There are some problems with any half-float (16-bit) values; they're not read correctly.
A coin should not look like this...
but rather like this:
I took NumPy's implementation of reading HalfFloats. It's basically what is described in Wikipedia; I couldn't check your code yet, but there seems to be a bug in it, as it happens with any such attribute formats.
A coin should not look like this...
but rather like this:
I took NumPy's implementation of reading HalfFloats. It's basically what is described in Wikipedia; I couldn't check your code yet, but there seems to be a bug in it, as it happens with any such attribute formats.