A major problem with testing temporal resolution is that when we make temporal changes in a frequency-specific sound, we also make spectral changes. For example, a long duration tone has a single peak in its spectrum, but when we make the same sound short in duration by turning it on and off abruptly, sound energy spreads to adjacent frequencies . This phenomenon is called Òspectral splatterÓ. The problem is that a listener could tell, for example, that the duration of the sound changed by detecting the change in the spectrum, not the time waveform.