Do magnets redshift light? Do magnets redshift light?   
Suppose we have an extremely powerful magnet (say the size of the Sun) and we have a smaller paramagnetic  material above it (say. Titanium Brick which is indestructible) . Due to magnetic attraction it will fall toward the big magnet. 
Now suppose that when it hits the surface of the bigger magnet, the smaller paramagnetic material (Titanium Brick) is somehow converted into photons (light) equivalent to its mass using $E=mc^2$, plus its kinetic energy, $KE_1$. These photons are shot upwards via an perfect mirror perpendicular to the paramagnetic material above. When they reach the magnet's original position they are again converted into mass in the form of the titanium brick.
If this is repeated many times it seems we will gain free energy, since the iron brick gains kinetic energy when it falls, but light is not effected by magnetic fields. (Or am I wrong?) But this will breach the energy conservation laws in specific "Energy cannot be created nor destroyed". 
Therefore I must conclude light is (or should be) redshifted (loses energy) as it travels through magnetic field. The problem arises when you take into consideration that light is not charged so it should not be effected by any magnetic fields but my intuition tells me light should still be affected so is my conclusion valid or not? Is light red shifted in presence of an magnetic field? 
If light is not red shifted then what is lost to ensure the system is not gaining energy but in this experiment it does not make sense since we are using photons which are not effected by magnetic fields and then recreating the para magnetic material (like Iron brick we used) so we should not loose energy as we are simply recreating the same object again. 
 A: I love the question! I wanted to share my thoughts and a possible explanation of this phenomena.
Lets look at the process(we'll start from a non-magnetized object):


*

*Two bodies, one of which is a huge magnet and the other a very small ferromagnetic object are in free space.

*No forces act upon them in the starting moment (gravity is unneded). The ferromagnet is not yet magnetised(the magnetic domains are still random)

*We assume a swift magnetisation of the small object, they begin to attract each other magnetically. Energy used up can be calculated from the magnetic energy/coenergy in the hysteresis loop of the material. $W_m^*=\int_V\int_H B\mathrm{d}H\mathrm{d}V $

*The actual force that attracts them magnetically is the Relucance force, this force wants to reduce the total magnetic resistance of a path and moves two objects into a position of minimal resistance for the magnetic flux.

*The attraction process happens because of the same energy/coenergy mechanism. A force results from the change of coenergy $F=-\frac{\mathrm{d}W_m^*}{\mathrm{d}x}$. 

*As the small object gets closer the value of $W_m^*$ is decreased (you must consider the volume). Intuitive explanation: everything wants to get to the lowest energy state. 

*The increase in the kinetic Energy happens in parallel $E_k=\int_xF\mathrm{d}x$, therefore we can look on the initial value of $W_m^*$ as the potential energy.

*In the moment the small object is converted to photons with $E=mc^2$ the $W_m^*$ ist mostly (depending on the moment) converted to  $E_k$.


If I didn't make any mistakes, there should be no energy loss. The big magnet supplies the energy
A: Remember that in addition to the pure mass of your smaller "test" magnet, you need to create it's energetically non-trivial magnetic structure. This should account for the energy gained through the electromagnetic attraction.
Barring non-linear quantum optics effects light is not influenced by magnetic fields. This is because the equations of motion are linear, therefore one electromagnetic field can not influence another without the help of charged matter.
A: There are a lot of things wrong with this thought experiment, imo:

*

*a piece of ferromagnetic material will immediately display a north and a south pole to the magnet lines. The north pole will be attracted the south will be repelled, and it will end up like iron filings when it hits the surface.


*in the real world if it hits a surface it will break up into its component parts, depending on the energy. To get annihilation the energy should be huge  and there will be many fragments atoms and photons


*with the break up the ferromagnetism disappears and the individual particles will be bouncing off according to the rules of particles scattering ( look at crossections in the particle data book)


*These will have every which angular distribution ( photons too) and will fly off according to their momentum and energy given by the big bounce. They will never cohere enough to fall back to any structure called solid matter to envisage it ending up at its previous spot whole . Energy is conserved in elementary particle interactions which means if you gather all the bits you will find the rest mass and kinetic energies all there.


*Supposing the big magnet was made of antimatter, again the same problem of not reconstructability: annihilation of protons and neutrons gives mostly pions , the charged  decay into muons and neutrinos and pi0 into two photons, that will never join up at the top, both because of angular distributions and because the crossection of gamma gamma to pi0 is tiny ( higher order diagrams and the electromagnetic coupling is 1/137 ). Nothing annihilates into just photons, and even if it did, the photons would disperse worse than a bagful of cats.
