.net - C# float max parse error -

i declared byte array contains 4 bytes.

byte[] bts = new byte[] { 0xff, 0xff, 0x7f, 0x7f };  float f1 = bitconverter.tosingle(bts, 0);  string s = f1.tostring();  float f2 = float.parse(s);  byte[] bts2 = bitconverter.getbytes(f2); 

after conversion, realized output changes

{ 0xff, 0xff, 0x7f, 0x7f } 

to

{ 0xfd, 0xff, 0x7f, 0x7f } 

why did happen?

if @ f1 , f1.tostring() in watch window see that

f1 = 3.40282347e+38

f1.tostring() = 3.402823e+38

which means tostring method outputs string representing trimmed number , not accurate 4 byte float, but why?

the default float.tostring uses the g specifier in standard numeric format strings has 7 digit accuracy floats.

you can use other overload of tostring specify format specifier , provide amount of digits want represent:

string s = f1.tostring("g10"); 

a better way use "r" specifier lossless

string s = f1.tostring("r"); 

msdn: round-trip ("r") format specifier used ensure numeric value converted string parsed same numeric value.