Lyle: Just played with the NR in both SSB and CW, on 80 and 40.
First, I can hear the change of the filters now without turning things on and off, as you mentioned it would work. The frequency response is decidedly less "flat" (less bassy) than the sound of 3.25. For CW, the 3 and 4 settings are much more efficient, not so in SSB, where the 1 and 2 settings are much cleaner sounding and better copy. I found that with no antenna and with the RF gain turned to 12 o'clock... Turn on the NR and all receiver white noise mutes and slowly ramps up. That's a recursive effect, I guess. I prefer the sound of the 3.25 version for ssb, its cleaner and somewhat less bright (flatter in response). I had to tweak the RX EQ to add some bass in the 3.27 version or it sounds "harsh" in my headsets (I dont use speakers hardly ever, btw). Inteligibility is marginally better with 3.27 on ssb and quite a bit better on CW, so you have drifted a bit to the CW side of the equation, but with a decent compromise for SSB. No more boingy peakyness in any mode, which is good, and the level while it still drops a bit, is better behaved. Can I assume from some comments I have read here that the NR process is pre EQ and AGC? How exactly do the processes "stack" in the radio architecture? If NR is at the top of the stack, a decent "fix" might be to somehow gang these three processes using presets so that they can be set up ahead of time and recalled from a memory button by the user. As an old brodcaster, we used to preset things in Switchers (vision mixers to UK readers) using a process called E-MEM... Which could recall preset parameters in "salvos" to preset multiple settings. This might work here. This would be handy in a contest environment where a minimum of tweaking and rapid adaptation to changing conditions is needed... You could play outside of a contest and create the settings then in the heat of battle, recall them with a single button push from a "canned" setup. It wouldnt be perfect for any environment, but it might mean the difference between working a mult and not working a mult. And multipliers, after all, are.... :) As you said, every receiving environment is different, but some generalizations can be made and being able to recall the multiple settings would be a definite plus feature of the radio. Thanks for letting us test these iterations. Lu Romero - W4LT K3 # 3192 Lyle Johnson wrote: > >> ...For now I'll treat the parameter as an opaque >> series of magic numbers. > > The way the new beta NR works is: > > Fx-y > > x selects the length of the filter. > > F1 = 121 taps, F2 = 91 taps, F3 = 61 taps, F4 = 31 taps > > (The Beta 3.25 release used FIR filters of 61 taps.) > > y selects values of Beta (gain), decay, and delay (how long the NR > algorithm waits to process a signal) > > (The Beta 3.5 release used the "x" parameter for these selections) > > 73, > > Lyle KK7P > > ______________________________________________________________ > Elecraft mailing list > Home: http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/mmfaq.htm > Post: mailto:Elecraft@mailman.qth.net > > This list hosted by: http://www.qsl.net > Please help support this email list: http://www.qsl.net/donate.html > > -- View this message in context: http://n2.nabble.com/K3-NR-tp3515659p3520809.html Sent from the [K3] mailing list archive at Nabble.com. ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
