In a device designed to automatically change audio levels (i.e. a compressor, gate, de-esser, limiter, expander, etc.) the detector is the circuit that senses the current audio level and sends control signals to a gain cell to tell it when and how it needs to act on the signal. Take a compressor for example. In order for it to know when to lower the level of the signal, it needs to take constant measurements of how high the signal is and compare those to the threshold. When the signal level goes beyond the set threshold the detector sends a voltage to an amp (gain cell, VCA – whatever you want to call it) that causes it to ‘turn down.’ This is the essence of how most compressors work. Since the detector is a distinct function inside these units it is possible to apply a signal to them that is different from the signal passing through the audio section of the device. This enables the user to change the characteristics or sensitivity of the detector and thus change how it acts on the main audio signal. A de-esser uses this principle. The audio signal is multed (split) and frequencies relating to “s” sounds are boosted on the portion going to the detector. This effectively makes it more sensitive to “s” sounds, and thus causes the compressor to act more sensitively to them than the rest of the signal. The result is reduced sibilance without actually changing the EQ of the main signal. Frequency-selective gates work on the same principle. In many dynamics processors the detector can be accessed separately through a unique input, which makes it possible to design your own custom processor. There are thousands of tricks you can do with these.