Another gain staging question

[Brad Weber]

Originally Posted by audiomatic

I wish there were a formula for input levels and gain staging through the mixer, amps and speakers. Unfortunately, the only thing that really matters is how everything sounds in the room. I try not to pay too much attention to polls. It's more important for it to sound good than to have "proper" gain staging.

There are formulas and a reason that many mixer manufacturers include the nice little gain/level graphs for their mixers but most people don't want to get into that much calculating and planning, not to mention things always ending up being different than expected. And while I agree that how it sounds is the overriding factor, gain structure can be a critical element of that. I personally don't think it is as important to 'buy into' one particular approach as to understand the general concepts and the advantages and disadvantages of each so that you can then apply them in the most effective manner for any situation you encounter. However, not applying a single specific gain structure approach to every situation is quite different than not implementing some effective approach for gain structure.

A factor that I don't think has been mentioned is summation. When you start combining signals they can sum with a resulting higher signal level. If you have sixteen 0dBu sources all mixed to one bus then depending on the sources and their signals at any point in time the level of that combined signal could be significantly greater.

An adjunct to this is that you may correct for this by reducing a Group fader or something similar, however that is only reducing the level after that fader and not for any of the signal path before it. You could have 0dBu at the input, 0 at the Group meters and at the Master meters and so on but if the faders aren't at 0 then you don't have 0dBu in between. Pretty easy to think that the meter levels and related fader levels are directly related. a +10dB signal could be a fader at 0 and +10 on the meters after that fader or it could be 0 on the meters with the fader at -10. Unless you consider the interrelationship of the levels and the faders then it is very possible to be clipping at different points in the mixer without recognizing it. One reason for setting gain structure with all faders at 0 is that it lets you directly meter the actual levels, however once you know that you have good basic gain structure that may not be where you run the faders all the time.

[bobbycito]

the thing i feind when setting up gain is what am i doing am i mixing FOH only, monitors only or mixing FOH with foldback montiors and sending lineds out to a multitrack all affect the way i need to set my gain. for ex: when only doing FOH i want to be at rougly 0 (unity) on all my faders and adjust my gain from there, when mixing FOH with foldback and an aviom system i need to keep in mind every time i need to adjust my gain beacues i need more or less then i am effecting everyone on the signle chain and in turn there mixes are being changed so i try to get a good level and not touch it with thos setup.

the issue of where to set outputs of devices is for best s/n ratio but when you are not in control you want to protect yourself while mixing and see what the hottest level they will send you is. also keep in mind with keys and guitar borads is that when the change patches if they are not leveled out you can get some clipping.

[audiomatic]

Originally Posted by Brad Weber

A factor that I don't think has been mentioned is summation. When you start combining signals they can sum with a resulting higher signal level. If you have sixteen 0dBu sources all mixed to one bus then depending on the sources and their signals at any point in time the level of that combined signal could be significantly greater.

Thanks, Brad. Summation was mentioned by Tim Corder in the post I linked: (0dB mixing) "produces the maximum signal to noise ratio, but there is absolutely no headroom at the mix bus summing amplifier." I didn't specifically mention that because it's beyond my level of understanding. It's good to hear that there are reasons to experiment with unity mixing rather than blindly follow the signal to noise argument for 0db mixing.