ICEinLAVA's Domain

In last month’s column, we took a look at the two most common applications of dynamics: compression and limiting. This segment will deal with some familiar dynamic types and some more esoteric varieties as well as specifying applications for each.

Before getting in to the nitty-gritty of this month’s focus, let’s take a closer look at a typical dynamic “engine”. Fig 1 shows a typical VCA-based dynamic processors block diagram. Take note of the side chain path, which examines the input signal and generates the control voltage for the VCA based on the settings of the threshold, ratio, attack and release controls. In most dynamic processors, this path is normalized to the input, but can be broken and fed by an alternate signal, often referred to as the external input, or key input. Although only occasionally used on compression and limiting, these inputs are more often used when gating or expanding to produce special effects which will be covered later.

Gating and Expansion

With compression and limiting, we are dealing with over-threshold conditions, in gating and expansion, we work below the threshold. That is to say, below the threshold setting, the gain of the signal is adjusted by the processor, while above, the gain is linear or unity with the input signal. With expansion, the ratio can be varied from 1:1.2 up to 1:50. Employing low ratios can give a track more dynamic “life” allowing the dynamic peaks to jump out a bit more. Additionally, expansion can be used to remove the effects of too much compression on a track. Gating is similar to expansion, but implies the use of a fixed, high ratio achieving more of an audio “switch” opening and closing, rather than the smooth, continuously variable transitions found on the expander.

Ducking (the reverse of gating) is done by causing the gate to activate above-threshold to a side-chain-signal. A common application for a ducker is in a paging system where the voice signal is sent to the side-chain input of a noise gate, in addition to having a path to the mix bus. The program, usually background music, is processed through the gate. The presence of the voice mic at the input of the ducker triggers the dynamic, which can then be set to reduce the program gain by a preset amount, say 3 to 6 dB. The effect is of the background music lowering in volume when the voice mic kicks in. Similar set-ups have been used in live sound to ensure that the lead vocal mic stays above the mix, no matter what. For more side-chain and special effect applications, see below.

Leveling is achieved by taking a high ratio and low threshold and using slow attack and release constants. This way, the processor is almost always in gain reduction, with plenty of gain to give up if the level from one track to the other varies. However, due to the slow speed of the time constants, any fast change will get through, only the slow change of overall level, like that of one CD changing to another, with cause the leveler to give back a little gain to gently, and transparently keep the overall gain around the 0 VU point. Sometimes this process is called AGC, and may employ some other specialized controls such as, target output, gate, max gain, or floor threshold. Fig 2 shows the control settings for a leveler set-up on a compressor.

Dynamic EQ

One of the more unique applications of dynamics, and most powerful to the FOH mixer is dynamic EQ. This process allows compression or expansion within a specific frequency or EQ band. That might not sound like much, but this is process is often the “Swiss Army Knife” of the touring or recording engineer. For instance, an example of dynamic EQ as applied to a vocal microphone would be to tune the low shelf filter to 100 Hz, set the DEQ to compress and a medium-to-high threshold. This will take care of proximity effect and popping while the artist is up close to the microphone, but will allow the mic to perform “flat” when the performer backs off. This will enable the microphone to sound more natural within a greater performance distance.

Another great trick for dynamic EQ is sibilance control or de-essing. Set the high filter for the sibilance frequency; usually between 4 and 8 KHz. Again, setting the dynamic controls for compress at a high threshold will effectively “slice out” the offending freq, while leaving the frequencies around intact and everything below the threshold is, of course, unaffected.

As an all around microphone processor for horns, dynamic EQ can quickly adjust the frequencies when a sax player decides to “go down” on a mic.

Using expansion and dynamic EQ is another trick entirely. If compression to dynamic EQ is like subtracting or tuning down the frequency of the EQ, then expansion is the additive process. Ever want to bring out the breathy whisper of a vocalist during a soft dynamic, but were afraid to use EQ because of the scream in the next chorus? How about bringing out the pick and the top sound of a delicate acoustic guitar part, only to have to back it off when the heavy strumming comes in? Well, take your handy dynamic EQ and set to expand below threshold, set the filters to enhance the frequency range you wish, then, by adjusting the threshold to the level of the loud section, the EQ is removed.

On a sound system, dynamic EQ can increase the clarity and the headroom of a horn-loaded PA system. Every compression driver has an overload point, where the diaphragm no longer acts as a piston, but the frequencies “wave” the membrane. This is called the break-up mode. It is often the first harshness heard as a PA is pushed to the upper limits. Dynamic EQ can be tuned in to this primary break-up frequency (usually between 2 and 4 KHz) and forestall the onset of the break-up by gradually tuning out the resonant frequency. Fig 3 shows a functional block diagram of a dynamic EQ unit.

Some interesting processing can be achieved when applying a side-chain signal to control a dynamic processor. The most common form of side-chain processing used today might be the application of filters, either a low and high pass pair or a sweepable parametric, to a noise gate. These “frequency conscious” gates are found to be more accurate in triggering to open off the dominate frequency off the program, by eliminating the inevitable leakage from other instruments. Sending a mult of the drum mix channel to the key input of a gate, and processing the reverb return with that same gate gave us that famous Phil Collins non-linear reverb sound. Take that same processing path, but insert a compressor in place of the gate, and you can modify the attack of the reverb envelope to be triggered as in the snare hits on Joe Walsh’s “Rocky Mountain Way”. Still others have linked kick drum control signals to gate the bass guitar to produce a drum/bass combination so tight, the strongest solvents couldn’t pull them apart.