Here is a dataset coming from Fuse. It is the object: SK-66D172, ID: P117-22, and the segment is LIF 1a:
The extraction consist in calculating the 1D spectrum from this 2D spectrum.
The easiest way to do that is to do what I call a ``classical'' extraction: just sum up all the pixels in a column.
Another way is the ``box'' extraction, which consist in putting a box surrounding the useful part of the 2D spectrum and sum the pixels inside the boundaries of this box.
The third way is to do optimal extraction, which is discussed here.