On 'Gain Staging for the DJ' and Not Blowing Your Speakers
So, having just wrapped up a weekend where I was involved in a couple of parties (including being a part of setting up the soundsystems), it's come to my attention that most DJs really don't understand gain staging at all.
In short, this has led to one of the speakers used blowing itself and a bunch of us are having to foot the bill - not so much fun (I'd much rather buy a bunch of records, thank you).

So... What is this 'gain staging' all about? you ask.

Gain staging is the process of ensuring that the signal is being sent and received at an appropriate level for the next piece of gear to be happy.
Correct gainstaging can do lots of things to make the performance of your system better, including:
* Bettering the signal to noise ratio (and thus getting a little more sound out of your speakers)
* Reducing distortion at every gain stage
* Reducing the likelihood of square-wave production, thus increasing the life of your amps, speakers and ears!

More or less, any piece of gear is expecting to see a certain level of signal down the line. There are four reasonably commonly used signal levels:
-10dBV (the standard Consumer Line Level you'll see from an unbalanced RCA or 1/4" TS out)
0dBV (a far less common level, but my mixer puts that out to the unbalanced 1/4" TS outs)
0dBU (also less common, but out there - a few dB lower than 'Pro Line Level')
+4dBU (Pro Line Level, what you'll see on most XLR outs or balanced TRS 1/4" outs of mixers)

When interfacing with other gear (eg. When you put your mixer into powered speakers, or an amp, or a crossover or whatnot) it is expecting to see one of these levels (it will tell you which). If you haven't delivered this level, you'll have one of two problems:
* You've put through too much level, and you'll be distorting on the input of the next piece gear.
* You've put through too little level, and you'll need to amplify the signal at the next stage, thus increasing the noise level on the line (every amplifier in the world produces noise.

When you retain the expected level, you are running at what is known as unity gain - there is no loss or gain on the line, and thus signal to noise ratios are kept low.
All amplifiers (and there'll be some in each piece of gear you're using) has what is reffered to as headroom, a kind of safety net in which you won't reach a point of a noticeable distortion - this is space to play. Often better gear has more headroom, but most works on quite fine lines.

More often than not, when setting up soundsystems, the main speakers will be set up 'fully open' - that is, the amplifiers for them are set to maximum gain and the Master Volume on the DJ mixer is used as the attenuator. This isn't ideal - really, the DJ mixer should output at unity gain and the amps' attenuators should be used for final level changes, but this just isn't a realistic move and wouldn't allow the DJ the control that they require.

So I can hear you all saying "Get to the fucking point already, How do we avoid blowing our gear!?"

Alright, alright! I'm getting there...

So, as a DJ, what you should be worrying about is keeping out of the red.
Remember that headroom thing? Yeah, so your mixer only has a little headroom - it sounds best when signal comes in to the masters (from the channel fader) at 0dB. You want to be seeing your level peaking somwhere from -3dB to +3dB or thereabouts. More than that and you may already be distorting at the channel fader, before we've even hit the master fader!
On my mixer there are only channel meters from -6dB to +3dB, that tells me to keep the fuck away from the +3dB side of things. If you want it louder, ride the master fader/pot.

At the master fader/meter, things are just as important, so KEEP THE FUCK OUT OF THE RED... it's really pretty simple. Once again, things will sound optimal at 0dB out of the master fader - this is the level the system should be tuned for. That said, it wouldn't make sense to start your night at this volume, you can start down at -10dB or -20dB even... you mightn't have many bars on your meters, but don't worry... that's the point - you've got headroom to burn; When you want to, you can safely double your volume. Sure, there won't be the most optimal signal to noise ratio, but that little bit of noise is a SHITLOAD better than having distortion pumping from the speakers or pounding it out to an empty room (a pet hate of mine).

Back to the headroom thing, it's usually safe to run the master between negative infinity and around +4dB... most gear is built with enough headroom at their input amplifiers to withstand an extra 4dB - nothing too bad will happen. That said, 0dB as the MAXIMUM on the master meter is still the most optimal - you will be maintaining unity gain. This is CORRECT GAIN STAGING.

The more level you pump out of your mixer past 0dB - the more it distorts, once you've hit the 'Peak' light you're certainly producing square waves, but you may already be earlier.
Additionally, you could be gainstaging just fine up until the amp but just not have a powerful enough amp. It's often better to get an amp that rates higher than the maximum input of the speaker so as to not cause squarewaves at the amplification stage.

So what is a square wave and why do you keep rambling on about them?

A square wave is a wave that is... errr... squared off at the top and bottom.

Generally speaking, the purest sound that can be produced (theoretically speaking) is a sine wave - a single tone at a set frequency and amplitude. On paper, it looks like this (image one):

Alternatively, a square wave looks like this (image two):

When discussing sound reproduction, we often look at sine wave vs. square wave reproduction as a way of showing distortion.
In an analogue amplifier, as you overdrive the amplifier a perfect sinewave will start to square off at the top and bottom - the more distorted (the more you peak), the more your sinewave (image one) starts to sound like a squarewave (image two).

This is bad for a few reasons:
* Your music isn't sounding the way it's supposed to (notably, the top end starts sounding extra harsh and the low end starts to fart and get a bit whimpy).
* The amplifiers hate seeing squarewaves and heat up a lot, reducing their life or causing them to cut out as a protection mechanism kicks in.
* Your speakers are being forced into a very ugly movement and the cone may just jump off the magnet (as happened to us at the weekend).

So, what is this ugly movement that happens to the speakers when they see square waves?

As shown in image one, a sine wave is a smooth, even movement. They are easy for speakers to reproduce as the speaker just moves in and out evenly (if you can imagine, the movement of the cone out is the top peak in the graph, the movement in is the bottom peak).

As you start to clip, the sinewave becomes a square wave - the signal is being amplified and going clean, clean, a little distorted, fucking nasty sounding (image three):

Now, what had been an easy movement for the speaker is getting much harder; Instead of simply going in and out at an even rate, the speaker is now being told by the amp to go to a certain distance and stop there for a little while and then to come back in (meanwhile we aren't even reaching the peak that we'd wanted to - your music is just sounding wrong now.
Now this whole stop-start thing mightn't sound like that much of a task, but it might have to do this a couple of thousand times a second, and REAL music signals are much more complex than sine waves - you're sure to kill your speakers with this exertion.

Sadly I can't find a picture of a dead speaker, but with any luck, Ollie will upload a picture of the one we killed.

So, Why are squarewaves bad for my ears?

While not all the research is conclusive of such, my ears definetely agree with some of the research that has been done...
More or less, when a squarewaves hit your ears, they put undue strain on some parts of your ear - I feel this as a pressure over my ears and temples, I can hear squarewave distortion and overcompression (which produces a squarewave-like signal) and it aint pretty.
Anytime I feel this pressure on my ears it translates to tinnitus... my ears can handle a fair bit of volume, but it's gotta be clean signal. I can tell you that on Saturday I complained about it... it aint fun to play or listen to music with pressure behind your ears, and it aint easy to sleep with tinnitus in your head.

Anyways, that's it for the minute - any questions, I'm more than happy to answer (and I'm sure some others are more than qualified... Kieren, some of the MPT boys, etc) - but just remember

Last edited by ferretrock: 20-Oct-09 at 09:53pm