The output is a binned dataset of the same dimensionality as the input file, with the spline fit instead of data. There are no bad quality points in the output, all have been fitted by the spline. The positions of the knots, and the values of their coefficients are tabulated in the MORE structure of the file.
Subtopics: