Best Practices in Q-SYS Gain Structure (Part 4)

Welcome back! Let's pick up where we left off.

Alright, so far we’ve talked about INPUT gain structure and processing,

now let’s talk about the other side of this equation.

Typically, there is going to be some EQ, some other user level control,

and peak limiters to ensure we don't clip anything on the output,

and then perhaps you also add some delay in if you are dealing with a larger distributed audio system.

There are varying thoughts on how to equalize a given system.

Some EQ in the output signal chain is typically for “loudspeaker EQ”,

which are filter sets typically recommended by the manufacturer, and then there's Room EQ,

which accounts for the size, shape and reflectivity of the room.

This work can get cut in half if you are using QSC loudspeakers in the installation

because you can enlist Intrinsic Correction blocks for your loudspeaker EQ.

These are custom voicing for QSC loudspeakers.

For user gain controls,

you should generally aim to allow front end control of the individual sources rather than an overall gain.

If the user adjusts each source before the mixer, it keeps the ratio of each input in the mix relatively constant.

Allowing the user to control the output level to boost a quiet source

will cause all the other sources to become incredibly loud.

And almost equally important is to properly set reasonable upper and lower limits for those gain blocks.

No user needs -100 to +20, period!

You want to avoid giving the user license to put the system in feedback everyday.

In addition, output limiting will prevents a signal from going beyond a predetermined level,

it'll prevent clipping, and prevents overloading the amplifiers.

The peak limiter block default threshold is zero.

We did a few experiments and we found what would happen if you do use zero as that threshold level.

Below the threshold, the signal is unaltered and everything is fine.

However, beyond the threshold, the limiter will start to take effect.

As you can see, while the 0 dB threshold seems like a good number to start, 0 dB cuts it a bit too close.

So we recommend you change the threshold level to -3 dB to allow for a bit more margin on the limiter,

which should help keep you out of clipping.

However, when you are using QSC amplifiers,

you don't really need any sort of limiting circuitry on the output side.

QSC amplifier limiters are automatically set if the QSC loudspeaker blocks are configured correctly.

There are a few types of limiting.

The “DAC limit” prevents clipping at the input stage of an amplifier.

Next is “Protect”, which will provide Limiting due to driving low impedances.

And lastly you have “Limit” which is the computed limiting based on load time.

However if you are not going to use a Q-SYS network amplifier,

then eventually you need to convert the digital audio back to analog.

In this limiting example, we saw clipping at 0 dBFS (as you would expect).

In keeping with our -20 dBFS gain structure, the digital ‘Gain’ should be left at 0dB.

If we look at our controls here,

this is perfect for a standard professional downstream audio device (+4 dBu nominal, +24 dBu Max).

However, what if the next device in line was a consumer level device?

Ideally, we’d like all the components in the chain to clip at the same point.

Assume that Consumer line level is about -8 dBu, and you still want to leave 12 dB of headroom.

You need to also assume that the maximum input level is +4 dBu (RMA).

To scale the output at this level, you would use the ‘Output Gain’ control below to top meter.

This is an analog gain stage after the Digital to Analog convertor.

Sometimes we must drive a mic level input on another device from Q-SYS.

There are two major concerns here.

1) Phantom power is not enabled on the input Q-SYS is feeding. That will damage the Q-SYS output circuit.

2) The ‘Output Gain’ will need to be adjusted to reduce the max output to mic level.

As you can see in this diagram,

moving to Consumer or Microphone level greatly degrades the signal-to-noise ratio.

So whenever possible, use Professional Line Level.

Let’s wrap up everything that we’ve learned, shall we?

For your input calibration, make sure that you bring in everything at -20 dBfs.

Then trim high and low frequency noise using low-pass and high-pass filters.

Parametric EQ can be used for manufacturer suggested curves.

And then Notch Filter or Parametric EQ can be used to trim out feedback frequencies.

Automatic Gain Control or Compressors can be used to get consistent levels from a wider range of talkers.

Next, don't place user level controls or gates ahead of the automixer because that's just terrible.

And then, of course, use the fewest user gain controls possible to get the job done

and manage the range to avoid letting users ruin the gain structure.

For Output limiting, using QSC amplifiers and loudspeakers together

will get you limiting automatically when those blocks are configured properly.

We always recommend -3 dBFS as that safe limiter threshold.

And then, finally, you're going to scale the analog section of the output circuitry

to match your downstream device.

Ideally, you want Q-SYS to clip at the same point as the next device downstream.

And that's it for Gain Structure through Q-SYS