koobas.hobune.stream
The 'fixed' numbers in shaders: what are the 11 bits used for?
I've just started getting into the magical world of shaders, and this line from the UnityDocs bothers me:
fixed: low precision fixed point. Generally 11 bits, with a range of –2.0 to +2.0 and 1/256th precision.
How are those 11 bits being used? To get 1/256th precision 8 bits must be for the fractional part, 1 bit for the sign, leaving 2 bits for the integer part. However that would make the true possible range [-4.0, 4.0) though! Am I missing something that requires the use of the extra integer bit? Or is this one of those "the precise implementation depends on the platform" things?
The closest thing I can find that makes sense is from here:
The fixed data type corresponds to a native signed fixed-point integers with the range [-2.0,+2.0), sometimes called fx12. This type provides 10 fractional bits of precision.
That makes perfect sense to me. 12 bits: 1 for sign, 1 for integer part, 10 for fractional part. Resulting in the expected range with "1/1024th precision" as the Unity docs would say.
Depends on the platform. The NVidea page is telling you about their exact specs on their best chip. Unity's page is probably giving you a worst-case, for an older cell phone.
My (not extensive) experience is that the specs on a graphics chip are only a very rough guide as to how it really works, anyway.
