Fix lerp function

For avoiding floating point errors it was being used an inverse lerp function. The new one still benefits from being precise while not needing further inversions / a change in the parameters order