Just to confirm your information, despite the information given in the online help ("Multiplier and offset for the data contents. These may be positive or negative floating point values."), this means that the scale factor can only be an integer and not a float value? I mean that I won't be able to have a scale like 0,1 or 0,05?
This was my original question as a 16 bits value (or 8 bits) defined as an integer is not the same that a 16 bits floating point value (splitted as a sign, exponent and mantissa and with different formats if not IEEE)...
the manual is correct, you can also use "floating point values" for the scale and offset.
But you must it type in in the normal decimal format, it´s not possible to type in splitted in sign, exponent and mantissa.