ICEinLAVA's Domain

One of the tricks a live sound person has to make the mix sound better than what is coming off the stage is the selective use of compression. However, let me jump up on my soapbox to preach that compressors are not a cure-all for other production and performance deficiencies. The thing that irks me the most about newbies in this business is that they feel the need to have at least one dual-channel compressor, just because they see the “pros” using them in their FOH racks. Of course, not knowing where a compressor is inserted, they throw it across the main mix and “tweak it in.”

What entry-level sound persons need to remember is that compressors come last on the list of procurement priorities. If you wasted $200 on some cheap compressors, that money could have bought you a larger power amp or better speakers that could really make your production stand out. Living with inadequate racks and stacks and squashing your mix with a compressor to make up for the inadequacy is something that infuriates me. I am now stepping off my soapbox.

Compressor Guts

Before we jump into where and how to use compressors, let us review the internal function and how the controls work. Compressors have two paths for the signal once buffered at the input. The main signal path is pushed into a Voltage Controlled Amplifier/Attenuator (VCA) stage. The controlling voltage for the VCA comes from second path, or “sidechain” that takes the input signal, converts it to a logarithmic representation (dBs), threshold detects the signal, and gains the signal to the VCA to amplify or attenuate control inputs. Audio signals exiting the VCA are fed to a last stage for “make-up gain” if the user desires. Main compressor controls such as the threshold and compression ratio work upon the sidechain signal to set the points of compression onset and the rate of compression.

Additional controls like attack and release work on the sidechain signals between the threshold and ratio controls to place smoothing on the signal as it passes the threshold and returns from the signal peak amplitudes. Many compressor manufacturers offer an auto-attack/release feature that continually adjusts the attack and release time constants based on the rate of change of the incoming sidechain signal. If you are not looking for a specific attack or release time constant, or just do not know the best settings, consider switching in the auto-attack/release function.

Another optional control that sometimes appears on compressors is the gate/expander control. Mostly used for recording, the gate expander is just another dynamics alteration in which low-level buzzes, hums or non-musical noises can be un-amplified until gating signal amplitude (threshold) has been reached. I personally leave the gates off for live sound use, due to my own uncertainty that I will remove the set gating at the end of the gig, and at the next gig I will be in the midst of level-checks and won’t find the channel’s signal until I have dialed in way too much gain. These sudden blasts of channel levels are unwelcome to my PFL headphones and my ears, and I can live with a little more noise floor to prevent the occurrence.

Settings

The majority of compressor channels are used as channel inserts for key signal types with sometimes nasty dynamic signal levels—mostly inputs like vocals, bass guitar, keyboards, brass instruments and some percussion. If you use insert compression, you may want to bypass the compressor or raise the compression threshold way up when doing line checks. Also, you want to set the channels a bit lower than your normal levels, such as zero dB for average levels. The rationale for this is the nominal +20dBu clip level at the channel preamp that gives you the 20dB gain reduction headroom before clipping. For example, a 4:1 compressor ratio at a zero dBu threshold brings the maximum compressor output level to +5dBu before clipping.

Compressor make-up gain settings on insert channel compression should be used sparingly, if at all. Gain settings from zero dB to +3dB are typical and prudent. The rationale for this is occasionally you may have to un-insert or bypass the insert jack signal processing, and the sudden drop in level may not be something you can tolerate in mid-performance. Generally your channel fader is your best tool for channel level adjustments. If you find yourself often tweaking preamp gains or compressor makeup gains, you probably have a gain structure problem best addressed with channel, subgroup and master faders.

Vocals

The biggest payoff in compressor usage is with insert compression on lead vocals. In general, the compression ratio for live performance lead vocals varies between 2:1 and 4:1, with the threshold set for about six to 10dB gain reduction on the peak vocal signals. Most kinder, gentler music genres err towards the 2:1 side, with rock vocals toward 3:1 or 4:1. If your compressor has the hard-knee/soft-knee threshold switch, choose the soft-knee for vocals. The effect of compression on vocals is best described as a “more professional sound,” due in part to the fact that most of us hear plenty of compressed vocals on television and radio broadcasts and recorded music.

Backing vocal compression is nice touch, especially if you have the luxury of plenty of compressor channels. You can use the same kinds of settings as lead vocalists and the compression helps keep the backing vocals from crowding the lead vocal. In a pinch, some compressor-poor sound persons will combine all the backing vocals to a subgroup, and use a subgroup insert to compress the vocals together. This can work well if the backing vocalists live on the same monitor mix or are bunched together to hear each other’s vocal levels and self-adjust their voices. Where things can get vicious is when sub-group compressed vocals get dominated by one singer, who then causes the compressor to focus on that voice, thus squashing the other singers further down into the mix.

Inserting compression on bass guitar direct-inject (DI) signals is the natural next choice on compressor channel allocation. Much depends on the style of the bassist. For example, jazz bassists may not need any squashing at all; and on the other side, a thumb slappin’ and string pullin’ bassist may never get enough dynamics control. Bass guitar signals typically get a 5:1 to 10:1 ratio and gain reduction in the 10 to 15dB range. This is especially handy to smooth up less talented bassists or bassists who alternate from plucking to slapping technique. Here a hard-knee threshold setting is appropriate.

Some keyboards, especially synthesizer keyboards, are ripe for compression. Not so much for the performing style of the instrumentalist, but because from synth-patch to another patch the levels may vary widely. Here the compressor is set up more like a limiter, with a higher ratio such as 6:1 to 10:1. The softer patches may not see any gain reduction and the louder patches should be adjusted for no more than 15dB of gain reduction. Also using newer and nicer multi-band compressors (e.g. TC Electronic Triple-C) for low/mid/high frequency bands may help, especially in sequenced multi-part synth playback. With keyboards, a soft-knee threshold helps soften the gain reduction onset.

Optionally, the kick drum may get a taste of compression if a “punchy” kick drum sound is desired. Most percussion is generally not compressed, but some may have compressors set up as limiters with thresholds at +10dBu and ratios at 10:1 or higher. Kick drum punch is usually dialed in with 4:1 to 10:1 and thresholds set to taste, typically in the six to 15dB of gain reduction. The hard/soft-knee switch can be set arbitrarily to whatever appeals to the user.

Lastly, brass instruments are another compressor application. Like on keyboards, compressors can be used to strangle direct hits on the brass microphones, especially when the instrumentalists tend to be dancing to the groove. Setting a compressor channel with 8:1 or higher ratios, and 10dB or more of gain reduction on the peaks, helps keep distortion on the brass channels from getting to be a problem.

Squashing Thoughts

While I do not expect everyone to have a dozen channels of compressor channels, it is nice to have at least a few to handle the critical offenders of excessive dynamic range. A typical hire soundco will usually have one or two premium (“diva”) compressors for the star vocals, and a bunch of general-purpose dual or quad channel compressors. I find that having a couple quality name compressors for vocals plus a couple adequate compressors for instruments works well in most gigs. One thing to avoid are compressors using external power supplies, such as wall-warts and line-lumps, as they generally do not provide the needed +18 to +22dBu headroom for live sound mixing.

Good uses for compressors placed on console outputs are generally few, but sometimes warranted. Last September I mentioned comps for “catch and hold” ring-out tactics on aux send masters. Another use for compressors are as limiters when the downstream amplifiers are without clip limiters or too large for protecting the mains speaker drivers. Thankfully, the emergence of DSP speaker processors with onboard limiters is beginning change things.