How to Set Up a Livestream Mix That Actually Works (and Stays Good Week After Week)
If your livestream mix sounded fine three months ago and now sounds like a mess, you're not alone. The most common way to build a livestream mix - push every fader on the stream send to a level that seems OK that day - is also the most reliable way to end up with a stream that drifts further from "good" every week.
I've spent ten-plus years setting up livestream mixes for churches, conference rooms, lobby feeds, and overflow spaces. I keep coming back to the same approach. Boring, predictable, and it sounds great every week without anyone babysitting it.
Before the how, a quick note on the why. I covered the full version in Why is Mixing for Livestream So Different Than a Live Mix in the Room?, but the short version: in the room you and your listener hear the same PA in the same space. On a stream you have no idea what the listener is on - good headphones, a $200 Chromebook speaker, a phone, a desk speaker against a wall, a cranked sub. Your job is a mix that holds up across all of it.
The walkthrough below uses a Yamaha DM3 (see the companion video), but the approach maps cleanly to any digital console. And the same method works for any clean output that isn't the room PA - a stream feed, a lobby or overflow room, a conference room, a cry room, a recording feed. If you have a matrix output and a need for it to "just sound good," this is the recipe.
Start With KISS - the Fewest Active Faders Possible
The cornerstone of a consistent stream mix is keeping it simple, sweetheart. The more faders you have actively in your stream send, the more drift you get over time. Volunteers reach for the closest fader. Levels creep. You come back two months later and the bass is way too hot or the pastor's mic vanishes the moment the band lifts.
The fix is to base your stream mix on the mix you're already taking care of - your main PA mix - and only add a few specific things on top.
Here's what that looks like in practice.
Step 1: Send Your Main Mix Into a Matrix
On most digital consoles, you can send a group, subgroup, or your full stereo LR mix into a matrix output. Find the bus you're using for the livestream feed (in the companion video it's matrix 5/6 on the DM3) and bring your main stereo mix in at roughly -5 dB.
That's it. That single send is doing the heavy lifting. Every fader move, every mute, every level ride you do for the room is now also happening on the stream.
That's the whole point. You're not building a second mix from scratch. You're starting from the one you already trust.
Flexible alternative: if your console gives you groups (or DCAs feeding subgroups) you can send those into the matrix instead of the LR - vocals, guitars, drums, and cymbals each as separate items. This lets you do one thing the main-mix approach can't: pull cymbals down independently for the stream. More on why in Step 4.
Step 2: Add a Touch of Effects
In the room, the building gives you natural decay - early reflections off walls, ceilings, and the people in the seats. The stream gets none of that. Run your reverb and delay returns into the matrix slightly hotter than they sit in the room. Just enough to put space around the vocals so it doesn't sound like everyone is singing in a closet.
Step 3: Bring the Spoken-Word Mics Up
This step surprises people, and it's the single biggest difference between a stream that sounds professional and one that doesn't.
In the room, the volume gap between full band and a single pastor on a wireless headset works fine. The PA pushes the band, the mic is naturally softer, your ears adjust, and the room itself smooths the transition.
On a stream, that same gap is brutal. People are listening on phones, laptops, and TVs. They turn up the volume to hear the message, then the band kicks in at full level and blows them off the couch. Or they turn down for the band, miss half the sermon, and tap out.
The fix: in your matrix, push your pastor, speaker, and host mics noticeably higher than the room sends them. Start a few dB above the main mix's level for those channels and tune by ear. Listen on a phone. Listen on headphones. You want the music big and the spoken word clearly audible without anyone reaching for a remote. Same trick for lobby feeds and overflow rooms - those listeners are at the same volume the whole time too.
One thing not to do: don't try to fix the music-vs-speech gap by crushing the main mix with heavy compression so spoken word "comes up to meet the music." It flattens the band, kills the dynamics, and the speech still sounds underwater. Solve it at the send, not with brick-wall compression.
Step 4: Apply Light EQ at the Matrix Output (and Mind the Cymbals)
Listen to your stream on a handful of devices - phone, laptop, Bluetooth speaker, consumer headphones. While you're at it, listen in mono on at least one of them. A surprising number of stream listeners are hearing the mix mono on a phone speaker, and what sounds great in stereo can collapse on a single driver.
Two patterns show up nearly every time:
Bass that sat right in the room is heavy on small speakers and phones.
The top end - especially cymbals and high-hat bleed - takes over on cheap speakers and earbuds.
Pull a few dB out of the low end on the matrix output if needed. Tame the very top if it's stinging. If cymbals are the specific problem and you're on the groups-based version of this setup, pull the cymbals group down directly. Resist the urge to carve the EQ aggressively. The goal is to nudge the matrix so it translates, not to remix it.
Step 5: Add Compression - But Set It So It Rarely Works Hard
A limiter or gentle compressor on the matrix output is non-negotiable. It catches the unpredictable moments (a shouted "amen," a snare crack, a feedback duck) and keeps the stream from clipping.
On the DM3, the default dynamics curve on the output is a great starting point. It lets the signal breathe and only hits hard once you push past 0 dB. That's what you want. If your compressor is constantly working, your output level is too hot or your matrix sends are out of balance. Pull things back until the compressor is mostly resting and only catching peaks.
Step 6: Confirm Post-Fader Behavior and Test
Walk through every channel and make sure your matrix sends are post-fader. That's what makes the rest of this work - every move you make on the main mix during a service tracks to the stream automatically.
Then test on headphones tapped off the matrix. Mute and unmute channels, push faders up and down, and confirm the stream is actually following along. If anything's off, the most common culprit is a send stuck on pre-fader.
The Honest Part: You Probably Don't Need a Dedicated Stream Mixer
A lot of churches and venues put their newest or least experienced volunteer on "the stream" because the main mix is too important to risk. That's understandable, but in practice it tends to make the stream worse than it would be if no one touched it at all.
If you build the matrix-based mix above and leave it alone, you'll get a more consistent stream than most volunteer-mixed streams I've heard - across small churches, midsize churches, conference centers, and corporate AV rooms. There are actually some very large churches that don't have anyone mixing the stream at all, on purpose. The setup carries it. Free that person up to learn the main mix instead. The stream gets better and you grow a future FOH engineer at the same time.
Want to Go Deeper?
The recipe above is the foundation. There's a lot more to unpack - matrix routing on specific consoles, handling multitrack stems on a stream, EQ for different player apps, building presets you can recall service to service.
That's what we built Above AVL Academy for. In-depth training on stage audio and lighting from me and the rest of the team, plus a community forum where you can ask questions about your specific room, console, and mix and get real answers from people running the same gear in the same kind of space.
There's a free trial so you can poke around first. Try the Academy free here, and once you're in, jump in the forum and ask. That's what it's there for.
