What is Gain Staging and how to use it for voice over.
Ok, so everyone has their 'special sauce' when it comes to mastering their voice over. I do and like everyone else, I have it dialed in. Or at least I hope you do. Its a long time struggle to get that 'perfect sound' out of your equipment from the mic down to the rendered file. Everything in your toolbox needs to be honed and if there is a new technique, its hard not to adopt it. I know. I've been there and I guess I'm still traveling. To say that what I'm presenting here is my favorite method is to say that its my favorite--today. So, I'll not be going into what I believe to be the ultimate trick to use, but I will say its made my voiceover mastering quite fluid and almost plug-and-play. So hang on with me. We have some work to do.
Lets start with what I've learned
Its taken me a lot of trial and error to arrive at my method and madness. It will probably take you a while too. But overall, one thing I've learned; and still learn about is the use of gain staging. Gain staging is all about getting the right levels at the time of recording. Like the old saying "You can't make a silk purse from a sow's ear" rings true. Poor recording levels can be much worse than having a bad mic or even a bad environment. Well, they can be worse, but lets not add to the problem. Its very important to set the proper levels when making a recording. Audio engineers always set the input levels to their desired specs and try to maintain those specs throughout recording. Mostly this is done for music recording, but the same can be said for voice and voice over recording. Engineers also like to have headroom when it comes to what's recorded. By that I mean, an amount of decibels below the target or final dB that the product will be finalized at. There are many opinions on where these initial levels need to be set and most of the time, it will have to be set individually. I have to even if I record in my booth with the same mic and at the same distance. My formula is one that there is about a 6-10 dB range between the recorded input dB level and what my final max dB level will be. For example, I record with an initial level between -12 dBm and -9 dBm and I also use an external vu meter to nail the level down before I start recording. Here is what it looks like:
As you can see, its a simple, dual channel vu meter I got off amazon of which I am only using one channel. Since I record in mono, no need for the right channel.
So what's your secret sauce? So when I come in, I turn on the equipment, the vu meter lights up, I bring the mic close and record my favorite line "I eat my peas with honey. I've done so all my life...". Its something I can repeat with consistency. Once I have the mic at the right location as to get my voice to bounce around at about -3 to 0, I'm happy. I record the same line and view it in my DAW and look at an input vu meter from the FX there as well.
As you can see, the input vu meter and the dBFS (dB Full Scale; green bar on the left side) shows the level peaking just above -6 dBm. If it were lower than that or if the vu meter FX had the needle bouncing around -10 to -3, I might consider applying some pre-FX gain to the track:
As you can see here, I increased the pre-FX gain by +1.6 dB to compensate and increase the gain before any FX were applied. I want to see the input vu meter hover somewhere between -3 and 0 with peaks no greater than +3. This way, any levels in my compressor/EQ will be the same. This is gain staging. This is where the initial magic starts. Pre-FX.
Ok, but how does it taste once the sauce is slathered on?
Well, I'm glad you asked. Or rather, I asked on your behalf. See, once I get the input vu meter sorted out, I apply my compressor/EQ, an exciter plugin, and finally an output vu meter to check that the input and output levels match (as best as possible). I also don't want to see a bunch of large swings in the output vu meter unless there are a lot of dynamics in the voice over. If I am reading a story, I might yell or otherwise become dynamic and expressive. Possibly quiet. This is where dynamics are acceptable, but for straight voice over like in explainers, I'm not really trying to be dynamic. Therefor, keeping a nice even level is good for me.
But wait! There's more!
Ok, I get it. Too Long; Didn't Read. Well if you got this far, thank you (and you only have yourself to blame ;-) ). I haven't mentioned the last part of the mastering. This and only at this point will I set a mix volume level. I typically leave the master mix volume alone and only apply a -3 dB limiter (just in case...) and go to each track and set the track volume to be max -3 dBFS, nominal -6 dBFS. If I can get that to show that, I'm golden.
The only thing left to do is to check how it renders. By render, I mean what does the track look like and is supposed to sound like based on max dB and RMS. RMS is Root Mean Square--something I know absolutely NOTHING about, but its related to LUFS; a Perceived Loudness. I shoot for about -18 RMS at -3 dB.
Is that all?
Yeah. That's about it. Oh, I use a DAW called Reaper. I also want to see a big, fat, fuzzy caterpillar shape when its rendered. This way I know I haven't clipped anything. I can now feel confident that the files are produced to my specifications and standards.
Good luck and remember, gain stage and use a vu meter! It'll save you a lot of headaches.