[CQ-Contest] Summarizing the Skimmer Accuracy Thread

[VR2BrettGraham]

UA9CDC added:

>> Another dimension to this skimming thing is spotting on wrong 
>> frequencies. Or the wrong band.  Take the RBN data for a day - a 
>> non-weekend day is better, as few stations actually on >1 band at 
>> once then - and look for the same call spotted say 3 kc or more from 
>> others at exactly the same time.
>>
>> Open that window to a more realistic minute or two & see just how 
>> many there are!
> Every contest I run into several stations on CW who actually transmitt 
> 3 signals. Main signal is usually 59+30 or more and two side bands 
> spaced 1-2 kHz from the main can be as loud as 59.
> No wonder, close by skimmers decode then on these sideband frequencies.

That doesn't surprise me, given all the fecal signals I see on the bands 
in my reincarnation as an SWL.

Though what I was referring to are skimmers reporting signals on 
frequencies that the signals are not really on.

How many wrong frequency spots are there on RBN?  Well, back when I was 
looking for just same-call-different-band & 
spotted-on-exactly-the-same-second, I found 140 such spots in a day (out 
of 120k).

Open that window up to ~1.4 minutes & there were ~3700 spots of the same 
call on a different band than other skimmers had spotted it on (and 
probably more, as that excludes all spots of calls that might be on >1 
band at the same time [DXpeditions, special event stations, etc]).

Of those 3700 same-call-different-band spots, about 100 were for 
UA9CDC.  UA9CDC very likely only operating on one frequency during the 
~1.4 minute period concerned.  Where UA9CDC was said to be:

20m: 14042.2, 14042.3

15m: 21039.9, 21040, 21040.1

10m: 28041.9, 28042, 28042.1

There are a number of mechanisms involved & it isn't just the poorly 
implemented but popular quadrature sampling detector-based receivers 
that do this.

73, ex-VR2BG/p.