[pd-ot] electric circuits in software

Miguel Ramos pd at anjos.strangled.net
Sat Apr 29 10:29:50 CEST 2006


Sex, 2006-04-28 às 09:19 +0800, Chris McCormick escreveu:
> On Thu, Apr 27, 2006 at 02:27:29PM +0100, Miguel Ramos wrote:
> > Never forget that if you put an analog square wave on one end of a long
> > cable, on the other end what you will receive won't be a square wave,
> > but a rounded version of it.
> 
> Yeah, and the rounding as a result of an analogue signal coming through
> a cable will probably be different to the rounding generated by a dac
> from a digital signal. I guess this is my point. There is going to be a
> difference in the wave forms, but it comes down to whether we can hear a
> generic difference in a waveform generated from analogue gear, or from
> digital gear.

Yes, it *may* be different, if nothing else is working as a low pass
filter. Each subsequent stage of an amplifier also has a frequency
response, and will also work as a low pass or band pass filter.
But the square wave and saw wave at 22.05kHz you mentioned, only differ
at frequencies above 44.1kHz! It's possible to accept that 22.05kHz is
an audible frequency, for the sake of argument, but above 44.1kHz??
And if that would be audible, do you think there's any chance this
difference of >=44.1kHz frequencies will pass on to the speakers and
from then to spl?

> > Moreover, a squarewave is a digital concept mainly. There's really no
> > way to do it in analog.
> 
> And all digital signals are actually analogue on the wire.

Well, but there's a big difference, don't you see? A signal processing
chain has many elements. With a digital square or saw wave, your
processing is working on an as-close-as-possible approximation of a
square or saw wave. When converted to analog, the low pass filter does
the best job possible of approximating the audible part of the timbre of
a square or saw wave to analog.
In analog, you start working with a crude approximation of a square or
saw wave, full of peaks; as the signal goes through more and more
processing elements, the crudeness increases...

Miguel





More information about the PD-ot mailing list