[mamon]

I think this is because our memories are not videos, we do not store them pixel by pixel, because that would require huge amount of storage space our brains do not have. Instead we use the approach similar to 3d games: remember rough shape of an object as one piece of information, location as second piece, and color or texture as a third. When recalling a memory your brain renders the whole scene, and if some piece of infomation is missing it makes it up on the fly to fill the gap. Then you overwrite your original memory with memory of this most recent rendering.