In subtractft8.f90: refine DT for early decodes before subtracting them from dd().

lib/ft8/ft8b.f90

@@ -418,7 +418,7 @@ subroutine ft8b(dd0,newdat,nQSOProgress,nfqso,nftx,ndepth,nzhsym,lapon,     &
      endif
      nbadcrc=0  ! If we get this far: valid codeword, valid (i3,n3), nonquirky message.
      call get_ft8_tones_from_77bits(message77,itone)
-     if(lsubtract) call subtractft8(dd0,itone,f1,xdt)
+     if(lsubtract) call subtractft8(dd0,itone,f1,xdt,.false.)
 !     write(21,3001) nzhsym,npasses,nqsoprogress,ipass,iaptype,lsubtract,   &
 !          f1,xdt,msg37(1:22); flush(21)
 !3001 format(5i3,L3,f7.1,f7.2,2x,a22)

lib/ft8/subtractft8.f90

@@ -1,66 +1,113 @@
-subroutine subtractft8(dd,itone,f0,dt)
+subroutine subtractft8(dd0,itone,f0,dt,ldt)
 
-! Subtract an ft8 signal
+! Subtract an ft8 signal.  If ldt==.true., refine DT first.
-!
+  
-! Measured signal  : dd(t)    = a(t)cos(2*pi*f0*t+theta(t))
+! Raw data         : dd(t)    = a(t)cos(2*pi*f0*t+theta(t))
 ! Reference signal : cref(t)  = exp( j*(2*pi*f0*t+phi(t)) )
-! Complex amp      : cfilt(t) = LPF[ dd(t)*CONJG(cref(t)) ]
-! Subtract         : dd(t)    = dd(t) - 2*REAL{cref*cfilt}
-
-  use timer_module, only: timer
   
   parameter (NMAX=15*12000,NFRAME=1920*79)
-  parameter (NFFT=NMAX,NFILT=1400)
+  real dd(NMAX),dd0(NMAX)
-  real*4  dd(NMAX), window(-NFILT/2:NFILT/2), xjunk
+  complex cref(NFRAME)
-  complex cref,camp,cfilt,cw
+  logical ldt
-  integer itone(79)
-  logical first
-  data first/.true./
-  common/heap8/cref(NFRAME),camp(NMAX),cfilt(NMAX),cw(NMAX),xjunk(NFRAME)
-  save first
 
-  nstart=dt*12000+1
+! Generate complex reference waveform
-  nsym=79
+  call gen_ft8wave(itone,79,1920,2.0,12000.0,f0,cref,xjunk,1,NFRAME)
-  nsps=1920
+
-  fs=12000.0
+  if(ldt) then                           !Are we refining DT ?
-  icmplx=1
+     sqa=sqf(dd0,cref,f0,dt,ldt,-300,dd) !Yes
-  bt=2.0 
+     sqb=sqf(dd0,cref,f0,dt,ldt,300,dd)
-  call gen_ft8wave(itone,nsym,nsps,bt,fs,f0,cref,xjunk,icmplx,NFRAME)
+  endif
-  camp=0.
+  sq0=sqf(dd0,cref,f0,dt,ldt,0,dd)       !Do the subtraction with idt=0
-  do i=1,nframe
+  if(ldt) then
-    id=nstart-1+i 
+     call peakup(sqa,sq0,sqb,dx)
-    if(id.ge.1.and.id.le.NMAX) camp(i)=dd(id)*conjg(cref(i))
+     if(abs(dx).gt.1.0) goto 100         !No acceptable minimum: do not subtract
-  enddo
+     i1=nint(300.0*dx)                   !First approximation of better idt
+     sqa=sqf(dd0,cref,f0,dt,ldt,i1-60,dd)
+     sqb=sqf(dd0,cref,f0,dt,ldt,i1+60,dd)
+     sq0=sqf(dd0,cref,f0,dt,ldt,i1,dd)
+     call peakup(sqa,sq0,sqb,dx)
+     if(abs(dx).gt.1.0) then             !No acceptable minimum
+        sq0=sqf(dd0,cref,f0,dt,ldt,0,dd) !Use idt=0 for subtraction
+        go to 100
+     endif
+     i2=nint(60.0*dx) + i1               !Best estimate of idt
+     sq0=sqf(dd0,cref,f0,dt,ldt,i2,dd)   !Do the subtraction with idt=i2
+  endif
+100 dd0=dd                               !Return dd0 with signal subtracted
+
+  return
+end subroutine subtractft8
+
+real function sqf(dd0,cref,f0,dt,ldt,idt,dd)
+
+! Raw data         : dd0(t)   = a(t)cos(2*pi*f0*t+theta(t))
+! Reference signal : cref(t)  = exp( j*(2*pi*f0*t+phi(t)) )
+! Complex amp      : cfilt(t) = LPF[ dd(t)*CONJG(cref(t)) ]
+! Subtract         : dd(t)    = dd0(t) - 2*REAL{cref*cfilt}
+  
+  parameter (NMAX=15*12000,NFRAME=1920*79)
+  parameter (NFFT=NMAX,NFILT=2800)
+  real dd(NMAX),dd0(NMAX)
+  real window(-NFILT/2:NFILT/2)
+  real x(NFFT+2)
+  complex cx(0:NFFT/2),cref(NFRAME)
+  complex camp,cfilt,cw,z
+  logical first,ldt
+  data first/.true./
+  common/heap8/camp(NMAX),cfilt(NMAX),cw(NMAX)
+  equivalence (x,cx)
+  save first,/heap8/
 
