Gain structures

[BigYinUK]

Evening everyone.

 

Just been having a lively discussion regarding mixing console gain structures.

 

The way I was taught and the way I've always mixed for the past 37 years is to set the desk up as so:

 

First set each of the channel gains so that the VU meter is reading (roughly - for the sake of example) +3 or +5 or last two yellow LED segments, maybe a flicker of red on an analogue desk - Do this for every channel

 

Add EQ for each channel if necessary

 

Set up the mix with the channel faders

 

Set up your foldback if necessary, add fx etc etc

 

DO NOT touch the channel gains again unless some peaking/overloading is occurring or if you do adjust, just set it back to where it initially was I.e +3 or +5

 

Now, this means of course that the channel fader positions will be all over the place.

 

There seems to be a "modern" leaning towards starting off by setting all the faders to Zero and mixing on the channel gains, leaving the channel faders more or less on or around the zero position, which seems to me to be looking for trouble as it'll affect monitor levels if the gains are adjusted mid show and properly p off the talent.

 

Is there a right and wrong way or is it down to personal preference?

 

Regards

[timsabre]

Your way is the right way in my opinion but I have come across the "faders in a nice line at 0db" approach too. However this does not make best use of the desk's headroom/noise floor.

 

I can see an advantage where inexperienced operators are mixing (eg churches) and the inputs are very predictable - set the gains so that 0db is about the right volume, then the operator can be told to set the faders at 0 and just move them a bit if a slight adjustment is needed.

[chelgrian]

 

Is there a right and wrong way or is it down to personal preference?

 

 

I don't think there is a hard and fast rule really the points to ponder are:

 

Having faders all over the place limits the degree of useful control you have on the surface so if all you care about is control via the surface faders the keep everything at one level be it zero and set the head amps appropriately might be the right solution.

 

Having the head amps low limits the usefulness of in channel led meter strips.

 

If you have dynamic processors in the channel it may be that you need to setup the head amp as you would on a analogue desk as you need sufficient gain in to things like compressors and sufficient dynamic range between the point it comes in at and the noise floor that it can do something. Then you end up in the faders all over the place situation.

 

If you have post monitor feeds from the channel then to get sufficient level to them again you need the analogue method.

 

Personally I like to have the best of both worlds. I run the head amp hot enough to have reasonable level to auxes and so the led meter strips are useful. Then start the faders at a reasonable point, say zero or -5 which gives move a greater degree of individual channel control. Before going to FOH I run the channels through an audio group or mix or whatever your desk calls it which itself is set to FOH. That mix is dialed down to an appropriate level such that FOH's heads are not blown off and I don't have feedback.

 

At least in my world I generally don't even up with very low faders from this point except for things like kick drums and bass guitars where the audience can hear both the direct and indirect sound

[Oldradiohand]

I'd never adjust gain live using the presets - too many desks I've used have crackly pots. Faders are designed for gain riding! I like to set the desk so the faders are near their normal stop giving me a nominal 10db in hand, leave foldbacks pre fader send if possible so the mix stays the same, set fx sends post fader and then balance on the channel faders. Yes, they end up all over the place. It's the way I was taught to do mixes for broadcast. Perhaps I'm just old fashioned too.

[chelgrian]

I'd never adjust gain live using the presets - too many desks I've used have crackly pots. Faders are designed for gain riding!

 

Not as much of an issue when the desk is digital and there is no pot. However the point still stands for example the head amps on say an M7CL are not smooth. Quite apart from the point where the relay goes to change the amp voltage the resolution of the control of the head amp is not fine enough and you can hear audible jumps in volume as you adjust it.

[pete10uk]

Your way is the way I was tought, and the way I do it for most things.

I was always told it's called a structure for a reason, everything is structured for optimal performance and the faders control how much level is required into the next part of the structure, get the early part wrong and no matter what you later on, you can't bring it back.

 

I came a cropper a while back, everything seemed fine in rehearsal but on the show it was evident that the amps were to low, 2 lessons learned, 1 don't do free jobs where the organiser has no clue and asks to to sort too much out for the time available and 2 don't put the amps backstage without a tech to do some running mid show. A few tears later everything worked out.

 

All that said if I'm feeling lazy and there are just a couple of inputs I have been known to set all the faders to 0 for easy reference.

[Electrolytic]

There seems to be a "modern" leaning towards starting off by setting all the faders to Zero and mixing on the channel gains, l

 

This method is not modern and has no place in digital audio, a weak signal is inefficient use of the bit depth.

the method is from the analogue days when historically fader carbon tracks were of poorer quality, thus staying around the nominal mark negated issues.

 

Its a certain type of engineer that uses the fader-line method, and it works for them.

 

Your way is the right way in my opinion but I have come across the "faders in a nice line at 0db" approach too. However this does not make best use of the desk's headroom/noise floor.

