dB.

Audio levels are represented using dB (decibel), which is a relative unit of measurement on a logarithmic scale.

A +1dB difference in level is equivalent to the power of the original signal being multiplied by 10^1/10 (≈ 1.259). +10dB equates to 10 times the power, +30dB is 1000 times the power and when you get to +60dB you’re multiplying the signal by one million!

This scaling leads to a rule of thumb that each +3dB increase equates to a doubling of power as 10^(3/10) ≈ 1.995 ≈ 2

i.e. to make something twice as loud, increase the level by +3dB.

The reason dB is used to measure audio level is because our ears work in a similar way, that is to say our hearing sensitivity to changes in level is also roughly logarithmic.

Gain stages.

There are multiple gain stages throughout the signal path, and each can increase or decrease signal level, but at 0dB (or ‘unity gain’) the signal is passed through with no change. The effect of each gain stage is cumulative, so considering a simplified path of a single input to output:

Optimal signal level and headroom.

Both analogue circuitry and digital processing will work to their best with a specific signal level.

For example, with a digital console, the tiny signal from a microphone must be increased using a preamp to get it to the ideal level for the input of the ADC (analogue to digital convertor). If not increased enough then errors or noise may appear, but if it is increased too much, the ADC will distort.

This can be thought of as similar to a light in a room, if the light is not bright enough then it’s difficult to see properly, but if it’s too bright then you will be blinded and also not be able to see.

So it is best to start with the optimum signal, keep it near to this level throughout any processing and only attenuate it at the end of the signal path. This is the reason faders have more attenuation than boost and it’s also the reason 0dB on a meter is often considered the optimum signal level, as it means no change has occurred.

Headroom is the level above this optimum (0dB) that you can go to without ruining the signal. This can be important because you want your average (RMS) level to be close to the optimum 0dB level, but unexpected peaks can easily go way above this, even if only in for a very (very!) short period of time.

In the analogue domain, these peaks in level can introduce distortion or compression (which can sometimes be desirable, like a guitar overdrive), but in the digital realm, reaching the maximum just results in a flat waveform and lost information.

For analogue connections between equipment it is also important to run optimum levels, this ensures the best signal to noise ratio is attained in the system without introducing unwanted distortion.

For digital connections between equipment however, the data is binary, so other than transfer errors (which are not related to signal level) the information either arrives at the destination or doesn’t.

Different dB’s.

As dB is relative, a reference point is required to make any meaningful measurement.

Our digital mixers include analogue inputs and outputs, and they process and control signal in the digital domain, so we use two main references.

dBu:  A measure of voltage, where 0dBu (i.e. the reference point) is approximately 0.775 V (RMS). This is used in the analogue parts of the signal path to measure electrical energy. There is no upper limit for dBu (i.e. a signal could be at +100dBu ≈ 78000 volts!), but every analogue input and output will have a maximum rating.

dBFS: This is a digital measurement of level relative to the maximum possible value that can be represented using bits. i.e. when all the bits of a sample are ‘1’ (or ‘on’).

As these are not directly comparable (one is analogue and one is digital), the design of the equipment is the only thing that dictates how the two measurements align with each other.

We have ensured it works in the same way with all our digital consoles however, and a -18dBFS digital (internal) signal will appear at an analogue output as a +4dBu professional line level signal.

What’s important at an analogue input?

This is arguably one of the most important stages of a signal path, as bad signal in means bad signal out. The aim is to get the signal average to the optimum working level for the analogue to digital convertor.

XLR Mic/Line inputs: Preamp gain and/or pad should be used to adjust the signal average level to around 0dB on the preamp meter. Refer to the reference guide or technical specifications document to check maximum input level for the socket you are connecting to. With SQ local inputs for example, this is +30dBu, so you should not connect anything which could be putting out close to 25V.

TRS Line level inputs: Designed to be used with professional +4dBu line level equipment. Refer to the reference guide or technical specifications document to check maximum input level for the socket you are connecting to. With Qu and SQ local inputs for example, this is +22dBu, so you should not connect anything which could be putting out close to 10V.

3.5mm (Mini Jack/ST3) Input: At 0dBu this input is more sensitive than TRS line level inputs as it’s designed for used with unbalanced consumer line level equipment, which has a lower output level. This increased sensitivity translates to a lower maximum input level of +18dBu.

What’s important at an analogue output?

Our analogue outputs are at +4dBu (professional line level) when the meters are showing 0dB.

These outputs can be connected to +4dBu line level inputs on other audio equipment.

For use with -10dBV (consumer line level) equipment, the output level can be reduced using either the master send level (for a mix) or a direct output trim (for a direct out).

What’s important when processing the signal?

Any digital clipping will result in lost information and audible digital distortion.

To prevent this occurring, 18dB of headroom is provided, meaning a signal can peak at up to 18dB over the optimum level (0dB on the meters) without ill effect.

The main thing to watch out for here is when one processing ‘block’ (e.g. an EQ) is increasing the level to the point of clipping but a subsequent processing ‘block’ (e.g. a compressor) is reducing the level. In this scenario, although the channel output level may be ok, the signal has already been affected and any clipping or distortion will still be there, just at a lower level.

What do the meters show?

On A&H digital consoles, the metering works in the same way as with our analogue consoles.

• A 0dB reading on an input meter means the gain and/or trim has been set optimally.
• An output level showing at 0dB on the meters equates to a +4dBu signal at an analogue output socket, meaning the next piece of equipment (with a +4dBu input) is being sent the optimum signal level.

Then throughout the digital path, a signal measuring 0dB has 18dB of headroom available.

What do the peak (Pk) indicators show?

Peak indicators warn that the signal is at -3dBFS, so very close to digital clipping.

The channel peak (‘Pk’) indicators on our digital desks use multi-point sensing to show that peaking is occurring at any one of the metering points in that channels’ processing.

For example, if an PEQ is peaking, but the compressor is subsequently reducing the level, the ‘Pk’ LED will illuminate to show an issue even if the channel output level might look ok.

On Qu and SQ there are additional Pk leds next to dedicated controls for any processing which has the potential to increase signal level.

Why do the USB output levels seem really low?

The digital outputs on CQ, Qu and SQ (including AES, Qu-Drive, SQ-Drive, USB-A, USB-B and Option Cards) are taken straight from the mixers’ core. The internal 18dB of digital headroom is included, so a signal reading 0dB on the meters is at -18dBFS in the digital domain and this is the level being sent or recorded.

Increasing the digital signal level before it is sent would not provide any extra information or resolution. It would only increase the risk of clipping.

So this means there is a built-in ‘safety net’ at no other cost than sometimes needing to normalise or otherwise adjust levels in post-production. As an example, when recording drums this could mean that a take with some unexpectedly loud snare hits is still usable.

If required however, it is possible to sacrifice some of the headroom in exchange for an increased digital output level:

• Direct out levels can be adjusted for individual input channels in the ‘Routing’ screen of the Qu and SQ, with up to +10dB of gain available.
• Channel faders control a gain stage which can be used to boost a signal by up to +10dB.
• Make-up gain is available in every compressor (at the end of every channel) which can provide up to +18dB gain. The compressor can also be set to a ratio of 1:1 for no compression, or used as a ‘brickwall’ limiter with Infinite ratio and a higher threshold.

In Summary

For all Allen & Heath Digital Mixers

and

Illuminated Pk LED = -3dBFS