Surfer
asked on
Float conversion to 4 bytes?
I need to comvert 1 float into 4 bytes and then back to 1 float again. Someone told me to try this:
FLOAT f1 = 123.456, f2 = 0.0 ;
BYTE byByte1 = HIBYTE( HIWORD( f1 ) ) ;
BYTE byByte2 = LOBYTE( HIWORD( f1 ) ) ;
BYTE byByte3 = HIBYTE( LOWORD( f1 ) ) ;
BYTE byByte4 = LOBYTE( LOWORD( f1 ) ) ;
//rebuild
f2 = byByte1 << 24 | byByte2 << 16 | byByte3 << 8 | byByte4 ;
The rebuild answer for this is f2 = 123.000 and not 123.456. How can a get the rest of my float? What am I doing wrong? Also what is required to break down to bytes, a double (8bytes) and rebuild the 8 bytes into one double? All sugestions welcome and sample code is greatly appreciated!
FLOAT f1 = 123.456, f2 = 0.0 ;
BYTE byByte1 = HIBYTE( HIWORD( f1 ) ) ;
BYTE byByte2 = LOBYTE( HIWORD( f1 ) ) ;
BYTE byByte3 = HIBYTE( LOWORD( f1 ) ) ;
BYTE byByte4 = LOBYTE( LOWORD( f1 ) ) ;
//rebuild
f2 = byByte1 << 24 | byByte2 << 16 | byByte3 << 8 | byByte4 ;
The rebuild answer for this is f2 = 123.000 and not 123.456. How can a get the rest of my float? What am I doing wrong? Also what is required to break down to bytes, a double (8bytes) and rebuild the 8 bytes into one double? All sugestions welcome and sample code is greatly appreciated!
I don't understand what you want to do.
variable (as you probably know) of type float takes 4 bytes in memory. So if you need access to each byte of them:
BYTE *p = (BYTE*)&f1;
p[0] - first byte and so on.
the code in your question doesn't work because bit operation OR can't return float type.
may be something like this
FLOAT f1 = 123.456, *f2;
int *pf, i2;
f2 = new FLOAT;
pf = (int*)&f1;
BYTE byByte1 = HIBYTE( HIWORD( *pf ) ) ;
BYTE byByte2 = LOBYTE( HIWORD( *pf ) ) ;
BYTE byByte3 = HIBYTE( LOWORD( *pf ) ) ;
BYTE byByte4 = LOBYTE( LOWORD( *pf ) ) ;
//rebuild
i2 = byByte1 << 24 | byByte2 << 16 | byByte3 << 8 | byByte4;
memcpy((void*)f2, (void*)&i2, 4);
variable (as you probably know) of type float takes 4 bytes in memory. So if you need access to each byte of them:
BYTE *p = (BYTE*)&f1;
p[0] - first byte and so on.
the code in your question doesn't work because bit operation OR can't return float type.
may be something like this
FLOAT f1 = 123.456, *f2;
int *pf, i2;
f2 = new FLOAT;
pf = (int*)&f1;
BYTE byByte1 = HIBYTE( HIWORD( *pf ) ) ;
BYTE byByte2 = LOBYTE( HIWORD( *pf ) ) ;
BYTE byByte3 = HIBYTE( LOWORD( *pf ) ) ;
BYTE byByte4 = LOBYTE( LOWORD( *pf ) ) ;
//rebuild
i2 = byByte1 << 24 | byByte2 << 16 | byByte3 << 8 | byByte4;
memcpy((void*)f2, (void*)&i2, 4);
ASKER
I tried to modify your code sample that works with type float to work with type double and it dose not work. All suggestions welcome.
double r1 = 123.456, *r2 ;
r2 = new double ;
_int64 i2 ;
BYTE *pr = (BYTE*)&r1 ;
BYTE byByte1 = pr[7] ;
BYTE byByte2 = pr[6] ;
BYTE byByte3 = pr[5] ;
BYTE byByte4 = pr[4] ;
BYTE byByte5 = pr[3] ;
BYTE byByte6 = pr[2] ;
BYTE byByte7 = pr[1] ;
BYTE byByte8 = pr[0] ;
i2 = byByte1 << 56 | byByte2 << 48 | byByte3 << 40 | byByte4 << 32 | byByte5 << 24 | byByte6 << 16 | byByte7 << 8 | byByte8 ;
memcpy((void*)r2,(void*)&i
delete r2 ;
double r1 = 123.456, *r2 ;
r2 = new double ;
_int64 i2 ;
BYTE *pr = (BYTE*)&r1 ;
BYTE byByte1 = pr[7] ;
BYTE byByte2 = pr[6] ;
BYTE byByte3 = pr[5] ;
BYTE byByte4 = pr[4] ;
BYTE byByte5 = pr[3] ;
BYTE byByte6 = pr[2] ;
BYTE byByte7 = pr[1] ;
BYTE byByte8 = pr[0] ;
i2 = byByte1 << 56 | byByte2 << 48 | byByte3 << 40 | byByte4 << 32 | byByte5 << 24 | byByte6 << 16 | byByte7 << 8 | byByte8 ;
memcpy((void*)r2,(void*)&i
delete r2 ;
Your problem is due to the way floats are stored internally - IEEE format.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER