Re: LF: WSPR Reports

[Andy Talbot <andy.g4jnt@gmail.com>]

The WSPR software does its own digital filtering to the 200Hz wide input band - with associated passband edges,  and calculations are based on the bandwidth defined  there.   

It will therefore work, and noise level calculation will be correct, provided the input signal cointains a flat noise spectrum wider than the internal filter bandwidth and encompasing the roll-off.

 

Signal level is measured in the 1.46Hz FFT bin then just normalised to a convenient reference 2.5kHz bandwidth used for reporting purposes.  Chosen because its nar enough an SSB filter's BW   It is a pity this bandwidth value was used as it has lead to misunderstandings amongst several ops, but has become the norm and we have to just accept 2.5kHz reference bandwidth for nearly all datamode operation now.   

 

Hindsight suggests we should all have normalised to 1Hz, quoting S/N in this width, but what's a constant 34dB offset  amongst friends :-)

 

Andy

www.g4jnt.com

 

 

On 4 January 2011 23:04, Piotr Mlynarski <sq7mpj@toya.net.pl> wrote:

Andy Talbot pisze:

300Hz wide CW filters made a fraction of a dB difference, compared with an SSB filter used as a reference;

well, indeed , in my free time i have to look inside the wspr code ..
i just thought  that it was simply 3000 Hz (ssb)/ 300 Hz (cw) = 10

                                                                    73, Piotr, sq7mpj
qth: Lodz /jo91rs/