Twin Spheres Paradox
Assume we have two magic spheres that can pass through each other and we can get inside them. The inertial frame logic of light claims if either of us flash our light, all points on the inside of the sphere will be struck at the same time by the light regardless of our actual motion.
In other words, light expands spherically stationary to the frame from the point of emission in the frame. Now, somehow we put these spheres at the same point in space and I flash the light. Both of us then would see the light strike our sphere points all at the same time.
This time I am going to move very fast relative to the other sphere. However, your sphere looks like it is moving very fast relative to me and I am at rest. At the instant we occupy the same exact space, I flash the light. Light speed is not infinite so it does take time to strike each point on the sphere. Well, since light expands stationary to me it hits my points all at the same time.
But what about you? It cannot hit your points all at the same time and mine also because I am moving relative to you and we no longer occupy the same point in space. But, light from a moving light source is supposed to proceed outward spherically from the point of emission in your frame which is the center of your sphere according to the SR light postulate.
Since one light sphere cannot intersect the two rigid body spheres at all points in each sphere simultaneously, we have a contradiction. In particular, since the center of the one light sphere is centered in each frame, then the center of the one light sphere would have to occupy two different points in space.
Try drawing a picture of the event even using length contraction, it is impossible. Make the figures anyway you like and you still cannot express a geometry/picture that works for the two spheres and one light sphere.
More specifically, SR encourages the use of disjoint frame to frame logic when deducing relative motion. However, given one light sphere and two rigid body frame spheres in relative motion, both frames are forced together to solve the problem and the problem must be solved with all known information. As such the following three conditions must be met with one light sphere.
1) Light must proceed spherically from the center of the "rest" sphere.
2) Light must proceed spherically from the center of the "moving" sphere.
3) Because there exists relative motion, the two centers of the rigid body spheres do not occupy the same point in space.
In fact, by looking at just the x-axis of the spheres, the at rest observer would claim both equal marks at a distance r were struck at the same time, and in addition the moving observer would make the same claim.
However, the relativity of simultaneity suggests if the at rest observer views the event as simultaneous, then an observer moving in collinear relative motion would not judge the same event as simultaneous it its frame. In fact, both observers would agree on the ordinality of events in the moving frame. In particular, in Einstein's train experiment, both observer M and M' judged M' did not see the front and back of the train struck at the same time and both observers agreed the front was struck first. They just disagreed on the timing, not the ordinality.
Yet, both observers view the event above from the x-axis in a simultaneous fashion which contradicts the relativity of simultaneity.
Basically, since SR is unable to produce a set of equations that satisfies all the 3 conditions listed above, then SR's understanding of light is false and it is that simple.