation peak NCOs

[Tom Wiltshire]

This is why I'd managed to use the 33FJ128GP802 dsPIC's DAC for years without having any trouble, and then as soon as I did a digital delay with it, I found it noisy as hell. Previously I'd done digital oscillators, envelopes, LFOs, and such like - all outputting signals that use the full range of the DAC.
As soon as I needed soft signals, I noticed how high the noise floor was.

Tom

On 23 Apr 2017, at 17:36, Eric Brombaugh <ebrombaugh1 at cox.net> wrote:

> This is an excellent point, and something that is often overlooked when designing digital oscillators. Given that you're likely generating a signal that's near 0dBfs and will almost certainly "suffer" from additional downstream processing, having some crud that's > 60-70dB down is not going to be nearly as much of a problem as it might be if you were working with a more general-purpose audio signal with much greater dynamic range that's going straight to the speakers.
> 
> Eric
> 
> On Apr 23, 2017, at 9:16 AM, Richie Burnett wrote:
> 
>> It's interesting the limitations of using just a single bit that Brian mentioned. All I can think is that the presence of a bit of low-level noise, limit cycles and distortion are not really inhibiting for the application of generating hot raw oscillator signals that will subsequently be put through noisy analogue filters and VCAs anyway!?!? 
> 
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy