Manual
Programming Reference
rg32_float
Ordinary format with two components stored as 32-bit (single-precision) floating-point numbers.