For a long time, we kept our church cameras in manual focus because it felt like the safer option. We have a large LED wall behind the stage, and whenever we experimented with autofocus, the cameras had a tendency to notice all of that bright, high-contrast content behind the person we actually wanted in focus. Lyrics, graphics, motion backgrounds, and hard edges on the LED wall could all become more interesting to the camera than the person standing twenty feet in front of it.
Manual focus solved that obvious problem because the camera couldn’t suddenly decide the LED wall was more important than the pastor. But it created another issue: nothing was ever quite as crisp as I wanted it to be. Our stage is wide enough that two people can easily be thirty feet apart, so even if the focus looked good at the beginning of a shot, it might not stay that way once someone moved across the stage.
Technically, the answer is simple: adjust the focus. Operationally, that’s not so simple. Like a lot of churches, we rely heavily on volunteers, and while we can teach someone how to frame a shot, zoom, and follow someone around the stage, expecting every volunteer to constantly evaluate critical focus while also doing everything else hasn’t been particularly successful for us.
That’s when I started thinking differently about the problem. Maybe the goal wasn’t to decide whether manual focus or autofocus was “better.” Maybe the goal was to configure autofocus well enough that the camera could handle the technical part while the volunteer focused on composition, timing, and storytelling.
Manual focus was safe, but not necessarily good
There was a reason we stayed with manual focus for as long as we did. It was predictable. Once the camera was focused, it stayed there, and we didn’t have to worry about the LED wall suddenly stealing focus in the middle of a shot.
The downside was that predictable did not always mean sharp. If someone moved forward, backward, or across the stage, the focus could quickly become just soft enough to notice. It may not have been dramatically out of focus, but it also wasn’t the crisp image I knew the cameras were capable of producing.
That was especially noticeable on tighter shots. Eyes, facial hair, clothing texture, and other small details are where slightly soft focus starts to show up. The average viewer may not identify the problem as focus, but the image still feels less polished.
We had invested in good cameras, lighting, an LED wall, switching, broadcast audio, and everything else that goes into making the stream look intentional. It didn’t make much sense to settle for “close enough” focus simply because autofocus had caused problems in the past.
A wide stage makes manual focus harder than it sounds
Our stage width was a major part of the problem. People can easily be thirty feet apart from one side of the stage to the other, and our cameras regularly move between subjects who are nowhere near the same focal distance.
One shot might be a vocalist on one side of the stage. The next shot might be the pastor standing somewhere completely different. Someone may walk toward the front of the stage or move deeper into the room, and every one of those movements changes what the camera needs to focus on.
In a perfectly staffed production environment, you could simply have a camera operator continuously ride focus. But that isn’t always realistic in a volunteer-driven church environment. Our operators are already thinking about framing, movement, zoom, shot composition, and following the person on stage. Adding constant critical-focus adjustments on top of all that creates another place where things can go wrong.
The more I looked at it, the more I realized that we were asking volunteers to compensate for a technical problem we might be able to solve another way.
The LED wall was the reason we didn’t trust autofocus
The biggest problem with autofocus in our room has always been the LED wall. To the camera, that wall can look incredibly attractive.
It has bright colors, sharp edges, text, moving graphics, changing contrast, and a huge amount of visual information. Depending on what is being displayed, the wall can present a much stronger autofocus target than the person standing in front of it.
That means simply turning autofocus on is not enough. If you give the camera the entire frame and let it decide what matters, there are situations where it may decide incorrectly.
That was exactly what we wanted to avoid. We did not need autofocus that was technically active. We needed autofocus that behaved properly in our environment.
The goal became lower-maintenance operation
This is where the philosophy of the camera setup started to change for me. I stopped asking, “What gives the operator the most control?” and started asking, “What gives the volunteer the best chance of getting a consistently good shot?”
Those are not always the same thing.
A camera setup that works beautifully when an experienced production person is running it but falls apart when a volunteer sits down is not necessarily a good system for a church. The technical setup needs to support the people who actually operate it every week.
If the camera can reliably keep a subject sharp, I don’t need to make focus another constant task for the operator. I would rather have the volunteer spending their attention on the creative decisions that matter more: framing the shot, anticipating movement, following the right person, and knowing when to move.
That doesn’t remove the operator from the process. It just lets the camera handle a repetitive technical correction that it may actually be better equipped to make.
What we actually changed on our cameras
For context, our cameras are Panasonic AG-DVX200s, so the exact menu names and numeric values in our setup are specific to those cameras. If you’re running Sony, Canon, Blackmagic, PTZ cameras, or something newer, your autofocus controls may be labeled differently.
