Coming from an audio background, this explanation clicked immediately. You can't determine anything about frequency from a single audio sample. And conversely, you can't determine anything about signal phase from a frequency spectrum.

Is this just an apt analogy or is it basically the same concept at work in QM?