For now, change to BT=1.5 for FT8. This commit enables the waterfall analysis window. This decreases sidelobes significantly in exchange for increased noise bandwidth. Sensitivity to weak tones will be degraded somewhat.

lib/ft8/gen_ft8wave.f90

@@ -17,7 +17,7 @@ subroutine gen_ft8wave(itone,nsym,nsps,fsample,f0,cwave,wave,icmplx,nwave)
      twopi=8.0*atan(1.0)
      dt=1.0/fsample
      hmod=1.0
-     bt=1.0
+     bt=1.5
 ! Compute the frequency-smoothing pulse
      do i=1,3*nsps
         tt=(i-1.5*nsps)/real(nsps)

lib/symspec.f90

@@ -48,10 +48,23 @@ subroutine symspec(shared_data,k,ntrperiod,nsps,ingain,nminw,pxdb,s,   &
   if(nfft3.ne.nfft3z) then
 ! Compute new window
      pi=4.0*atan(1.0)
-     width=0.25*nsps
+!     width=0.25*nsps
+!     do i=1,nfft3
+!        z=(i-nfft3/2)/width
+!        w3(i)=exp(-z*z)
+!     enddo
+! Coefficients taken from equation 37 of NUSC report:
+! "Some windows with very good sidelobe behavior: application to 
+! discrete Hilbert Transform, by Albert Nuttall"
+!
+     a0=0.3635819
+     a1=0.4891775;
+     a2=0.1365995;
+     a3=0.0106411;
      do i=1,nfft3
-        z=(i-nfft3/2)/width
-        w3(i)=exp(-z*z)
+         w3(i)=a0-a1*cos(2*pi*(i-1)/(nfft3))+ &
+                  a2*cos(4*pi*(i-1)/(nfft3))+ &
+                  a3*cos(6*pi*(i-1)/(nfft3))
      enddo
      nfft3z=nfft3
   endif
@@ -86,7 +99,7 @@ subroutine symspec(shared_data,k,ntrperiod,nsps,ingain,nminw,pxdb,s,   &
   enddo
   ihsym=ihsym+1
 
-!  xc(0:nfft3-1)=w3(1:nfft3)*xc(0:nfft3-1)    !Apply window w3
+  xc(0:nfft3-1)=w3(1:nfft3)*xc(0:nfft3-1)    !Apply window w3
   call four2a(xc,nfft3,1,-1,0)               !Real-to-complex FFT
 
   df3=12000.0/nfft3                   !JT9-1: 0.732 Hz = 0.42 * tone spacing