multi-path-interference
Short delay multi-path – This is always caused by something directly in front of the antenna. One common cause is a tree in front of the antenna. There will be chaotically overlapped signals behind a tree. This will mainly affect UHF reception. The solution is to relocate the antenna (or cut down the tree). If the antenna stays behind the tree, you will likely see dropouts on UHF channels when the wind blows. And that’s for strong-signal areas. In weak-signal areas you will likely get no UHF reception at all behind a tree.
The other common cause is an irregular horizon line (structures and trees in the distance). These will cause overlapping fields, which will result in a regular pattern of strong and weak spots. For UHF, moving the antenna right or left three feet or so can make a huge difference. Moving the antenna is usually the solution. Unfortunately a strong spot for one channel can be a weak spot for a different channel. The same phenomenon happens for VHF, but since wavelengths are ten times as big, the strong and weak spots are ten time further apart, so moving the antenna to a strong spot is often too far to be practical. For VHF the solution is usually a bigger antenna. (The author faces a severe case of horizon-induced multi-path. His UHF strong spots are always about 12 feet apart, but they can move with the weather. His solution: He erected two UHF antennas 6 feet apart, selected by a switch. Whenever one antenna is in a weak spot, the other is guaranteed to be in a strong spot, so at least one antenna always works.)
Long delay multi-path – This is caused by a large obstruction like a hill or a large building. If you tune in an analog channel close in frequency and from the same direction, you will see ghost images. The offending signals are approaching the antenna mostly from the sides, but also in rare cases from the rear. Actually all analog images have these ghosts, but without the direct path blocked they are too dim to see.
One solution is to move the antenna to a new spot where signals from the offending directions are less strong. A move of 20 feet or more will likely be necessary.
The other solution is to select an antenna with little or no reception in the offending directions. There are two workable strategies here. If the analog channels show one really strong ghost (and maybe a number of weaker ones) then selecting an antenna with a null will work. See “Nulls in radiation pattern”. Otherwise an antenna with minimum radiation to the side and rear is the way to go. The higher the antenna’s gain, the less reception it will have to the side.