Hello everyone, I've got a simple question.

I am studying about DACs, actually I am working at one. So every book I get brings the same definition about INL, - "the difference between the actual output value and a straight line between the first and last point, or the best-fit line (this one is rarely used at the books), after gain and offset errors have been removed"

but on practical examples, they always just make the difference between the ideal values and actual values outputs, after removing offset and gain errors. i can't see how it fits the definitions above.

could anyone please give me a light?

i guess it all comes down to the definition of gain and offset that are used.