[600MRG] WSJT-X 1.9 vs 1.8 WSPR test summary

[N1BUG]

Thanks to the WSJT development team for the (mainly LF and MF) WSPR 
decoding enhancement!

I am concluding my side by side LF and MF WSPR decoding comparison 
between WSJT-X 1.8.0 and 1.9.0-rc2. Here are results from the six 
night test. Each line specifies number of decodes in 1.9 / number of 
decodes in 1.8, percentage of decodes in 1.8 compared to 1.9. I 
never saw 1.8 decode anything 1.9 failed to decode.

02/26
LF 253/183   72%
MF 933/883   95%

02/27
LF 178/174   98%
MF 890/838   94%

02/28
LF 077/076   99%
MF 347/330   95%

03/01
LF 059/051   86%
MF 627/598   95%

03/02
LF 093/088   95%
MF 786/746   95%

03/03
LF 075/064   85%
MF 490/465   95%

On average over the six night test, 1.8 decoded 89% of what 1.9 
decoded on LF, 95% on MF. I have no explanation or theory for why 
the ratio remains nearly constant on MF but varies widely on LF. I 
will add that my man made noise level varies more at LF than at MF 
but I don't know if this is a factor.

The geomagnetic field was quiet the first night, active to minor 
storm the second and has not yet recovered fully at LF and MF (at 
least for high latitude paths from my location).

The number of decodes should not be considered valid for comparing 
one night against another. My WSPR transmitting activity on LF 
varied from WSPR-2 only to a combination of WSPR-2 and WSPR-15 which 
reduced available time slots for receiving and therefore the number 
of total decodes on some nights. Within a given night the numbers 
should provide a valid comparison between the two versions.

This is what I saw at my station. Your results may vary. I am now 
running only 1.9.0-rc2 for LF and MF WSPR monitoring.

73,
Paul N1BUG