Lin Davis wrote:
> the NR will quickly increase attenuation during the
period between CW elements and during pauses/low level
points in speech
Sinisa YT1NT wrote:
>How quick is "quickly" in real Orion operation?
Good question Sinisa. At reasonable
CW speeds, this would need to be in the order of
milliseconds to quiet noise between "dit" elements.
Wouldn't it be be simple to measure
NR effectiveness? Measure MDS (Minimum Discernable
Signal) with DSP BW = 100 Hz and NR Off. Then
switch NR On and measure it again. If NR improves
the S/N by lowering noise, MDS will become lower.
I still believe we are up against some
fundamental communications theory issues...at least
for CW. You cannot narrow CW bandwidth much more
than is already done with DSP = 100 without getting
into ringing problems caused by attempting to force
the CW signal's actual bandwidth into a filter that
is too narrow (and hence rings).
Regarding the manual's description of NR,
I've not seen anything that makes me believe V2's
NR works fundamentally differently. Most of the V2
changes were aimed at better partitioning of some
inter-communications timing issues between the three
processors and also a rewrite of the code using
a more code-efficient method (Weaver vs Filter).
Where possible, I understand existing V1 code was
copied. It could be that the NR code had to be
rewritten, but I doubt it would use a different
fundamental DSP technique than is in V1.
A simple MDS measurement might tell us
a lot about our perceptions, but then we wouldn't
have all this fun debating! ;-)
73, Bill W4ZV
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|