On Monday 06 June 2005 10:04, Alfons Adriaensen wrote:

> The most difficult situation is for n nearly equal to one.
>
> But even this is not really correct if you want precision.
>
> Consider an Fs/4 sine with samples -0.9 -0.9 +0.9 +0.9 -0.9 -0.9 +0.9 etc.
> Now what is the (peak) amplitude of this sine ? Not 0.9, but something like
> 1.27. If you send this signal through a unity gain filter that produces 45
> degrees of phase shift, you will get samples at +/- 1.27, and your DAC will
> clip.
>
> So even a simle digital peak meter requires interpolation of peaks to
> be accurate. I don't know any in the linux world that do this.

Good points.
What kind of interpolation is required to visualize the DAC output of a 
sampled waveform?


Gerard


Reply via email to