voicin.c

局域网ICQ的源代码

/* would be of much interest to an application in which LPC10 is */
/* embedded. */
/* listl and lincnt are not needed for an embedded LPC10 at all. */
/* 	integer nframe, nunsfm, iclip, maxosp, listl, lincnt */
/* 	common /contrl/ fsi, fso, fpi, fpo, fbi, fbo, pbin, fmsg, fdebug */
/* 	common /contrl/ quant, nbits */
/* 	common /contrl/ nframe, nunsfm, iclip, maxosp, listl, lincnt */
/* 	Parameters/constants */
/*       Voicing coefficient and Linear Discriminant Analysis variables: 
*/
/*       Max number of VDC's and VDC levels */
/*       The following are not Fortran PARAMETER's, but they are */
/*       initialized with DATA statements, and never modified. */
/*       Actual number of VDC's and levels */
/*       Local variables that need not be saved */
/*       Note: */

/*       VALUE(1) through VALUE(8) are assigned values, but VALUE(9) */
/*       never is.  Yet VALUE(9) is read in the loop that begins "DO I = 
*/
/*       1, 9" below.  I believe that this doesn't cause any problems in 
*/
/*       this subroutine, because all VDC(9,*) array elements are 0, and 
*/
/*       this is what is multiplied by VALUE(9) in all cases.  Still, it 
*/
/*       would save a multiplication to change the loop to "DO I = 1, 8". 
*/
/*       Local state */
/*       WARNING! */

/*       VOICE, SFBUE, and SLBUE should be saved from one invocation to */
/*       the next, but they are never given an initial value. */

/*       Does Fortran 77 specify some default initial value, like 0, or */
/*       is it undefined?  If it is undefined, then this code should be */
/*       corrected to specify an initial value. */

/*       For VOICE, note that it is "shifted" in the statement that */
/*       begins "IF (HALF .EQ. 1) THEN" below.  Also, uninitialized */
/*       values in the VOICE array can only affect entries in the VOIBUF 
*/
/*       array that are for the same frame, or for an older frame.  Thus 
*/
/*       the effects of uninitialized values in VOICE cannot linger on */
/*       for more than 2 or 3 frame times. */

/*       For SFBUE and SLBUE, the effects of uninitialized values can */
/*       linger on for many frame times, because their previous values */
/*       are exponentially decayed.  Thus it is more important to choose 
*/
/*       initial values for these variables.  I would guess that a */
/*       reasonable initial value for SFBUE is REF/16, the same as used */
/*       for FBUE and OFBUE.  Similarly, SLBUE can be initialized to */
/*       REF/32, the same as for LBUE and OLBUE. */

/*       These guessed initial values should be validated by re-running */
/*       the modified program on some audio samples. */

/*   Declare and initialize filters: */

    dither = (&st->dither);
    snr = (&st->snr);
    maxmin = (&st->maxmin);
    voice = (&st->voice[0]);
    lbve = (&st->lbve);
    lbue = (&st->lbue);
    fbve = (&st->fbve);
    fbue = (&st->fbue);
    ofbue = (&st->ofbue);
    olbue = (&st->olbue);
    sfbue = (&st->sfbue);
    slbue = (&st->slbue);

    /* Parameter adjustments */
    if (vwin) {
	--vwin;
	}
    if (buflim) {
	--buflim;
	}
    if (inbuf) {
	inbuf_offset = buflim[1];
	inbuf -= inbuf_offset;
	}
    if (lpbuf) {
	lpbuf_offset = buflim[3];
	lpbuf -= lpbuf_offset;
	}
    if (ivrc) {
	--ivrc;
	}
    if (obound) {
	--obound;
	}
    if (voibuf) {
	--voibuf;
	}

    /* Function Body */

/*       The following variables are saved from one invocation to the */
/*       next, but are not initialized with DATA statements.  This is */
/*       acceptable, because FIRST is initialized ot .TRUE., and the */
/*       first time that this subroutine is then called, they are all */
/*       given initial values. */

/*       SNR */
/*       LBVE, LBUE, FBVE, FBUE, OFBUE, OLBUE */

/*       MAXMIN is initialized on the first call, assuming that HALF */
/*       .EQ. 1 on first call.  This is how ANALYS calls this subroutine. 
*/

/*   Voicing Decision Parameter vector (* denotes zero coefficient): */

/* 	* MAXMIN */
/* 	  LBE/LBVE */
/* 	  ZC */
/* 	  RC1 */
/* 	  QS */
/* 	  IVRC2 */
/* 	  aR_B */
/* 	  aR_F */
/* 	* LOG(LBE/LBVE) */
/*  Define 2-D voicing decision coefficient vector according to the voicin
g*/
/*  parameter order above.  Each row (VDC vector) is optimized for a speci
fic*/
/*   SNR.  The last element of the vector is the constant. */
/* 	         E    ZC    RC1    Qs   IVRC2  aRb   aRf        c */

