Re: Re:Re: Looperlative LP1 - sample rate

["gareth.whitcock" <gareth.whitcock@ntlworld.com>]

Hi Andy - all is well in Wales - you should pop over sometime!

Sorry to harp on but the dfference is that a d to a converter may attrempt 
to emulate a sine based on A NUMBER
of squares not just one - hence my statement. I'm sure we agree but there 
seemed to be a lot of mis- information happening around here -
uncharacteristic for this list it has to be said!!

Gareth

>
>>  No sampler can "guess" a sine wave from a square.
>>How would it know that you wanted a sine wave? Think of sines and 
>squares 
>>as just different shapes.
>>
>>Gareth
>
> hi Gareth,
> well believe it or not that's exactly what a good D/A converter does,
> (and what a bad one tries to do)
> The converter assumes that the signal it is trying to re-create is
> band limited, i.e. that it is describable as a mixture of sin waves all 
>of
> which are below the Nyquist frequency.
> In essence, the digital output is passed through a lo-pass filter with
> a cut-off at around 20kHz.
> As you're no doubt aware, all filters "ring" somewhat, they have
> something of a tendency to oscillate at their resonant frequency.
> The converter can then guess at the waveform at high frequencies,
> even though it's only represented by a few samples.
> By varying the Q of the filter, the amount of "ring" is controllable,
> which is why some converters sound dull, and some sound
> bright but unnatural.
>
> hope all is well in Wales
>
> andy butler
>
>