Gain Structure Revisited

by Mark Amundson
in Theory and Practice
I addressed this topic more than two years ago, but took a historical and Q&A approach to justify things like 0dBu levels, gain/level settings and +22dBu brick walls. While nice and informative, a lot of readers said that they needed to delve into the real knobs-and-faders portion of setting up their consoles. This article will revisit gain structure in a straightforward way.

All Those Controls I'm an analog mixing console fanatic, but the sea of knobs frightens a lot of newbies and begets the oft-asked question, "Do you know what all those knobs are for?" Most key knobs are the gain controls for the mic or line pre-amplifiers. Most have a 40 to 70dB range of gain adjustment, and are meant to bring -60 to -30dBu microphone levels to line levels. I mention these levels as a range, because singers vary their acoustic input to the microphones greatly. So what starts out as gain adjustment for a subdued lead vocalist at sound check will likely be pegging past +22dBu after the preamp during the last set. But by trimming the sound check mic gain levels at 0dBu peaks means that I will have 22dB more headroom for that vocalist to blow their lungs out later at the gig.

But if I get a rational vocalist or musician who sound checks at performance volumes, I will dial-in their mic gains for 0dBu averages and 6 to 9dBu peaks. The reason for this is that I expect the levels to stay about the same through the performance, and I plan to have the channel fader at 0dB when that channel is the prominent feature in the mix. This allows me another +10dB of gain (level) increase if I need it at the fader and likely not clip out the channel circuitry. In English, the plan is usually to operate the fader lower than 0dB when the instrument/voice is supposed to be in the background in the mix. Or worse yet, the wash coming off the stage to your mix position (Front of House) buries the fader to nearly the off setting.

The Bass-Ackwards Method

Occasionally, you will see an anal-retentive mixing console operator who starts with all his/hers channel faders at 0dB, and then begins adjusting the gain controls from the lowest settings to the final setting. Now, this would be okay if the final settings where about 0 to +10dBu on the peaks, but then you expect the operator to begin to use the faders to mix the band at sound check and the show. But normally these bass-ackwards operators will mix using the gain controls on each channel, leaving the faders in a neat and pretty 0dB row. These are also the soundpersons you want to avoid as they have egos as large as army tanks, and think their way is the only right way to mix.

What the Designers Intended

As an electrical engineer and former console designer myself, the mixing console design priority goes to the microphone preamps first, and not the fader control electronics. The reason is that the best lownoise performance is made at the preamps, and the successive circuitry can be noisier without much impact on the whole console's performance. To grade a preamp, the term Noise Figure is used to describe how much loss of Signal-to-Noise ratio (SNR) a preamp contributes to the pristine input signals. To characterize both the mic preamp and the rest of the mixer/summer stages as Noise Figure, the following equation is used:

Fconsole = Fpreamp + [(Fsummer – 1)/Gpreamp]

From the above equation, the overall Noise Figure is the preamp Noise Figure plus the mixer/summer Noise Figure divided by the preamp gain. What you should conclude from this equation is that strong preamp gain settings do the best to reduce overall system noise, and that using the faders to mix has less impact on noise than mixing from the gain knobs.

Crossing to the Other Side

As in the previous gain structure article, there are a lot more methods to sorting out gain and level settings once the signals leave the channel strips. Console designers start introducing attenuation strategically at the summing amplifiers for the subgroups, auxiliary masters and Left-Right-Center masters so that multiple high-level signals do not sum to a clipping level. So, a summing amplifier may lose a couple dB before the fader control. And the fader control loses 10dB while actually indicating unity gain (0dB), so the follow-on +10dB make up amplifier can provide the optional gain boost before hitting the output connectors, metering, and/or next summing amplifier stage.

From my experience, it is best to kick off the show with the vocal subgroup at the nominal level (0dB) and place the other groupings a few decibels behind it. Then, upon the start of the show, you can focus on vocal loudness with the master LCR faders, nail the mix loudness within a second or two and then begin to back or forward the non-vocal subgroup faders to rough in the mix. Then, finishing touches can be done using the channel strip faders, starting with key instruments like guitars, then bass, then keys, then drums. Most times, I will have the channel faders rough set from experience with the act onstage, so I am not at 0dB on every fader and scrambling to immediately lower half the channels at the start of the show.

For most rock/country/pop acts I do, there is one lead vocalist and the rest of the vocalists backing. For this configuration, the lead vocalist is pegged at 0dB with all other vocals starting at -5dB. This way I can still place the lead vocal on top or just within the other instruments, and use the subgroup and master faders to set the overall level of the show. If other vocalists switch off taking lead vocal duties, everyone goes to -5dB until the lead vocalist is known. (Hopefully, before they start singing, but that is another matter.)

Drive Chain

At the master LCR faders, this is where opinions vary. I am a devoted -5dB to -15dB master fader guy, as I like my master metering to barely bubble beyond 0dBu depending on the venue and stage wash conditions. But if you like backing off your power amplifiers by about 10dB, then you are running close to 0dB on the master faders and living on the threat of having only another 10dB of console "gas" before clipping console circuitry. The lower master fader levels are a hallmark of us old farts who lived and died by VU meters and amplifiers with only a couple hundred watts of amplifier power to perform on. Today, amplifiers require +4 to +10dBu of input signal, but they can knock hundreds to thousands of watts into the mains speakers at those levels. If you go the old fart way, you may have a touch more hiss in your speakers than if you trim back amplifier level/gain controls. But if you do run attenuated at the amps, keep a watchful eye on your console, equalizer and speaker processor levels for clipping or limiting.

If you are following my way and leave your amps wide open, make sure your speaker processor hits limiting at or just before the amplifier delivers full unclipped power. The limiter settings are generally chosen by either speaker driver maximum program power rating or amplifier power ratings--whichever gets in trouble first. For example, if a 1,200-watt amplifier that clips at +7dBu may deserve a processor limiter setting at +6 or +6.5dBu. And some of the better processors provide both compressor and limiter settings to better tailor the driver power dissipation characteristics as a measure to prevent voice coil destruction.

Conclusions

Please use this discussion as a food-for-thought piece, as there is no absolute way to get gain structure perfect. I encourage you to comment on your methods via e-mail or share your expertise on FOH Online's Wild and Woolly forum.