Filter PDF
Document Details
Uploaded by LawAbidingNeodymium1944
HTW Berlin - University of Applied Sciences
Tags
Summary
This document provides an overview of different filter types used in audio engineering applications. It explains how filters work, including their characteristics and importance in signal processing. The document covers topics such as low-pass, high-pass, band-pass, and band-stop filters, along with their practical significance in audio engineering.
Full Transcript
Filter Filters are not only needed for creative use for frequency adaptation in a production situation, but are the basis for almost all functions in terms of audio engineering or are at least involved in it. In analogue audio technology, they are used to remove noise and not burden systems with ene...
Filter Filters are not only needed for creative use for frequency adaptation in a production situation, but are the basis for almost all functions in terms of audio engineering or are at least involved in it. In analogue audio technology, they are used to remove noise and not burden systems with energy that is irrelevant to human perception. In the digitization filter stages must be used to meet the sampling theorem and thus to allow any digital processing. They are used in almost every process stage and are an irreplaceable tool for digital signal processing. In most cases, the primary task of a filter is the targeted intervention in the amplitude frequency response: It must always be remembered that most filters make an (often unwanted) intervention in the phase frequency response: This always leads to a distortion of the waveform in the passband of the filter. However, this distortion does not always have to be negative, it is often imperceptible. Filter Shapes The 'classic' filter shapes can be divided In practice, it is advisable to into five groups: choose the name which is most easily understandable for the application. So it will usually be easier to understand that one Low Pass / High Cut removes the hum interference with a Low Cut, as if you do this with a High Pass. Notice: The most important distinguishing criterion High Pass / Low Cut for different filters is the steepness of the falling edge. This edge steepness is always specified in audio engineering with the unit: Bandpass dB/octave. Filters are called with different degrees of order. For the rough distinction: Band-Stop / n * 6dB/octave where n specifies the filter order. Band Rejection Allpass If one tries to determine the filter order from a frequency representation, one must first look for the point on the curve at which the level decreased by 3 dB. From this point the edge steepness is determined. The differences in the curves are relating primarily to the passband. Known filter curves are: Butterworth, Tschebyscheff, Cauer, Bessel … They all have advantages and disadvantages in the phase frequency response and the amplitude frequency responses in the blocking and transmission range. Equalizer The most obvious and perhaps most popular application for filters is the equalizer. Here, different, controllable filters interact with each other, giving the user the ability to precisely affect the frequency response. The frequency curves used here are always the result of a complex circuit and can not be realized with passive filters. The filter types most commonly used in equalizers are: Bell Shelf/Shelving Low and High Pass Notch Filter Notch Filter Parameter Depending on how extensive the tax possibilities are one speaks of Semi- or fully parametric filter stages. The general distinction is: filter frequency gain quality (Q) semi parametric variabel variabel fixed fully parametric variabel variabel variabel grafic fixed variabel fixed For bell filters, the 'width' of the boost can often be defined. This value is specified as the filter quality (Q) or bandwidth. Notice: Quality and bandwidth are related, but do not represent the same information. The following applies: Bandwidth: B = upper limit frequency f2 - lower limit frequency f1 Center frequency: Notice: geometric mean, not arithmetic mean Quality: Q = f0/B Bell filters can keep their quality constant via the gain (or dip) (constant Q filter) or narrow or wider with increasing gain (or dip). Notice: The quality also refers to the point 3 dB below the peak value and thus is not easily measurable for small gains. In addition, there is a dissonance in the understanding: A filter with constant quality will affect a larger frequency range with increasing gain. A variable-Q filter affects the same frequency range for each gain. Constant Quality Variable Quality The analog circuit diagrams of the simplest filter forms: Low Pass / High Cut High Pass / Low Cut Bandpass Band-Stop / Band Rejection These analog filter shapes can also be used in the digital world. In principle, all known analog filters can be implemented digitally. Conversely, not all digital filters can be realized in the analogue field. Bildnachweise: Seite 1: Handbuch der Tonstudiotechnik, Band 1 Dickreiter et.al., Saur 2008 Seite 2 & 3: Studio Magazin, Friedemann Kootz Ausgabe 08/09 Seite 4: Handbuch der Tonstudiotechnik, Band 1 Dickreiter et.al., Saur 2008 Seite 5: Studio Magazin, Friedemann Kootz Ausgabe 08/09 Seite 6: Handbuch der Tonstudiotechnik, Band 1 Dickreiter et.al., Saur 2008 Seite 7 & 8: Studio Magazin, Friedemann Kootz Ausgabe 09/09 Seite 9 & 10: Wikipedia http://de.wikipedia.org/wiki/Transienten