Hello Edgar and all,
Interesting animated GIF - wonder how you create those :-)
One suggestion: Even though it's incoherent averaging, a curve with
the long-term average spectrum plotted along with the recent
spectrum (graph) sometimes helps to identify very weak peaks,
because in the average of all amplitude spectra, 'sigma' (the
standard deviation) will gradually drop, the noise in the spectrum
will be ironed out making weak peaks easier to spot. At least, a
'long term average' doesn't hurt. You could reset the averages
(accumulators) once per day or so.
A very long FFT, covering an entire day, may be even more useful
(coherent, more 'FFT gain' than with incoherent averaging). But in
the recent version of SL there is no such feature. Possibly easier
to do this in the post-processing, using the decimated I/Q wave
files for input, and an FFT with arbitrary length (SL's FFT is the
normal implementation, which needs a power of two for the FFT
length).
All the best,
Wolf .
Am 08.12.2015 um 21:54 schrieb edgar:
Hi Joe,
Result from last night.
https://dl.dropboxusercontent.com/u/101251787/VO1NA/VO1NA.gif
Some adjustment in Spectrum Lab settings still required.
Regards, Edgar
Moonah, Tasmania.
|
|
