Aug
26
Back in the olden days, when high-quality audio equipment wasn’t nearly as readily available as it is today, gain structure could make or break the sound quality of a show. Good gain structure was necessary to ensure that the audio signal voltage did not reach the console voltage rails and clip, while at the same time keeping it as high as possible above the noise floor, maximising the signal-to-noise ratio (SNR) and keeping it clean and free from distortion. Now that you can buy an off-the-shelf audio console on the high street with better than 100dB SNR, minimal THD+n and full-bandwidth frequency response, gain structure may not seem to be so important, but it can still ruin your day.
Head Amp Gain
The head amp is the first amplifier a signal encounters after entering a mixing console. As microphone level signals are low (to the order of -30dBu), they must be amplified to line level for processing within the console. By applying different amounts of head amp gain to each channel depending on how high the source level is, each input can be normalised to a standard level so that fading each channel up to 0dB individually results in the same volume at the mix output.
Each time the signal passes through a component, it adds noise. The higher the signal level, the lower this relative noise is, hence a higher signal-to-noise ratio. The ideal place to add gain is at the head amp as this both prepares the signal for processing through the console and will be the highest quality gain stage in the signal chain.
Metering
Console metering is essential to see how far from the dynamic extremes the signal is. Metering structure varies from console to console, but in general each channel will have a cue or PFL button, which when activated will display the cued channel level on a ‘cue meter’. This could be a PPM (Peak Program Meter), VI (Volume Indicator) or proprietary meter with proprietary ballistics and scaling. Generally, the better quality the console, the better quality the metering with more resolution and potentially dedicated channel meters.
The point at which the meter source is taken may or may not be variable, but will generally be pre-EQ or pre-fade. For initial level setting, it is most useful in the pre-EQ position as this monitors the raw source directly after the head amp, giving a clear indication of what the gain control is doing to the signal.
Equal Gain Techniques
There are several techniques for setting gain structure, and like everything involved in mixing, no technique is wrong and the choice is down to personal preference and workflow.
The first method requires the input source to play and raising the gain control until the input clips before turning it down just enough to prevent distortion. On the face of it, this technique looks ideal as it gives you maximum SNR whilst still preventing clipping, the drawback being that it results in zero headroom. This means that if you apply any EQ boost, it is likely to overload the channel. It also means that if you have a multi-channel mix and set all of the channels up to this same standard, the chances are that the summing bus will overload when adding them all together. The internal voltage rails don’t just count at the channel input, they apply at every single point in the mixing console.
A more sensible method for gain setting is to increase the gain until the meter reads nominal, with occasional transient peaks being allowed over. Nominal could be equal to 0VU on a VI meter, ‘0’ on a peak level meter or ‘6’ on a PPM meter but is essentially where the manufacturer recommends based on the internal gain structure of the console. 0 on a peak meter may actually represent an internal level of +4dBu, but this is the intended nominal level.
This method provides around 20dB more headroom, reducing the chance of the summing bus overloading significantly. The downside is that if you gain 20dB headroom, you simultaneously raise the noise floor by 20dB. This may not necessarily be problematic however and will still be adequately low to be inaudible with well designed equipment.
Unity Fader Techniques
Audio faders are typically designed with a logarithmic track so that around the 0dB mark, a given length of fader may change the level by 5dB whereas near the bottom of the track, the same length may change the level by 20dB (for example). Therefore, finer resolution is available when the faders are around 0dB.
Both of the equal gain methods require setting the gain appropriately before raising the input fader to set the desired volume. Depending on the source and the spill off stage, this could vary considerably and result in different channels having wildly different fader levels. This can potentially become a hindrance to workflow if you have a dynamic mix with lots of fader movement as the lower faders will have coarser resolution.
A common method among live FOH engineers is to start with the gain control fully anticlockwise, set the input fader to 0dB and increase the gain until the volume in the room is adequate. Applying this method to each channel in a mix should result in a good starting point as all the inputs will be at the desired volume with all the faders at 0dB, the ideal position for fine balance adjustment. It also means that if you put them all back to 0dB, you will return to a decent mix.
Care must be taken over the master level with this method. The dominant channels (e.g. vocals) should have nominal gain, faders at 0dB and adequate volume. If they have particularly low gain, the master volume should be decreased and compensated with a corresponding gain increase, and vice versa. Remember, it is essential that the range of signal levels across channels isn’t too wide and they are kept away from the dynamic extremes of noise and clipping.
