2 ms·

All those annoying trig functions get in the way since they're implemented for floats/doubles.

• veltas 4y ago
  No reason you can't have intermediate int's. And if there's some nearby frame to do calculations against then you don't get the loss of precision.