Hybrid Church Service AV

The rule that saves churches is the headroom multiple. A stream needs roughly twice its bitrate in measured upload, because the building is also carrying Wi-Fi, phones, the office and the card reader in the lobby. A plan advertised at 50 Mbps is a download figure with a much smaller upload number behind it, and the upload number is the only one that matters here.

Measure it on a wired connection, at the encoder, at service time on a Sunday. A Tuesday afternoon reading from the office Wi-Fi is not evidence.

This is the part that gets skipped, and it is the reason most hybrid audio sounds wrong. In the room, a drum kit arrives at the congregation acoustically at 105 to 120 dBA at three feet. A guitar amplifier arrives at 95 to 110 dBA. The operator therefore barely pushes those channels through the loudspeakers, because the room is already full of them.

The stream has no acoustic path. It only receives what is in the mix. So the same fader position that sounds balanced in the room produces a stream where the drums are almost absent, the guitar is a rumour, and the vocal is naked and loud. Turning up the stream master does not fix it, because the proportions are wrong, not the level.

The fix is a second, independent mix. On a digital console that means a stereo matrix or a pair of aux sends set to post-fader-independent, fed by the same input channels, with its own balance. This is the same problem covered at length in how to mix for livestream, and the signal path itself is laid out in church livestream audio setup.

A console that cannot produce a genuinely separate mix is the real constraint on hybrid, and it is worth more than any camera upgrade. The XR18 at $509.00 has six aux buses and will do it. The X32 has 16 aux sends and six matrix outputs and will do it comfortably. A small analog mixer with two aux sends already committed to monitors will not.

Read this table as a list of the sources most likely to be wrong on your stream this Sunday. Each one is a source whose acoustic level in the room is very different from its level in the console.

Say this out loud in the committee meeting. At 1080p30 a single stream needs 4.66 Mbps. At 4K30 the same service needs 15,192 kbps, over 15 Mbps, and about 30 Mbps of measured upload with headroom. Almost nobody watching a church service on a phone over cellular data will receive that, and the platform will transcode it down to something close to 1080p anyway before it reaches most viewers.

The money that a 4K camera would absorb is better spent on the broadcast mix, on two room microphones, and on lighting the platform so a 1080p sensor has something to work with. A well-lit 1080p image beats a dim 4K one on every device your congregation actually owns. If a 4K camera is bought, buy it for the optical zoom range and the auto-tracking, which are real advantages, and then stream 1080p from it.

For churches that do want a controllable camera, a PTZ such as the PTZOptics Move SE 20x gives a single volunteer three usable framings from one body. Compare the options in PTZ cameras for churches and livestream cameras for churches.

The stream is quiet and the room is loud. Shared mix. Build the separate bus described above.

Audio leads video by a quarter of a second. Video processing takes time, audio does not. Delay the audio at the switcher until the lips match. Never try to fix this by moving the microphone.

The stream buffers during the last song. Something else on the network woke up, usually a backup or a room full of phones joining the guest Wi-Fi. Put the encoder on a wired connection on its own segment, and give it priority.

Nobody online can hear the congregation. No room microphones. Two microphones and a low fader fix it permanently.

It works when one person is there and fails otherwise. This is the real hybrid failure mode and it is a people problem. Fix it with a written path, a saved scene and a rota, as set out in how to train church sound volunteers.