1. While data can be converted from integer to floating point without any loss of resolution, the precision of the reverse operation from floating point to integer conversion is not guaranteed and should be avoided unless absolutely necessary. It is strongly recommended that the results of any conversion to integer are stored in a new file, preserving the original data.
2. Always create and use the appropriate integer format values for the parameters ANALOG:GEN_SCALE, ANALOG:SCALE and ANALOG:OFFSET when storing analog data in floating point C3D files. These parameters contain useful information about the original source of the analog samples. Ideally these parameters will contain values that scale the analog data correctly regardless of whether the storage format is integer or floating point format.