Hi Bill,

> Steve, I am unsure about a couple of the changes I have made, particularly 
> the limited slice of the cdat complex array used when applying the Rayleigh 
> fading model on line 275 (necessary to avoid an array bounds violation when 
> using 11025 Hz rate), and whether I should apply clipping to 16-bit full 
> scale after applying the gain offset at line 288. I also don't understand 
> what line 287 is doing as it seems to apply a normalization factor 'fac' that 
> I would expect to be applied on line 288 yet that line used a fixed factor of 
> 100.0 (`rms`). Perhaps you can review my changes and advise?

I think that jt65sim is working properly in r8391. The fac vs rms thing was 
taking care of a hidden feature, whereby the simulator generates a noiseless 
signal whenever the requested SNR exceeds 90 dB. That wasn’t working properly 
with nonzero Doppler spread, even in the original version, because the peak 
amplitude is random - so I made some changes that will cause a noiseless 
(SNR>90 dB) signal to be scaled so that the (random) peak magnitude is 32767. 

I also changed the hardwired Nyquist frequency from 6000.0 to (fsample/2) so 
that the noise should be scaled properly for either sampling rate.

With the new option for an arbitrary “gain offset", I decided that it would 
also be worthwhile to print a warning if the simulated data is clipped in the 
wav file. The user should probably re-run with a smaller gain value if clipping 
is detected.

Steve k9an



------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
wsjt-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Reply via email to