Yes, you certainly can. This training method is based on spotting the difference between two audio samples, so the listening environment is not that critical.