It could be the case, that in my own postings I referred to a ‘Scale Factor’, which would come prior to Quantization. In other works of reference, the term ‘Quantization Step’ might appear. As far as I am concerned, these terms are synonymous.
The goal could be to start with a maximum value as input, and to find a way to quantize it and all lower values, to arrive at a maximum quantized integer value. One would divide the (absolute) first by the second input value, to find this parameter, for an interval of time.
Dividing all the sample-values in the interval by the resulting parameter, will yield the maximum, quantized integer. And this parameter will also be equal to the minimum difference between the sample values, leading to two different quantized integers.