Setting Proper Gain Structure

[flemingmras]

GAIN STRUCTURE


One of the most overlooked yet most important things in a typical band's PA system is gain structure. Gain structure is the practice of matching the levels throughout your system so that:

1) Each device in the chain clips at exactly the same time

2) The only place the signal receives ANY kind of gain is at the trim(channel input gain) control

3) The signal level stays exactly the same as it is coming out of the trim control all the way to the power amps

In other words, if you have a +22dBu signal coming out of the trim control, a +22dBu signal will be seen at the amp's inputs. The ONLY place the signal should receive ANY kind of gain is from the trim control.

Again, another example...a +22dBu signal comes off the trim control. It should maintain this value through the channel EQ, the channel fader, subgroup fader, main mix fader, house EQ, crossover, and right to the amps. When the gain stays the same throughout the whole system, this is known as a "unity gain" system.

On each fader on most mixers, there will be a "0" or a "U" mark(most Mackie boards use the "U"). Setting the channel, subgroup and master faders to this mark will ensure unity gain through the mixer. In other words, with the fader set to the unity gain mark, whatever is coming from the trim control will appear at the outputs of the mixer.

On all PA gear, you have gear that is either +4dBu or -10dBv compatible. Some pieces of gear have a switch where you can switch it to either +4 or -10. This matches the inputs and outputs to the type of mixer you have. Most "consumer" gear is -10dBv while "pro" gear is +4dBu. In this tutorial we will be working with +4dBu gear.

Most +4dBu gear has an input and output rating of +22dBu. 0dBu=0.775 Volts, so +22dBu is roughly about 9.756 Volts. Mackie mixers actually have a +28dBu output rating, however that's just to give you another +6dBu of headroom to keep from clipping the mixer's output stages. The max output from each channel's trim control is actually +22dBu.

Power amplifiers, however, have an input sensitivity rating that is much lower than that. The input sensitivity rating on power amps is the minimum voltage required on the input to cause the amp's output to put out full clean power. Most power amps have an input sensitivity rating of about +3dBu-+8dBu. This is why power amplifiers typically have input attenuator controls on them, which are the volume knobs on the front panel. A VERY common misconception is that if you don't set the input attenuators all the way up that you won't get full output from your amps. However, power amps are fixed gain devices...they will put out full power no matter where the input attenuators are set, however, depending on where they're set, it will just take a much larger signal to push the amp to full power. Judging by this statement, we now know that all the input attenuators do is raise and lower the amp's input sensitivity. The input sens rating that is advertised in the power amp's user manual is the sensitivity of the amp's input with the attenuators fully clockwise(all the way up). This is why they're there. I mean, for example, let's say that your mixer is capable of putting out +22dBu of signal, yet your most sensitive amp clips at +4dBu at it's input with the attenuators fully up. Well, now you're sacrificing +18dBu of gain/headroom you could be getting out of your mixer because once your amp is at it's clip point...that's all you get. This is why the input attenuators are there...so that you can match your amp's input sensitivity to your mixer/crossover.

The reason proper gain structure is important is so that we get max gain from the mixer, giving us more dynamic range(headroom) and a lower noise floor. Max gain from the mixer will increase your signal to noise ratio, giving you better seperation between instruments. It also makes your channel EQs MUCH more effective. I see a LOT of people use WAY too much boost from the EQ itself than from the gain control. If you flatten the EQ, then maximize the gain from the trim control, your EQ will be MUCH more effective to where on some instruments you may have to use cut more than boost, which enables you to get even more gain from the channel's trim control. Too much boost from the EQ rather than from the trim will cause phase distortion, which will cloud up, or muddy up your signal to where it almost sounds buried in the mix.

Following is a step-by-step procedure for setting proper gain structure. You will need a test CD with a pink noise track to do this. Pink noise is every frequency within the human hearing range played simultaneously(at the same time). You can download a full blown test CD featuring all kinds of test signals, including pink noise, frequency sweeps, left/right test(to make sure you wired the system so that left is left and right is right).

Test CD Download

Test CD Info/Track List Download

For this procedure, set all channel EQs flat, all channel Low Cut filters off and all faders all the way down:

1) Play the pink noise track through two channels. Pan one channel hard left and the other hard right.

2) Rotate the left channel's trim control clockwise until you see the channel's "Clip" or "Overload" indicator start to come on, then back off until the indicator JUST turns off. You are now at the max gain setting on the trim.

3) Repeat step 2 for the right channel.

4) Set both the left and the right channel's faders to the "0" or "U" mark. This puts the channel at unity gain and sends the full signal from the trim control to the main mix buss.

5) Make sure both the left and right channel's are assigned straight to the main mix buss.

6) Set the main mix fader at the "0" or "U" mark. Now the trim control's full signal will appear at the mixer's outputs. Run this signal through your crossover. Set your crossover's input and output controls to the point where the input/output clip indicators JUST start to come on, then back off until they JUST go out. If your crossover's max input rating is +22dBu(it should be), then it like the mixer should be JUST going into clip.

7) With the speakers disconnected and the amp's input attenuators fully counterclockwise, turn your mains/subs amps on.

Rotate each input attenuator clockwise until the amp's clip lights are JUST starting to come on.

*Note: Most amps that are bridged will only use Channel 1's input attenuator and Channel 2's will not be active.

9) Turn on your crossover's built in limiter and set the threshold to it's highest point. Then slowly lower the threshold to the point at which the amp's clip indicators JUST go out.

Your gain structure is now properly set, and your limiter will protect your amps from going into clip.

10) Turn off your amps and reconnect the speakers, then turn them back on.

11) Now play a music CD of your choice through the system, and listen. Almost always I find that when the mains amp is set to clip at exactly the same time as the subs amp, the mains are almost always too loud, so you'll have to turn the mains amp input attenuators down more to balance the mains to the subs. This will give you even more headroom on the mains amp.



Now when you sound check your signal sources (i.e. instruments)...have each performer play the loudest that they'll play during the performance. Making sure each channel's EQ is flat at this point, rotate the trim on each instrument channel until the clip light JUST starts to come on, then back off a hair. GET AS MUCH GAIN AS YOU POSSIBLY CAN FROM THE TRIM CONTROL!!! THIS IS VERY IMPORTANT!!! Bring the channel's fader up to 0 or U, then slowly bring the master fader up to performance volume. Listen to the instrument and apply EQ if necessary, then reset your trim as stated above. Do this for all instruments.

Set all channel/subgroup faders to the 0 or U mark, then slowly bring the main mix up to performance levels. Listen to the overall mix, if you need more of a certain instrument, bring the other instruments down around it. It's always helpful to subgroup drums and vocals since they take up more channels than guitars usually(especially drums). ALWAYS keep the lead vocal channel and the vocal subgroup faders at 0 or U to get the max signal from your vocal mix. Balance your backup vocalists to the lead vocalist. The drum subgroup faders will typically be about 5-10dB lower than vocals, depending on personal taste. Then bring the guitar and the bass channel faders up to balance with the drums and vocals.

Basically the key is to start with all channel/subgroup faders at 0 or U, then listen to the live mix, and bring down the faders on channels with instruments that stand out too much. Drums and lead vocal channels, with the exception of snare, hi hat and overhead channels will almost always stay at 0 or U, then balance snare, hi hat and overheads to the kit. Then slowly bring the guitars and bass into the mix until they blend with the drums and vocals well. You may have to EQ them slightly to make this happen.

I hope you guys find the above information useful.

[rockstah]

thanks for the post!