Hi Grant,
I've done a bit of research into DSP filtering and even called Gary at Ten-Tec
to find out what the NR filter implementation is.
And you are correct, it is certainly not just a form of an expansion circuit!!
By Doug Smith, it's called an adaptive predictor. The theory behind it was
developed 40+ years ago by Kalman, who based his work on Wiener and Kolmogoroff.
To me, it's pretty cool stuff.
But unfortunately, I still thing the manual description of it's behavior is in
need of an update. I was hoping to contact Mr. Smith to help me understand it
better, since he has most likely coded and played with it.
I'll follow up with more details soon. And hopefully that post won't put too
many folks out there to sleep :)
73,
Lin
WB1AIW
Grant Youngman wrote:
but I DO still hear a difference. I can only conclude from
this that there is no adaption component to the NR function,
and therefore the NR value is not controlling an adaption
rate, but something else, which sounds to me to be an
amplitude threshold.
We seem to be talking around each other, but there's no reason to believe NR
level is controlling some simple threshold.
All I can suggest you do is take it up with Gary at T-T for a definitive
answer straight from the proverbial horses mouth :-)
My take on it, based on experience and T-T's own description is that:
1. It's adaptive
2. It builds dynamic filters around signal spectral components
3. Since the filters are dynamic, they are built and rebuilt as the
apparent spectral components change in the bandpass, at a rate determined by
the NR setting. (I'm not saying this is precisely the mechanism, but it is
the effect of the mechanism. T-T has not published, that I'm aware of, the
specific algorithms employed).
4. The closer the signal amplitude gets to the noise level, the longer
adaption will take -- and in some cases, it may not take at all, in which
case, turn it off or turn down the THRESHOLD level or turn up the RF gain.
And if that doesn't work, then turn it off.
5. It's primary function is to improve S/N ratio, not provide a "noise
free" environment for listening enjoyment :-)
Note that I am not making statements about how WELL it works, or doesn't.
Many seem to think it worked better in 1.371 and earlier releases than the
current realization in 2.xxx.
Grant/NQ5T
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|