In the second minimum (the 3rd step) there is a smaller decrease in light intensity. For this to happen, wouldn't you need to be looking at the plane of orbit from above rather than directly along the plane of orbit? Because if a person was looking directly along the plane, then it would just be a full eclipse of the star behind in step 3?
But then if a person is not directly looking along the plane but just sees the plane from above, doesn't this mean the definition of eclipsing binaries is wrong? The definition is 'when the plane of orbit of stars is in the line of sight from the Earth'. By 'line of sight' I assumed it meant a person from Earth is able to look directly along the plane. So was my assumption wrong?