/*  The VOICE array contains the result of the linear discriminant functio
n*/
/*   (analog values).  The VOIBUF array contains the hard-limited binary 
*/
/*   voicing decisions.  The VOICE and VOIBUF arrays, according to FORTRAN
 */
/*   memory allocation, are addressed as: */

/* 	   (half-frame number, future-frame number) */

/* 	   |   Past    |  Present  |  Future1  |  Future2  | */
/* 	   | 1,0 | 2,0 | 1,1 | 2,1 | 1,2 | 2,2 | 1,3 | 2,3 |  --->  time */

/*   Update linear discriminant function history each frame: */
    if (*half == 1) {
	voice[0] = voice[2];
	voice[1] = voice[3];
	voice[2] = voice[4];
	voice[3] = voice[5];
	*maxmin = *maxamd / max(*minamd,1.f);
    }
/*   Calculate voicing parameters twice per frame: */
    vparms_(&vwin[1], &inbuf[inbuf_offset], &lpbuf[lpbuf_offset], &buflim[1], 
	    half, dither, mintau, &zc, &lbe, &fbe, &qs, &rc1, &ar_b__, &
	    ar_f__);
/*   Estimate signal-to-noise ratio to select the appropriate VDC vector. 
*/
/*   The SNR is estimated as the running average of the ratio of the */
/*   running average full-band voiced energy to the running average */
/*   full-band unvoiced energy. SNR filter has gain of 63. */
    r__1 = (*snr + *fbve / (real) max(*fbue,1)) * 63 / 64.f;
    *snr = (real) i_nint(&r__1);
    snr2 = *snr * *fbue / max(*lbue,1);
/*   Quantize SNR to SNRL according to VDCL thresholds. */
    snrl = 1;
    i__1 = nvdcl - 1;
    for (snrl = 1; snrl <= i__1; ++snrl) {
	if (snr2 > vdcl[snrl - 1]) {
	    goto L69;
	}
    }
/*   	(Note:  SNRL = NVDCL here) */
L69:
/*   Linear discriminant voicing parameters: */
    value[0] = *maxmin;
    value[1] = (real) lbe / max(*lbve,1);
    value[2] = (real) zc;
    value[3] = rc1;
    value[4] = qs;
    value[5] = ivrc[2];
    value[6] = ar_b__;
    value[7] = ar_f__;
/*   Evaluation of linear discriminant function: */
    voice[*half + 3] = vdc[snrl * 10 - 1];
    for (i__ = 1; i__ <= 8; ++i__) {
	voice[*half + 3] += vdc[i__ + snrl * 10 - 11] * value[i__ - 1];
    }
/*   Classify as voiced if discriminant > 0, otherwise unvoiced */
/*   Voicing decision for current half-frame:  1 = Voiced; 0 = Unvoiced */
    if (voice[*half + 3] > 0.f) {
	voibuf[*half + 6] = 1;
    } else {
	voibuf[*half + 6] = 0;
    }
/*   Skip voicing decision smoothing in first half-frame: */
/*     Give a value to VSTATE, so that trace statements below will print 
*/
/*     a consistent value from one call to the next when HALF .EQ. 1. */
/*     The value of VSTATE is not used for any other purpose when this is 
*/
/*     true. */
    vstate = -1;
    if (*half == 1) {
	goto L99;
    }
/*   Voicing decision smoothing rules (override of linear combination): */

/* 	Unvoiced half-frames:  At least two in a row. */
/* 	-------------------- */

/* 	Voiced half-frames:    At least two in a row in one frame. */
/* 	-------------------    Otherwise at least three in a row. */
/* 			       (Due to the way transition frames are encoded) */

/* 	In many cases, the discriminant function determines how to smooth. */
/*	In the following chart, the decisions marked with a * may be overridden
.*/

/*   Voicing override of transitions at onsets: */
/* 	If a V/UV or UV/V voicing decision transition occurs within one-half 
*/
/* 	frame of an onset bounding a voicing window, then the transition is */
/* 	moved to occur at the onset. */

/* 	P	1F */
/* 	-----	----- */
/* 	0   0   0   0 */
/* 	0   0   0*  1	(If there is an onset there) */
/* 	0   0   1*  0*	(Based on 2F and discriminant distance) */
/* 	0   0   1   1 */
/* 	0   1*  0   0	(Always) */
/* 	0   1*  0*  1	(Based on discriminant distance) */
/* 	0*  1   1   0*	(Based on past, 2F, and discriminant distance) */
/* 	0   1*  1   1	(If there is an onset there) */
/* 	1   0*  0   0	(If there is an onset there) */
/* 	1   0   0   1 */
/* 	1   0*  1*  0	(Based on discriminant distance) */
/* 	1   0*  1   1	(Always) */
/* 	1   1   0   0 */
/* 	1   1   0*  1*	(Based on 2F and discriminant distance) */
/* 	1   1   1*  0	(If there is an onset there) */
/* 	1   1   1   1 */

