While I'm not an expert at DSP, I know some time series tricks (time
series analysis is the basis for DSP) and I know some of the
statistics behind the time series analysis. Good DSP is the efficient
implementation of the time series stuff.
How the de-noised spectrum looks depends on how the correlation
filter is implemented, but the algorithms will typically raise the
amplitude of the coherent frequency component while decreasing the
amplitude of the incoherent components. In addition, the algorithms
are probably tailored for certain components. The filters aren't
perfect, because they're based on infinite convergent series that
must be truncated. Because of the truncation, stuff close to the
coherent component is also enhanced, so it looks a little like a
"peaky" bandpass filter with odd, lobed skirts.
A simple bandpass filter won't do any better at removing noise than
will any other narrow filter but, under the right conditions DSP
can. There are limits and, as Jerry said in a private exchange,
there's probably a Ph.D. dissertation behind each of the different
methods and their implementation. I have no idea how DSP is
implemented in the Orion or how configuarble the algorithms are.
There is, however, no free lunch and diddling with one aspect will
affect others.
Kim Elmore, N5OP
At 05:58 PM 12/8/2006, you wrote:
Jerry said:
"While the effect on the output spectrum is that of a bandpass filter, I
don't think NR works that way. I think it works on correlation of the
pass band with a time delayed copy of the pass band."
-------------------------------------------------------------------------------------------------
Jerry, I am curious as to why you believe this? Are there other
amateur transceivers - ICOM or Yaesu - where you have seen the
technique you described used before?
Thanks & 73,
Merle - W0EWM
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|