Talking Heads & Music Beds

In addition to the FOH gigs, I’ve been doing a fair amount of audio post-production lately, and, in fact, if your church could use a little help with the audio for your videos, I’d love to chat. For the rest of you, I have a little tip for the increasingly popular announcement video.

As I’m visiting churches, one of the biggest issues I notice in announcement videos is the music bed is often nearly buried in the video. This is usually done because the music bed was fighting the “dialogue” at a higher level. The downside to this approach, though, is it often robs the video of the energy the music is there to impart in the first place. So what’s going on here?

When it comes to music beds from music libraries, most of them do NOT have a lot of dynamics. This doesn’t necessarily mean they’re over-compressed or limited, although, some can be. I just mean they tend to be pretty consistent in level and loudness. It’s typically another story for most of the on-camera personalities in church announcement videos, though.

Most of the folks on camera and even doing voice-overs in church videos are not professional actors or voice over talent. I think this is great for these types of videos because we get to see people who are actually a part of the church on screen. However, these folks tend to be very dynamic when they speak.

Mixing a “talking head” with music is not much different from mixing a vocal with music. If the dynamics of the sources are too far apart, they won’t sit together, and the source with wider dynamics will typically need to be louder to be heard at all times in order to make that source’s entire dynamic range audible against anything else in the mix.

So a better solution than just turning the music down is to manage the dynamics of the talent. A lot of the videos I get to mix already have some level adjustments automated by the video editor to help with this, but that’s typically not enough to solve the issue. In my opinion, the best tool to employ is a compressor or leveler.

I often have three dynamics processors ready to go for the post work I do, but I rarely use all of them. First up I’ll have a limiter set up just to catch peaks if someone gets a little shouty. A lot of people will emphasize the first syllable of a sentence harder than the rest of it, and this limiter helps tame that while also helping later processing work better. Then I might follow it up with a bit of light, transparent compression. I don’t want it to sound processed to someone so I try not to overwork this, however, it’s always contextual.Personally, I find higher energy music beds usually need more compression on the talent. Finally I have a leveler. Leveler’s aren’t very common, but a leveler is similar to something like a really slow compressor, although, the one I use will turn things up AND down as needed. I look at a leveler as more of an automatic fader rider, and if I get everything in front of it right this can really help keep a voice up front while still feeling natural.

As I mentioned, I don’t use all of those dynamic processors on everything. A lot of projects just get the leveler. Some get a bit of compression added. Sometimes I only add the limiter with the leveler. In fact, a big reason I started using a limiter was because those shouty peaks can have a tendency to pump a leveler, but a limiter in front of the leveler will catch those peaks so the leveler works more efficiently. It’s a very similar technique to the ol’ 1176 into an LA-2A trick on vocals.

If you’re new to using compression on something like this, experiment a bit to get a feel for it. Overdo it and then turn it off and on to start learning to hear what it does. Then start dialing the compression back so you get the results you need without over-processing things.

Another approach you can use is to work backwards. Put the music bed up first around where you’d like it, and then add the voice on top of it so the relation between music and voice is nice when the voice is loudest. Then start adding compression until the voice is clearly audible at all times. Again, if this is new for you, do it right before lunch or at the end of the day. Listen again when you come back from lunch or the next day and see if it still sounds natural or if you overcooked it. If it sounds strange or unnatural or processed or clamped down or pick your favorite description of “not right”, then dial the compression back and adjust the relationship between the music and voice as needed.

I hope that helps a bit. If you have other questions about audio for videos, please add them in the comments, or if you’d like to get my help with videos your church is doing, send me a message through my Contact page.