In Conclusion together, and presents them to an analyzer, traditionally a 1/3-octave analyzer. The notion that this simple process can reliably predict what is perceived by two ears and a brain is preposterous. Using this information as a basis for equalization at middle and high frequencies compounds the error. Some elaborate equalizers make time-windowed measurements attempting to separate the direct and subsequent reflected sounds. This is a thoughtful move in the right direction, but the measurements are blind to direction: they have no idea where the sound is coming from, but the ear-brain system does. It also sacrifices frequency resolution to see into events in the time domain, meaning that the more precisely the sounds are separated in time, the less information we have about them (similar to what is shown in Figure 13.23). If we had detailed measurements on the loudspeakers to begin with, much of this would be unnecessary. Equalization can change frequency response--that is all. As can be seen in much data shown in Chapter 18, loudspeakers can have many problems that are not revealed in room curves, and they can have directivity problems that can show up in room curves but that equalization cannot address (Figure 18.10). The only cure for a loudspeaker with directivity issues is to take it back to the 519