It's all about the Software

Does anyone know the details behind the virtual SWR meter. The documentation says the noise is derived from 70 percent of the bins. I believe this is 70 percent of the total bandwidth. I am trying to make a sun noise measurement of an antenna. There are spurious signals near the frequency I am making the measurement at (solar cell inverter). I set the IF (the green box of the FFT display) to keep the spurs out, but it does not help. Any ideas on how to narrow the range of frequencies that the noise is calculated from?
I would like to set the modulation to USB and the filter to 2.4 kHz, like I was using an old radio, and just get a noise reading for that narrow piece of the band.

If you are looking for anything with a high degree of accuracy this is not the tool to use. Go get Linrad. Leif went to some effort to make it able to be very well calibrated.
{o.o}

Pardon the ignorance, what is Linrad, and where do I get it?
Thanks!

but what you want... is that even possible? There's always some kind of automatic gain adjustment inside the radio, RF or IF. Never seen a sdr radio or software where you can set both off to get a reliable noisefloor reading.

What SDR radio are you using with SDRC? Did you looked at the Signal History under tabs View - Spectrum?
It measures the signal/noise within the bandwith you choose, fe 2,8 KHz USB