Tweeter versus horn; that is the question?
“I hear 2 tweeters; the first one is a piezo, and the other one is a compression driver. They sound different even though the frequency response is the same (1,800 Hz, 20,000 Hz). Why?”
Aside from the fact that no two devices are going to sound the same anyway, there are fundamental differences between a tweeter and a compression driver that make them sound different. A tweeter radiates sound more or less directly into the air. There is usually little to control the dispersion of a tweeter other than the front surface of the cabinet in which they are mounted. With careful design tweeters can sound very nice and pretty, but their behavior and polar pattern can be inconsistent. It can also take a lot of power, which eventually translates to size, to make a tweeter loud enough to keep up with larger speakers. Where speakers can be redundant (i.e. more than one speaker can cover a frequency range), the anomalies in tweeter dispersion and frequency response can cause objectionable phase cancellation and comb filtering when multiple devices are used together.
Compression drivers, on the other hand, can have very controllable polar patterns and frequency response characteristics depending upon the type of horn they are coupled to. In addition, the compression driver makes the device much more efficient, so it takes less power to reach a certain SPL. Horns can also be used in multiples due to their controllability. Horns and compression drivers are used in virtually all high power PA applications for these reasons. It’s just not practical to use tweeters, especially in the critical 1k to 6k range. Above that tweeters are sometimes used. One issue designers of compression drivers must address is the fact that sound from all parts of the diaphragm must reach the throat of the horn at the same time (in order to be in phase with each other. A device known as a phasing plug, which has a series of different length passages through it is often used to accomplish this, but it’s never an exact science. There are side effects to this as well. Some compression drivers develop so much pressure at the throat of the horn that nonlinearities in the compression of the air are introduced. These can cause significant harmonic distortion and is partly responsible for the objectionable sound of some high power horn systems.
While any high power PA application is more surely going to call for horns and compression drivers, studio monitors and stereo speakers tend to rely more on tweeters. They are much less expensive to build and to most people’s ears sound better at the typical close listening distances used in near field monitoring. Do not assume, however, that tweeters are “better” than horns. Some of the most expensive and accurate midfield studio monitors in the world use horns and drivers. They are just designed with excruciatingly careful detail.