Hi Andy

 

>Is the S/N calculation being a bit optimistic?  The calculation in WSJT / WSPR gives an artificially high S/N value if the input bandwidth is restricted below >a complete 2.5kHz's worth, and if impulsive noise is present.  So could ROS be suffering some similar effect?

 

I think the reports should be taken as ‘As Indicated By’ - WSPR or ROS. I have no idea how the S/N figure is produced but by observation in both systems it is obvious that averaging goes on.

Last night some of my S/N figures from ROS of -31dB were from a frame which had QSB -25dB short peaks and -35db troughs.

 

I have a WSPR capture from 6m in which my signal is decoded 51Hz lower than the signal frequency, due to Aircraft Doppler. The reflection is logically and visibly much weaker than the direct path signal. Yet WSPR has reported it only 1dB lower. The report is enhanced by another reflection from another Aircraft which is much stronger and does not de-code due to a 15Hz plus varying Doppler shift, it sweeps down in frequency and joins the decoded reflection for less than on third of the data period.

 

These are real life happenings, quite different from laboratory figures.

 

If the AS description sounds like Chinese, I can see the critics saying 51Hzis due to mains hum already, it isn’t. I can point you to the screen grabs and tests we were doing.