To put it simply, in general, you need three signals protecting each junction. One at the head of the point, facing the forks, and one on each fork facing the head. With these in place, train 1 will see the red signal at the head of the point that is in front of train 2, and release the junction, behind it, allowing train 2 to pass, and vice versa. The first of the two trains to do this will continue on to the next passing loop. The stretch of track between the loops becomes a block on its own, which is empty, allowing one train into it. Right now, both trains see it as available, and try to lock the point at the other end.