[RTTY] RTTY Skimmer - DF4UE

[Salvatore Irato]

Hi Pete, Lee and all.
This was and is the RCKSkimmer made by Walter DL4RCK:

http://www.walter-dallmeier.de/software-by-dl4rck/rckskimmer/

I didn't have the tooling since then to observe statistically what's coming out.
But for a lot of time there was only the RCKskimmer, a brilliant way
to mix old and new in my personal opinion.
So, I couldn't say that what you noted is wrong, but well before the
advent of the now RTTY Skimmer the RCKskimmer worked and worked well.
I used to have my cluster connected to dl4rck.ham-radio-op.net:8000 or
S50ARX, if I recall right this last.
There where not that many errors as now. As I sad in a mail before,
due to my lack of tooling and thus evidences, it could be that the
large number of error and strange errors - as Lee VE7CC reports in
another mail - come from the simplest statistical rules: more stones,
then more diamonds in the bunch.

I don't know how Walter DL4RCK filtered out CQ a la carte:
TEST IW1AYD IW1AYD CQ #### IZ5RKW IZ5RKW ####
Never went out as IZ5RKW calling CQ.
I know that just the keywords to make the filers were CQ, TEST and a
variable filed reserved for user when Contest claimed for a dedicated
call in:
VOLTA IW1AYD IW1AYD CQ

It is not an all that right example but gives the idea.

Walter, DL4RCK, used MMVARI as the physical decoder, wrapping around
it his code and knowledge. Thus the RCKskimmer works also for any PSK
schema available into MMVARI.

That's, just to say. Not to say that anyone is better or worse. The
DL4RCK RCKskimmer have some features that are advantages over the RTTY
Skimmer and also the latest RTTY Skimmer have some features that give
advantage over the RCKskimmer. That is in the life. Just looking at
the two: timings, spreads, string checks are extremely different and
subjected to operators setup abilities.
I.E. playing with RTTY decoders "filters" and "features" gives
astonishing results that aren't to be applied steadily to any
different environment we face.
RTTY Skimmer is much more modern and untouchable by those aspects.

I wish to look forward to anyone that is willingly to give real
numbers to be compared with numbers by others about record quality.
Not a single source of data, or as in my case, personal sensations
from single places, could give the right view.

I still see now and since not that much time ago more wrong wrong and
bogus calls with clear error on each possible digit. Often, N1MM+
bandmap, the right call is there (just another call) followed by a
bugus callsign (that is marked as a multiplier).

BTW when we used IY1GMS, a pretty unknown special call in memory of
Guglielmo Marconi, I had seen any variable printout as anyone could
imagine. Next will be IY1A.
This is my personal point of view, I see much more error now and I
look further forward to understand why. I am just scratching the
surface, I know.

  73 de iw1ayd Salvo