Mixing for live stream and in-house (part 2)
When the spring 2020 pandemic hit we, like everybody else, had to shut down and figure how to set up a live stream of a bare-bones worship service. As things rolled along, we were gradually able to allow congregants back into the sanctuary for live services. But we still wanted to maintain a live stream for those who weren’t comfortable returning. Long term, while we’re at it, we could continue the remote feed for shuts-ins, people traveling, and others who just want to join in from wherever. A few decision points to consider:
What’s more important, the live feed or in-house? Seems obvious to say “both,” but that takes a level of resources and complication we didn’t have. Which brings up the second question: how many people can we realistic train and have available to run this stuff? A bigger team means we can have multiple job positions each week focused on a variety of tasks. If you only have one or two, however, your options are limited. And finally, do we have the resources (equipment and funds) to follow a more high-end approach?
In our case, we have a very small team and could not realistically foresee having three or four individuals serve each week on AV-related jobs. That narrows our options, which, by the way, is the main topic for this post. There are several approaches for mixing both in-house and live stream, and we’ll focus on two of them here. I’ll cover more later; they get progressively more complex (and better).
Our three options are:
Taking a copy of the main house mix
Creating an aux send mix for the stream
Setting up a second console for a completely independent stream mix
Option #1: Main mix
#1 is easy. Route your main mix to an additonal output on the back of the console. For digital boards, you can do this in a setup screen where you select where that main mix is going. I have a Yamaha TF5, which features Omni outs. This is just a bunch of XLR-M jacks that you can assign anything to. Run an XLR cable from there into an audio interface connected via USB to your streaming computer. Depending on the interface, you might need to use an XLR-F > 1/4” balanced cable so it plugs into the 1/4 jack on the box. This tells the interface that it’s a line level signal, not a mic level signal, and so will avoid potential distortion. Now select that interface in your streaming software, use the level control on the interface, and you’re done.
Everything you do for the house mix also goes to the stream, so you probably have to decide which is more important. Presumably the in-room mix is the priority, so mix to that and accept what gets sent to the stream. So what’s the big deal? Your house mix takes into account not only the electronic signals coming into the console, but also any acoustic sources live in the room such as drums, piano, horns, and so on. You won’t have to blend as much drums into the mix, but those at home can’t hear that live stuff, so the drums will be too low. It’ll sound drier without the natural room reverberation. It’s a huge compromise, but at least it’s easy, costs nearly nothing, and you only need your regular audio person to run things.
Option #2: Aux mix for stream
If you’ve used aux sends on your console to set up a stage monitor mix, you know what we’re talking about. Aux sends are on every channel (aux 1, 2, 3, etc). Turning up an aux on a channel makes a copy of that signal and routes it to a master aux summing amplifier (or virtual bus on digital boards). They’re essentially independent mixers, allowing you to create multiple mixes of the same material.
Select one of your auxes, say aux 1, and turn it up proportionately on each of your channels—maybe a lot of piano and guitar, but less cowbell (I know, I know). Digital consoles allow you to push a button and flip the faders, so you can mix this aux send on your main channel faders. Route the master aux send to a physical output on your console (one of the Omni outs on my Yamaha TF5). Connect this to the audio interface on your streaming computer.
Now, look at each channel’s aux send and find the button labeled “pre.” This stands for pre-fader, and it determines where the copy of the channel signal will actually be taken from the signal path. The two options are either right after the main channel fader (post-fader) or right before (pre-fader). For stage monitors, we want to set up a completely independent mix that won’t get changed as we adjust our house mix. Set the auxes to pre-fader for that.
But streaming is different; while we want to set up a different balance of channels, we also want some of what you do in the house mix to tweak that. Huh? First, a post-fader aux means your main channel mute will also mute the aux feed. This is a good thing here because you’re gonna turn off the band mics when they’re done, and therefore want those same mics muted on the live stream. You certainly don’t want to keep switching back and forth between main and aux layers to do this. You also want subtle tweaks in the mix balance to adjust the stream, since you’re not hands-on mixing both at the same time. When you need to turn up the lead vocalist a bit, or turn down the electric guitar a tad (happens all the time, of course), that will also tweak the stream mix. Once you set a good fundamental balance of everything in your aux mix, go back to your main mix layer and it should be pretty close.
The challenge for setting this mix is that you’re still in the room with all the live stuff going on. Even with headphones I have to push them against my head in the attempt to close out the drums and everything. During rehearsals, walk somewhere else in the building with the console iPad app and tweak it from there.
You don’t have a separate set of EQ and compression/effects for the stream, so it uses the same settings as the house mix. This might not be a big deal, but if you tend to EQ your channels to compensate for room issues such as feedback frequencies and such, then your aux mix will suffer accordingly. Try to get the system overall tweaked out, then EQ for taste. And yes, my channel EQs are a blend of both, so don’t fret too much.
Also keep in mind you can set what parts get sent to this aux mix. In our case, we play pre-recorded music in the room before the service starts. We’re not licensed to do that on live streams (yes, that legal thing, which is really important), so I have that audio channel turned off permanently on the aux mix. You can also add stuff just for the stream, such as an additional reverb or effect. Just make sure to listen to the stream periodically to make sure it’s getting what you intended.
What I decided to do in our case
We haven’t talked about option 3, a second console, but that wasn’t in the cards for us. And a main mix copy was pretty bad, to be honest. It wasn’t close enough. Would it work? Sure, but setting up an aux mix is not difficult, doesn’t require special training or a second person on the team, and it finesses things much better for the stream. Once I get it in place, the volunteers don’t even need to know it’s there, although I encourage them to take a listen once in awhile. For most of you, this is probably the best way to go, so give it a try and let me know what you run into.
