A signal processing apparatus includes a signal transforming unit transforming a first sound signal obtained by dividing a sound signal represented by a time function and by the predetermined time width into a second sound signal represented by a frequency function a calculating unit determining a third sound signal which has a reduced influence of the operation sound by using the second sound signal and a sound signal representing the operation sound a correcting unit performing a correction, by setting a sound signal representing the target sound as a reference signal, on the third sound signal based on the reference signal and a signal inverse transforming unit inverse-transforming a sound signal on which the correction is performed from the sound signal represented by the frequency function into the sound signal represented by the time function.