For me, the core is not that you can't "recognize" a face, it's that it is very very hard to form a memory of a face. It's not in the storage, it's in the retr

For me, the core is not that you can't "recognize" a face, it's that it is very very hard to form a memory of a face. It's not in the storage, it's in the retrieval. For example, if I try to picture a person's face, even someone I know well, it's kind of fuzzy. It is hard to describe, but again, it's more like you can't quite remember a face, more than you can't recognize a face. So, I put faces together into mini-shapes. Nose-shape, hair-shape, eye-shape, teeth shape (how I recognize Kathereine Heigl), etc. I can't put together a whole face into memory, but I can group these mini-shapes, features, etc. into a "faceprint" of sorts. The main thing I use though is sound/voice. I can recognize voices to an extremely precise degree, all the time in shows/movies, I can recognize actors in costume, full masks, animated movies, etc. without problem. I am sometimes fooled, actually, by people with similar voices, and I'm always saying "these people sound so much alike they should be cast as twins" to which everyone of course thinks I am crazy. (Orphan Black is an exception because the lead in that show changes her voice with the different characters so it keeps me guessing).