LF: Spectrogram technique]

[Andre' Kesteloot]

Mike Dennison wrote:

> I have used Spectrogram for ages and am currently on v4.2.6. It is
> a truly excellent program.
>
> I thought I had worked out how to optimise it but found an
> interesting dodge at the weekend.
>
> The averaging is normally set at about ten times the expected dot
> length (ie 25-30 for 3s dots). A lower setting gives less s/n ratio,
> and higher will fill in the spaces between dots. But when trying to
> separate DF3LP's signal from a Loran line on exactly the same
> frequency, I found that reducing the averaging to 10 improved
> things. Presumably, the fact that Loran is pulsed was the key here.
> I also found that, although I normally prefer 32 bit resolution, on this
> occasion 16 bit gave a better separation between the two signal
> sources.
>
> Both of these are obvious when you think about it but it is worth
> playing with the settings of Spectrogram to optimise a particular
> signal and noise combination.
>
> Mike, G3XDV (IO91VT)
> http://www.dennison.demon.co.uk/activity.htm