Re: [Discuss-gnuradio] FFT normalization issue: noise aliasing

[Jean-Michel FRIEDT]

> > Shouldn't my aliasing experiment raise the noise floor rather than keep it
> > constant ?
>
> By taking only every 10th sample, you also reduce the sample rate
> by a factor of 10. So you have the same power in 1/10th of the
> original bandwidth. In other words, noise power per Hz has
> increased by a factor of 10. This noise density (compared to
> signal power) is what matters in e.g. demodulation.

which brings me back to my question indeed: is the FFT sink displaying
dB or dB/Hz (or dB/sqrt(Hz) for voltages) ? In the latter case I agree
with your statement.

Marcus: I see that I will have no other choice than to start learning the
Qt interface ... I was so used to the WX one that I never bothered. I just
checked, the result is the same on both GUI so they are consistent.

Thanks, JM

--
JM Friedt, FEMTO-ST Time & Frequency/SENSeOR, 26 rue de l'Epitaphe,  
25000 Besancon, France

----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.