In the previous chapter, the f-x prediction technique and the t-x prediction technique were compared and contrasted. It was shown that an f-x prediction is equivalent to a t-x prediction with a very long filter length in time. This filter, with its long length in time, generates spurious events from parallel events embedded in noise on the input. Even with a short filter length in time, the wavelet of events may be corrupted, since a wavelet on an event appears as a series of parallel events. This chapter compares the action of these two prediction techniques, and shows examples of spurious event generation for both widely spaced events and for wavelets.