 

ANALOGUE:

the purpose of gaining a signal to around the manufactures mark of nominal is to utilise the optimal sound quality. This optimal level has been researched by the design engineers to be the best trade off between noise floor and distortion. every component in the signal path works at its best when receiving the optimal signal.

 

by using the fader-line method and under gaining a signal because thats all you require in the mix, surely you have not made best of the desk pre amps because you have not achieved the best possible Signal to Noise Ratio S.N.R.

[Bobbsy]

Can I be a fence sitter on this one? I'm sort of a hybrid of the two.

 

When setting up, I put the channel fader at zero then adjust input gain and EQ so that is "about right". In my case, I consider "about right" to be near the top of the greens--I like a bit of headroom since I'm frequently having multiple mics open at the same time.

 

However, from that point I only touch the input gain in a dire emergency and do my mixing on the faders. This means the faders are rarely, if ever, in a straight line but do tend to operate in that "sweet spot" where you can mix smoothly.

[Shez]

Your way is the right way in my opinion but I have come across the "faders in a nice line at 0db" approach too. However this does not make best use of the desk's headroom/noise floor.

 

I'll have to disagree with that one. A quiet acoustic sound will need a lot of gain to get it up to 0dB on the desk and if it needs to also be quiet in the mix, the fader will be down low - you've just added a load of unnecessary gain (noise) and then dropped it right back down again.

 

This method is not modern and has no place in digital audio, a weak signal is inefficient use of the bit depth.

All modern digital desks have 24 bit converters and work internally at an even greater bit depth - the digital noise floor is far too low to be of concern with low signal levels, no matter how wonky your gain structure.

 

I'm generally a fan of setting the gains to get the right output level with the fader around the 0 mark. I'll never touch the gains again after that - as pointed out that will mess out any sends from that channel. Mixing takes place with the faders - the scale is such that the most fine control can be found around the 0 mark. Small movements at -20 cause a much larger change in signal level.

 

There are compromises though - a loud source will be a lower signal level in the desk which may make sending high levels to monitors difficult although in practice, if the source is loud on stage, it won't be needed quite as much in monitors so that one sorts itself out to some extent. Inserted dynamics processors are the other one although I've never struggled to set the threshold sufficiently low.

 

In theatre, for mics, I'll always want to be able to push up anyone's fader to about -5 and have them come out at the right level. Working any other way would be a massive headache.

[Andrew Edwards]

Can I offer the words 'unity gain' into this interesting discussion?

Everyone seems to have alluded to this but not actually said it.

 

Personally, I aim for unity gain at sound check then leave it alone once monitor mixes are established. I expect some changes in fader level for both corrective and aesthetic reasons. However, if I find any particular ones significantly higher or lower then I try to find out why (for example mic placement has altered or a certain pedal attenuating too much when switched in). As Shez said, a major exception to this is deliberately high dynamic range for subjective reasons.

 

One other factor which makes the faders all over the place is not quite getting 'unity' gain out of the instrument at sound check. Or to put it a slightly less diplomatic way, they didn't play as loud as that in the sound check...

 

It doesn't really matter how you reach the best final product at the end of mix but I have a preference towards keeping it 'tidy'.

[Shez]

Can I offer the words 'unity gain' into this interesting discussion?

Can you define what you mean by "unity gain"? I hear the term bandied around a lot but I've never encountered a definitive definition. I suspect it's something we all know about but don't necessarily associate with that particular term. Linguistically, it suggests a gain of 1 or 0dB so in essence, a piece of wire - presumably not what we're talking about here.

[sam.spoons]

I'm with the 'gain structure' people here, everything in the analogue parts of the signal path generate 'noise', correct gain structure minimises the amount of 'noise' which arrives at the speakers. In simple terms sending a quiet signal to the power amps means you need to turn them up to max so the power amps 'noise' is at a max, if you send a loud signal to them you can turn them down which reduces the power amps 'noise' but also reduces the 'noise' in the input signal the net result being less noise coming out of the speakers. The faders end up where they need to be.

[Andrew Edwards]

I have always understood unity gain to be getting the maximum SPL of the sound to be at 0dB upon PFL. Therefore making the full dynamic range of the sound 'minus infinity' to 0dB to match on the fader. As mentioned, noise must also be mitigated during this process. I see unity gain as part of the overall gain structure, from DI/mic all the way to speaker.

[Jevans]

Unity gain is very simple - all it means when applied to an electronic component/circuit/system is that the signal level (measured in volts, dBu, whatever, as long as it maps back to volts) coming out the output is the same as the input ie. a gain of 0 dB. That's all. Nothing to do with dB SPL at all.

[Andrew Edwards]

Perhaps it should be unified gain instead for when we are not taking about voltage. It would appear that these two concepts have been somewhat merged based upon what I was taught.