[Skimmertalk] Help--What's Going On?

[Bob Wilson, N6TV]

On Wed, Jul 18, 2012 at 9:29 AM, David Robbins <k1ttt at verizon.net> wrote:

> maybe there is something in the call validation logic that is more
> sensitive to the 2x2 call


Actually the logic is *insensitive* to WS7*xx* callsigns in particular,
though this can easily be corrected.

The callsign recognition algorithm is based on callsign patterns in a file
named *Patt3Ch.lst*, which on my XP system is found in:

C:\Documents and Settings\[userid]\Application Data\Afreet\Reference\


(it will probably be in a different location on a Vista or Windows 7
system, maybe under

C:\Users\[Userid]\AppData\**Roaming\Afreet\Reference\


Inside that file you can see the problem. Any callsign beginning with WS7
is expected to have just *one* letter after the prefix, not two:

  WS0@
+ WS1@
  WS2@
  WS3@
  WS4@
  WS4@@
  WS5@
  WS6@
  WS6@@
+ WS7@
  WS8@
  WS8@@
+ WS9@

A plus sign in front of a pattern in the file means that the calls that
match the pattern are "most likely correct," that is they will be spotted
faster by the Skimmers.

Apparently there were too few WS7*xx* callsigns in use at the time that the
pattern file was created, so it was not included.

If you feel strongly about this, you try to ask all Skimmer ops to add WS7@@
to the list. :-)  Then WS7SX will be spotted by Skimmers.

73,
Bob, N6TV

On Wed, Jul 18, 2012 at 10:37 AM, Mark n2qt <n2qt at verizon.net> wrote:

> The call detection algorithm can be a bit fussy.   I watched vp9/wa4pgm
> roll
> in on
> 6M for ages with a good signal and good decoding on screen.  However he was
> never
> spotted with the full call, just a couple of times with wa4pgm.  I added
> vp9/wa4pgm
> to the watch list (which does seem to help with the /B beacons) but still
> nothing.
> I had the skimmer set to aggressive, a second skimmer running normal did
> much
> better, but also has more busted spots than I feel comfortable putting out
> on the RBN.
> Alex said he used a call history file to develop his algorithm so it seems
> likely that
> the more uncommon call structures have a higher decode threshold.
>
> Mark n2qt
>
>
>
> -----Original Message-----
> From: David Robbins
> Sent: Wednesday, July 18, 2012 12:29 PM
> To: Skimmertalk at contesting.com
> Subject: Re: [Skimmertalk] Help--What's Going On?
>
>
> well, that may be part of it, but not all since my skimmer doesn't use the
> supercheck file.  maybe there is something in the call validation logic
> that
> is more sensitive to the 2x2 call?  or some kind of sending quirk that the
> skimmer can't figure out??   maybe try some variations with extra half
> spaces in the ws7sx call to break it up differently.
>
> Jul 18, 2012 12:06:16 PM, k1ttt at arrl.net wrote:
>
>
> let me guess, you see more spots with one call than the other??? most
> likely
> skimmers that are using the supercheck database are finding one call in the
> database but not the other one so it doesn't get spotted.
>
> Jul 18, 2012 11:56:55 AM, w6sx at arrl.net wrote:
>
> Good Morning All,
>
> I have a club call which I sometimes use to tweak RBN for propagation
> checking. But I am often surprised by the lack of hits.
>
> Last night I ran tests with both W6SX and WS7SX. I alternated TEST TEST
> TEST W6SX W6SX W6SX TEST and TEST TEST TEST WS7SX WS7SX WS7SX TEST numerous
> times on various bands. Everything was the same: power, speed, memory keyer
> with auto character space. To see results, use Search Spots by Callsign.
>
> What's going on?
>
> Ski Exuberantly,
>
> Hank, W6SX
>
> Mammoth Lakes, California
>
> Elevation 8083 feet in John Muir's Range of Light
> _______________________________________________
> Skimmertalk mailing list
> Skimmertalk at contesting.com
> http://dayton.contesting.com/mailman/listinfo/skimmertalk
> _______________________________________________
> Skimmertalk mailing list
> Skimmertalk at contesting.com
> http://dayton.contesting.com/mailman/listinfo/skimmertalk
>
> _______________________________________________
> Skimmertalk mailing list
> Skimmertalk at contesting.com
> http://dayton.contesting.com/mailman/listinfo/skimmertalk
>