/*   Determine if there is an onset transition between P and 1F. */
/*   OT (Onset Transition) is true if there is an onset between */
/*   P and 1F but not after 1F. */
    ot = ((obound[1] & 2) != 0 || obound[2] == 1) && (obound[3] & 1) == 0;
/*   Multi-way dispatch on voicing decision history: */
    vstate = (voibuf[3] << 3) + (voibuf[4] << 2) + (voibuf[5] << 1) + voibuf[
	    6];
    switch (vstate + 1) {
	case 1:  goto L99;
	case 2:  goto L1;
	case 3:  goto L2;
	case 4:  goto L99;
	case 5:  goto L4;
	case 6:  goto L5;
	case 7:  goto L6;
	case 8:  goto L7;
	case 9:  goto L8;
	case 10:  goto L99;
	case 11:  goto L10;
	case 12:  goto L11;
	case 13:  goto L99;
	case 14:  goto L13;
	case 15:  goto L14;
	case 16:  goto L99;
    }
L1:
    if (ot && voibuf[7] == 1) {
	voibuf[5] = 1;
    }
    goto L99;
L2:
    if (voibuf[7] == 0 || voice[2] < -voice[3]) {
	voibuf[5] = 0;
    } else {
	voibuf[6] = 1;
    }
    goto L99;
L4:
    voibuf[4] = 0;
    goto L99;
L5:
    if (voice[1] < -voice[2]) {
	voibuf[4] = 0;
    } else {
	voibuf[5] = 1;
    }
    goto L99;
/*   VOIBUF(2,0) must be 0 */
L6:
    if (voibuf[1] == 1 || voibuf[7] == 1 || voice[3] > voice[0]) {
	voibuf[6] = 1;
    } else {
	voibuf[3] = 1;
    }
    goto L99;
L7:
    if (ot) {
	voibuf[4] = 0;
    }
    goto L99;
L8:
    if (ot) {
	voibuf[4] = 1;
    }
    goto L99;
L10:
    if (voice[2] < -voice[1]) {
	voibuf[5] = 0;
    } else {
	voibuf[4] = 1;
    }
    goto L99;
L11:
    voibuf[4] = 1;
    goto L99;
L13:
    if (voibuf[7] == 0 && voice[3] < -voice[2]) {
	voibuf[6] = 0;
    } else {
	voibuf[5] = 1;
    }
    goto L99;
L14:
    if (ot && voibuf[7] == 0) {
	voibuf[5] = 0;
    }
/* 	GOTO 99 */
L99:
/*   Now update parameters: */
/*   ---------------------- */

/*  During unvoiced half-frames, update the low band and full band unvoice
d*/
/*   energy estimates (LBUE and FBUE) and also the zero crossing */
/*   threshold (DITHER).  (The input to the unvoiced energy filters is */
/*   restricted to be less than 10dB above the previous inputs of the */
/*   filters.) */
/*   During voiced half-frames, update the low-pass (LBVE) and all-pass */
/*   (FBVE) voiced energy estimates. */
    if (voibuf[*half + 6] == 0) {
/* Computing MIN */
	i__1 = fbe, i__2 = *ofbue * 3;
	r__1 = (*sfbue * 63 + (min(i__1,i__2) << 3)) / 64.f;
	*sfbue = i_nint(&r__1);
	*fbue = *sfbue / 8;
	*ofbue = fbe;
/* Computing MIN */
	i__1 = lbe, i__2 = *olbue * 3;
	r__1 = (*slbue * 63 + (min(i__1,i__2) << 3)) / 64.f;
	*slbue = i_nint(&r__1);
	*lbue = *slbue / 8;
	*olbue = lbe;
    } else {
	r__1 = (*lbve * 63 + lbe) / 64.f;
	*lbve = i_nint(&r__1);
	r__1 = (*fbve * 63 + fbe) / 64.f;
	*fbve = i_nint(&r__1);
    }
/*   Set dither threshold to yield proper zero crossing rates in the */
/*   presence of low frequency noise and low level signal input. */
/*   NOTE: The divisor is a function of REF, the expected energies. */
/* Computing MIN */
/* Computing MAX */
    r__2 = sqrt((real) (*lbue * *lbve)) * 64 / 3000;
    r__1 = max(r__2,1.f);
    *dither = min(r__1,20.f);
/*   Voicing decisions are returned in VOIBUF. */
    return 0;
} /* voicin_ */