FIB is best way (and maybe only way), but all depends on what kind of substrate or/and material you are thinking of? If both are conductive the simplest way is to etch it with electrolyte and low current (the best stable current power source). Usually that way will give hole (from one side of sample to another) but near the hole you will have thick enough structure for (even low voltage) TEM observations.
Anyway, there a lot of substrates available so maybe some solutions of the producers may be enough?
Need more information on the film and substrate. General principle is taht the specimen has to be transparent to electrons ( 100 nm or so, depending on material). If the substrate can be selectively detached/ dissolved, and the film can be floated, then it is the easiest. Normally, that is what is done. If detaching is not possible, thinning is the only way.