   if(first) then
 ! Create and normalize the filter
      pi=4.0*atan(1.0)
      fac=1.0/float(nfft)
-     sum=0.0
+     sumw=0.0
      do j=-NFILT/2,NFILT/2
         window(j)=cos(pi*j/NFILT)**2
-        sum=sum+window(j)
+        sumw=sumw+window(j)
      enddo
      cw=0.
-     cw(1:NFILT+1)=window/sum
+     cw(1:NFILT+1)=window/sumw
      cw=cshift(cw,NFILT/2+1)
      call four2a(cw,nfft,1,-1,1)
      cw=cw*fac
      first=.false.
   endif
   
-  cfilt=0.0
+  nstart=dt*12000+1 + idt
+  camp=0.
+  dd=dd0
+  do i=1,nframe
+     j=nstart-1+i 
+     if(j.ge.1.and.j.le.NMAX) camp(i)=dd(j)*conjg(cref(i))
+  enddo
   cfilt(1:nframe)=camp(1:nframe)
+  cfilt(nframe+1:)=0.0
   call four2a(cfilt,nfft,1,-1,1)
   cfilt(1:nfft)=cfilt(1:nfft)*cw(1:nfft)
   call four2a(cfilt,nfft,1,1,1)
 
-! Subtract the reconstructed signal
+  x=0.
   do i=1,nframe
      j=nstart+i-1
-     if(j.ge.1 .and. j.le.NMAX) dd(j)=dd(j)-2*REAL(cfilt(i)*cref(i))
+     if(j.ge.1 .and. j.le.NMAX) then
+        z=cfilt(i)*cref(i)
+        dd(j)=dd(j)-2.0*real(z)      !Subtract the reconstructed signal
+        x(i)=dd(j)
+     endif
   enddo
+  sq=0.
+  if(ldt) then
+     call four2a(cx,NFFT,1,-1,0)                 !Forward FFT, r2c
+     df=12000.0/NFFT
+     ia=(f0-1.5*6.25)/df
+     ib=(f0+8.5*6.25)/df
+     do i=ia,ib
+        sq=sq + real(cx(i))*real(cx(i)) + aimag(cx(i))*aimag(cx(i))
+     enddo
+  endif
+  sqf=sq
 
   return
-end subroutine subtractft8
+end function sqf
-

lib/ft8_decode.f90

@@ -77,7 +77,7 @@ contains
     endif
     if(nzhsym.eq.50 .and. ndec_early.ge.1) then
        do i=1,ndec_early
-          call subtractft8(dd,itone_save(1,i),f1_save(i),xdt_save(i))
+          call subtractft8(dd,itone_save(1,i),f1_save(i),xdt_save(i),.true.)
        enddo
     endif
     ifa=nfa