High pass, low pass, and band pass filters are very useful in the field
of Digital Signal Processing (DSP). These filters remove specific portions of a
given wave length of data, and are therefore named accordingly to indicate their
purpose. For example, a low pass filter will allow only frequencies within the low
range to pass through it, thereby reducing or removing high frequencies, while a
high pass filter will do the opposite. A band pass filter allows a specific range
from a given wave length, known as a band, to pass through it. The range used for
a band pass filter can include any cross-section of the wave length's available
frequencies.
All of these filters are also useful outside of the DSP field of interest. One noted use of these types of filters is in financial charting, where filters are commonly used to remove short term value fluctuations in market segments, and stock values. Removing insignificant data points aids in easily identifying long term market trends. By reversing the above mentioned filtering process, you can quickly identify, or focus on, specific short term market trends. One example showing the effective use of filtering is when a high pass filter, with a low cutoff frequency, is used to remove trends and seasonal variations from raw market data. This approach effectively "cleans" the data. When you complement this approach with a low pass filter, you will also be able to isolate trends and remove the long term cycles that a high pass filter would inherently show.
For more information on how to use the digital signal processing utility classes, including source code and a sample application, please visit our support site.