If you like to define lipid structure/behavior in its natural fluid state, it is better to use cryo-technique. Otherwise you might introduce artifacts, witch is hard to interpreter.
In cryo-TEM you are limited to (1) structures less than micron- size (big particles are blotted out during cryo-specimen preparation); (2) not very viscous fluids, to be able to prepare a film, thin enough for TEM observation.
For bigger structures light microscopy could be satisfying. For viscous systems cryo-SEM or "wet" AFM is an option.