If you read the paper, the answer is yes, but you can work around it by basically breaking the symmetry between the "output" of a system and the "input"[0]. For

If you read the paper, the answer is yes, but you can work around it by basically breaking the symmetry between the "output" of a system and the "input"[0]. For the output and input, they will satisfy basic fourier uncertainty relations, but independently of each other. [0] I've scanned the paper a couple of times and it's still a little murky for me. I have other things to do, unfortunately. For a clue, see figure 2.