Forum

How to enter a full 32bit float value?

just downloaded kalles e, PI, golden section constants modules. the modules contain only a IOBox (Value), which holds the value, but only with 5 digits.

how can i enter the full 32 float value ?
and how i can be sure, that the value is sended and used by vvvv correct ?

thx,
tf

The displayed value in the iobox and in the tooltip is always rounded. This is basically stupid and will change. someday. The value itself should be untouched and you can display the stored value with FormatValue or by some appropriate mapping. you may also like to inspect the v4p file.

Internally vvvv uses ieee double floats, so precision basically shouldnt be a problem with nonfractal math. i admit there might be some bugs still floating around in the code, which set the 8087 control word sometimes to single, or pretend that 1.0 needs to be represented as 1.00000001123 or something.

If you are into setting the mantissa and exponent as bits, use Ord, which has a mode for setting the ieee floats directly.