What is the difference between the various Q types on the DN9340 Helix Equaliser?

 

The “Q” of an audio equaliser describes the steepness of the filter - the degree to which it will affect signals either side of its nominal or “centre” frequency. In general, the Q of a peaking filter is defined mathematically as , centre frequency / bandwidth where the bandwidth (in Hz) is the range of frequencies affected by the filter. Because the frequency response of such a filter is a smooth curve (not a sharp “brick wall” filter like the ones in an analogue-to-digital converter) we have to decide how we choose to define the bandwidth, and the established convention is that we use the bandwidth to the “-3dB” points on either side of the centre frequency, where the gain is 3dB less than the maximum gain.

 

fig_1 Q.jpg

In the example above, the filter is centred on 1 kHz, the lower 3dB point is at approximately 800 Hz, and the upper one is at approximately 1.25 kHz. This filter therefore has a Q of 1000 / (1250-800)=2.2 In a typical parametric equaliser (and in the case of the Helix system the graphic and dynamic sections too) we have a manual control for the Q of the filter, and this allows us to set any Q that we require. In general high-Q, narrow filters are used for notching out problem frequencies without affecting the programme material too much, while gentler low-Q filters are useful for adjusting the tonal balance. In the case of graphic equalisers there is another issue - that of interaction between adjacent bands. In general, lower-Q filters will blend together more smoothly, but higher-Q filters provide more selective control of problems - at the expense of more frequency response ripple.

So far so simple - but why the different types? This is due to the way in which the Q of the filter varies (or not) when the gain control is adjusted. There are three modes available in the Helix system, which we term Proportional, Constant, and Symmetrical Q.

 

Proportional Q

fig_2_Q.jpg

 

Proportional Q is the mode of operation familiar to users of the KLARK TEKNIK analogue graphic equalisers such as the DN360. As the amount of cut or boost is increased, the Q also increases. This has the effect of making the equaliser “focus” more tightly as the amount of EQ is increased. This allows a fairly low-Q filter at small cut and boost settings, providing gentle control of tonal balance and low ripple. At high gain settings, a proportional-Q equaliser “automatically” increases Q for more dramatic problem solving such as suppression of feedback or unwanted resonances. In the interests of clarity, the Q setting shown on the display is the Q at full cut or boost - the Q at lower gain settings will be lower than that shown on the panel.

 

Constant Q

fig_3_Q.jpg

 

A constant Q equaliser has the same Q at all cut and boost settings. In other words, the bandwidth between the 3dB points does not change at all as the gain is adjusted. The really important thing to notice about this is that the resulting frequency response is NOT symmetrical in cut and boost. This is because of the definition of Q which is based on the 3dB points relative to maximum gain. The maximum gain of the filter when in cut is, of course, 0dB, and the bandwidth is determined by the -3dB points relative to 0dB and NOT relative to the minimum gain (at the centre frequency). This makes a lot of sense musically too - if you listen to a music signal and apply a notch filter, and then change the shape of the curve around the minimum gain (centre) point, it will make little difference to the sound (since that area is already attenuated a lot). However, if you change the curve around the 3dB points, this will affect the sound much more, as more or less of the signal “falls into” the notch. It is this bandwidth that the constant-Q filter is keeping constant. Note that many equalisers that are described as “Constant Q” by their manufacturers do NOT fall into this category, and are what we would term symmetrical-Q designs.

 

Symmetrical Q

fig_4_Q.jpg

 

This class of equaliser has the same curves in boost as the constant-Q type, but then has cut responses that are symmetrical with the boost ones. In other words, the bandwidth in cut is defined not according to our usual definition of Q (see constant-Q above) but as “the point were the signal is cut by 3dB less than the maximum cut”. Most equalisers described by their manufacturers as “Constant Q” in fact produce symmetrical responses.

 

 


Share this page