Caption: "Comparison of the measured FitzGerald contraction of a cube versus its visual appearance as given by the Terrell-Penrose effect. The view is from the front of the cube at a distance four times the length of the cube's sides, three-quarters of the way from bottom to top, as projected onto a vertical screen (so that the vertical lines of the cube are initially parallel)." (Slightly edited.)

Features:

1. The left-hand side of the animation shows the lengths that would be actually measured by a true measuring technique. The cube FitzGerald contracts along the direction of horizontal motion of velocity v = (x * c).

2. The FitzGerald contraction formula is

     L(v)=L_0*sqrt(1-v**2/c**2) , where v is the observer's velocity relative an object,
                                  L(v) is the observer's observed length for
                                    the object along the direction of motion,
                                  and
                                  L_0 is the length along the direction 
                                    of motion measured in rest frame of the object.
                                  L_0 is called the
                                    proper length in
                                    Relativityspeak.  

3. Howsoever, what you measure by a true measuring technique is NOT what you see because of the finite vacuum light speed c = 2.99792458*10**8 m/s. There is a differential time delay for the light signals from different parts of the cube.

   If you correct for the differential time delay, then you get a true measurement of length and observe the FitzGerald contraction.

   What you observe sans correction is a rotation, the Terrell-Penrose effect.

   The Terrell-Penrose effect is illustrated by the right-hand side of the animation.

4. The Terrell-Penrose effect was fully worked out independently in 1959 by James Terrell (1923--2009) and Roger Penrose (1931--) (see Wikipedia: Terrell rotation).