If working in a large venue, this method should result in most of the channels being gained-up to similar levels as they must all be amplified around the same amount. In smaller venues where stage and backline spill is a major problem, some channels such as guitars may be gained much lower than nominal. This may mean the level has to be pushed more for solos, resulting in hitting the fader end stop or an exceptionally low SNR.
It’s tempting when using this method of setting gain to mix with the gain pots i.e. leaving all the faders at 0dB and applying any volume changes with the gain pots. However, it must be considered that this workflow will affect the entire channel path; it will increase the drive level to any inserted dynamics and the changes will be tracked on any pre-fade auxiliaries such as monitor sends. Once the gains, dynamics and monitor sends (if applicable) are set, the mix should be balanced with the faders to prevent compromising these aspects.
Digital Console Gain Structure
Gain structure for digital consoles isn’t much different to analogue, except that it’s even more important to ensure that it is correct; digital clipping sounds considerably worse than analogue clipping and is to be avoided at all costs.
The main operational difference with regards to digital console gain structure is the metering. Digital meters are calibrated to dB Full Scale (dBFS) where 0dBFS is maximum level i.e. red light. A digital sample is defined by a number relating to how loud it is at an instant in time, where the word length of the system defines the range of levels. A 24-bit system has 16,777,215 different levels, expressed as a binary number, meaning that the maximum level is 16,777,215. If a signal tries to go over this level, it will simply be truncated to maximum level, resulting in a harsh digital clip.
Considering that an analogue console would be calibrated such that a nominal level gives 20dB of headroom, an equivalent nominal gain structure for a digital console would be around -20dBFS. This is generally a good place to work at, although note that as with analogue consoles, metering is non-standardised; some digital console input meters are even calibrated in dBu to make it easier for the non-digital savvy mixer.
One benefit of digital consoles is that gain structure is much easier to monitor, as metering points are often selectable, and in/out meters may be provided on individual modules such as EQ or dynamics. This means that it’s possible to monitor the level at every point of the input path. However, with proper gain structure methods as previously discussed, it is not necessary to check the level so comprehensively, safe in the knowledge that headroom is available for safety.
Equal Gain vs. Unity Fader
Some FOH mixers live or die by their gain structure. They will swear blind that mixing with faders at unity improves the clarity of their mix. The opposition, the equal gain party will lament them for poor technique. Not one of them is better, and each method can pull of an equally great mix.
It’s also common for people to deliberately ‘cook’ their head amps by running them hard, believing that this results in a better sound. At least one series of console utilises soft-clip limiters on the input stage which will change the sound non-linearly as it reaches the clipping point. This is a case of personal preference and is a feature not to be expected on every console!
The equal gain method is the technically accurate way of running a console, optimising the signal for transmission through the console with maximum SNR. However, with today’s high quality, low noise floor consoles, it’s generally ok to throw away some SNR if it makes for a better workflow. I believe that mixing whichever way the mixer is most comfortable will result in the best mix.
I would suggest that with either method, once the show starts, the balance should be mixed on the faders, leaving the gains alone unless clipping occurs or extra level is required. Constantly adjusting the gain structure mid-show will detrimentally affect dynamics and monitor sends, not to mention rendering the faders an unnecessary feature!
Personally, in smaller venues, I adhere to the equal gain method as bigger level variations are necessary to compensate for large amounts of backline spill. In larger venues, I find that the equal gain and unity fader method tend to become the same thing as backline spill becomes less troublesome and each input needs equal amounts of amplification. If mixing monitors from FOH, or from side of stage in fact, I would always use the equal gain method as the added complication of multiple monitor mixes makes the benefits of unity fader redundant.
Excellent article….thanks! I personally prefer to set the gains at 0 on the meter (PFL), then set the faders at unity. Then I’ll mix from the 4 groups (drums, guitar/bass, keys/tracks, and vocals). I run the master out at around -10/-5 depending on the venue size (power amp trims maxed out…have to, so that some drunken curious fool can’t possible blow out our powered QSCs).
One of the techniques I used in the past was a similar idea, using the group masters almost as attenuators in order to get good input gain structure and faders in a useable range. I would always mix from the input faders though as the group outputs usually ended up with quite a wide range of values, with the lower ones not giving optimum resolution.
I concur about the power amp trims, though my reasoning is repeatability and consistency!