The 3-D geometry implies that the rays can now have different azimuth
and can propagate out of the vertical plane.
If the migration velocity is correct, the two
rays focus at the same point at zero subsurface offset
(Figure ).
By assuming the velocity is constant around the image point,
all the elements (rays, normal, image points)
are contained in one plane: the plane of the propagation.
By an appropriate change of
coordinates, this 3-D problem with correct migration velocity
can be locally transformed in
a 2-D problem, and the 2-D theory analyzed by
Biondi and Symes (2003) be applied.
The offset-to-angle transformation must be adapted though.

multi_correct_v_2
Figure 1
The velocity function is V(z)=1.5+.5z km/s.
The target has a fixed position, azimuth ()
and dip ().