[Discuss-gnuradio] Re: factors for gr.nlog10_ff [was: output of gr.argmax & gr.max]

[Patrick Strasser]

Tim Meehan wrote am 2008-03-12 11:28:

> BTW I sent a message to this list the other day questioning the 
> "-20*math.log10(self.fft_size)..."
> in the line below (taken from your code above).  I argue it should be 
> "-10*math.log10(self.fft_size)..."
> [from Brook Lin]  
> self.fft_size,-20*math.log10(self.fft_size)-10*math.log10(power/self.fft_size))

It is:

for tap in mywindow:
    power += tap*tap
self.log = gr.nlog10_ff(20,
                        self.fft_size,
                        -20*math.log10(self.fft_size)
                          -10*math.log10(power/self.fft_size)
                        )

The variable power is a square of amplitudes, which is a power value, so 
-10*math.log10(power/self.fft_size) is correct.

Patrick

Please do not top quote/full quote. See 
http://gnuradio.org/trac/wiki/MailingLists#Guidelinesforposting
--
Engineers motto: cheap, good, fast: choose any two
Patrick Strasser <patrick dot strasser at tugraz dot at>
Student of Telematik, Techn. University Graz, Austria