diff --git a/gr-sim/tfv_flying_6502.c b/gr-sim/tfv_flying_6502.c
index 9d9650ff..0c2ead25 100644
--- a/gr-sim/tfv_flying_6502.c
+++ b/gr-sim/tfv_flying_6502.c
@@ -398,7 +398,7 @@ static int fixed_mul_unsigned(
 	a&=0xff;
 
 //	product[0]=a;				// sta PRODUCT+0	;
-	_aa=a;
+//	_aa=a;
 
 	//sm3a:
 	a=square1_hi[sm3a+x];			// lda square1_hi,x	; 4
@@ -569,27 +569,27 @@ static void fixed_mul(unsigned char x_i, unsigned char x_f,
 
 	a=(x_i&0xff);			// lda T1+1			; 3
 					cycles.multiply+=12;
-	if ((a&0x80)==0) goto x_positive;	// bpl :+		; 3/2nt
-					cycles.multiply--;
+	if ((a&0x80)==0) goto x_positive;	// bpl :+		;^3/2nt
+
 	c=1;				// sec				; 2
 	a=product[2];			// lda PRODUCT+2		; 3
 	a+=(~y_f)+c;			// sbc T2+0			; 3
 	c=!(a&0x100);
 	a&=0xff;
 	product[2]=a;			// sta PRODUCT+2		; 3
-	a=product[3];			// lda PRODUCT+3		; 3
-	a+=(~y_i)+c;			// sbc T2+1			; 3
-	c=!(a&0x100);
-	a&=0xff;
-	product[3]=a;			// sta PRODUCT+3		; 3
-					cycles.multiply+=20;
+//	a=product[3];			// lda PRODUCT+3		;
+//	a+=(~y_i)+c;			// sbc T2+1			;
+//	c=!(a&0x100);
+//	a&=0xff;
+//	product[3]=a;			// sta PRODUCT+3		;
+					cycles.multiply+=10;
 
 x_positive:
 
 	a=(y_i&0xff);			// lda T2+1			; 3
 					cycles.multiply+=6;
-	if ((a&0x80)==0) goto y_positive;	// bpl :+		; 3/2nt
-					cycles.multiply--;
+	if ((a&0x80)==0) goto y_positive;	// bpl :+		;^3/2nt
+
 
 	c=1;				// sec				; 2
 	a=product[2];			// lda PRODUCT+2		; 3
@@ -597,12 +597,12 @@ x_positive:
 	c=!(a&0x100);
 	a&=0xff;
 	product[2]=a;			// sta PRODUCT+2		; 3
-	a=product[3];			// lda PRODUCT+3		; 3
-	a+=(~x_i)+c;			// sbc T1+1			; 3
-	c=!(a&0x100);
-	a&=0xff;
-	product[3]=a;			// sta PRODUCT+3		; 3
-					cycles.multiply+=20;
+//	a=product[3];			// lda PRODUCT+3		;
+//	a+=(~x_i)+c;			// sbc T1+1			;
+//	c=!(a&0x100);
+//	a&=0xff;
+//	product[3]=a;			// sta PRODUCT+3		;
+					cycles.multiply+=10;
 y_positive:
 	*z_i=product[2];
 	*z_f=product[1];
diff --git a/tfv/OPTIMIZATION b/tfv/OPTIMIZATION
index 6e6eeff0..e1d75c05 100644
--- a/tfv/OPTIMIZATION
+++ b/tfv/OPTIMIZATION
@@ -162,14 +162,13 @@ Remove unneeded precision in the 8.8 x 8.8 fixed point multiply
 	Cycles: flying=             187
 	Cycles: getkey=              46
 	Cycles: page_flip=           26
-	Cycles: multiply=        44,785
+	Cycles: multiply=        43,588
 	Cycles: mode7=          118,034
 	Cycles: lookup_map=      22,747
 	Cycles: put_sprite=       2,561
 	================================
-	Total =                 188,386
-	Frame Rate = 5.31 fps
-
+	Total =                 187,189
+	Frame Rate = 5.34 fps
 
 Each cycle removed from inner X loop saves
 	32*40=1280 cycles
diff --git a/tfv/tfv_multiply.s b/tfv/tfv_multiply.s
index c969474d..8130db9c 100644
--- a/tfv/tfv_multiply.s
+++ b/tfv/tfv_multiply.s
@@ -312,29 +312,30 @@ multiply:
 	lda	RESULT+2						; 3
 	sbc	NUM2L							; 3
 	sta	RESULT+2						; 3
-	lda	RESULT+3						; 3
-	sbc	NUM2H							; 3
-	sta	RESULT+3						; 3
+;	lda	RESULT+3						;
+;	sbc	NUM2H							;
+;	sta	RESULT+3						;
 								;============
-								;	 19
+								;	 10
 
 x_positive:
 
 	lda	NUM2H		; y_i					; 3
 								;============
 								;	; 6
-	bpl	y_positive						; 3/2nt
+
+	bpl	y_positive						;^3/2nt
 
 
 	sec								; 2
 	lda	RESULT+2						; 3
 	sbc	NUM1L							; 3
 	sta	RESULT+2						; 3
-	lda	RESULT+3						; 3
-	sbc	NUM1H							; 3
-	sta	RESULT+3						; 3
+;	lda	RESULT+3						;
+;	sbc	NUM1H							;
+;	sta	RESULT+3						;
 								;===========
-								;	 19
+								;	 10
 
 y_positive:
 	lda	RESULT+2	;	*z_i=product[2];		; 3