Microphone characteristics, harmonics from multiple voices and reverberations in the room can be compared nomatter how garbage the recording is. Even differences in humidity on different days can be picked up. It would be straight forward for an expert analyst to confirm it wasnt spliced.
Of course, these days all those characteristics can all be faked digitally by an expert
You can only do so much from a 1960s surveillance recording. No one is going to spend the money to analyze it anyway. This isn’t the JFK assassination where they struggled to just identify how many shots were fired from some crappy audio file.
Splice? No splice. As I said one could mix a recording of MLK laughing with a recordings of the so-called rape and it could be done such that no one could definitely prove it was doctored. Different experts could state opposite conclusions and they would just be opinions. An MLK audio tape that is damaging to MLK’s reputation is not going to be generally believed with experts disagreeing on conclusions. Analyzed by whom? The FBI? :-)