It is shown that the watermarking algorithm presented in another paper [Ganic and Eskicioglu, J. Electron. Imaging 14, 043004 (2005)] has a very high probability of a false-positive answer and has its limitations in practice. Furthermore, the intrins