The important part is not necessarily copying our numbers. It’s understanding what we were trying to make the autofocus system do.
We left the rest of the camera operating manually the way we wanted it. Exposure, iris, white balance, and the overall image were still under our control. We were specifically trying to automate focus, not hand the entire camera over to an automatic mode.
On the DVX200, we enabled Custom AF so we could control how aggressively the camera reacted. From there, the three settings that mattered most for us were AF Speed, AF Sensitivity, and AF Area Width.
Our starting point became roughly:
Custom AF: ON
AF Speed: -2
AF Sensitivity: 1
AF Area Width: approximately the center 30–40% of the image
Those settings are not a magic preset, and I would not suggest blindly copying them onto another camera. They are simply where we landed after trying to make our cameras behave the way we needed them to in our room.
We slowed down the autofocus response
The first thing we changed was AF Speed.
We ended up around -2, which slows down how aggressively the camera changes focus. We didn’t want focus snapping quickly from one distance to another every time something changed in the frame.
A slightly slower focus transition also looks more natural in a live broadcast. If somebody moves toward or away from the camera, I would rather see focus make a controlled correction than visibly jump.
On another camera, this setting might be called something like AF Transition Speed, Focus Transition Speed, AF Drive Speed, or Response Speed. The terminology will vary, but the principle is the same.
We wanted autofocus to move deliberately, not react instantly to every change in the image.
We made the camera less eager to abandon the subject
The next adjustment was probably even more important in our room: AF Sensitivity.
We set ours very low, around 1.
The goal was to make the camera less willing to abandon whatever it already had in focus just because something visually interesting appeared somewhere else in the frame.
That matters a lot when the background is an LED wall. The pastor might be standing in front of a relatively dark graphic, and then the next slide suddenly has bright white lyrics, hard edges, and a lot more contrast. We didn’t want the autofocus system seeing that change and immediately deciding the background deserved its attention.
We wanted the camera to essentially say, “I already have something here. I’m going to stay with it unless I have a good reason to move.”
Different manufacturers may describe this as tracking sensitivity, subject-switching sensitivity, responsiveness, stickiness, or something similar. Again, the exact label is less important than the behavior.
If your autofocus keeps jumping away from the person and grabbing the background, look for the setting that controls how easily it changes subjects or focal distance.
We stopped letting autofocus look at the whole frame
The third major change was AF Area Width.
Instead of allowing autofocus to evaluate the entire picture, we narrowed the active area to roughly the central 30–40% of the frame, where our subjects normally appear.
That was a balancing act.
If the autofocus area is too wide, the LED wall has more opportunity to influence focus. If it is too narrow, the volunteer has to keep the person perfectly centered or risk moving them outside the active autofocus region.
We wanted enough room that volunteers could still frame naturally without constantly worrying about an invisible focus box, while also excluding a significant amount of the LED wall from what the camera was evaluating.
Different cameras may call this AF Area, Focus Zone, Flexible Zone, Focus Box, AF Frame, or Focus Region. Whatever your camera calls it, the idea is straightforward: don’t ask autofocus to pay attention to portions of the image where your subject is almost never going to be.
That adjustment made a lot of sense for the way we shoot. Our pastor, worship leaders, and other subjects are generally somewhere around the middle of the shot. There was very little reason for the camera to be making focus decisions based on the extreme edges of the LED wall.
We automated focus, not the camera
I think that distinction is important.
When people hear “we switched to autofocus,” it can sound like we put the camera in full auto and let it start making all the decisions.
That isn’t what we did.
We still want intentional exposure. We still want intentional white balance. We still want the operator determining the shot, the zoom, and the composition. We simply decided that maintaining critical focus was something the camera could help us with.
We automated focus. We didn’t automate the camera.
That is really the philosophy behind the change.
Don’t copy our numbers. Copy the behavior.
I don’t want someone with a different camera—or even another DVX200 installation—to read this and assume that -2 / 1 / 40% is some universally correct autofocus formula.
It isn’t.
Those settings worked as a starting point for our cameras, our stage, our LED wall, our shooting positions, and the way our volunteers operate.
What is transferable is the behavior we were trying to create.
If autofocus feels too sluggish when someone walks toward or away from the camera, increase the focus speed slightly. If the camera keeps abandoning the person and jumping to the LED wall, reduce its tracking or subject-switching sensitivity. If the background still influences focus too much, tighten the autofocus region. If volunteers are frequently framing people outside that region, widen it again.
That is a much better way to tune autofocus than copying somebody else’s settings and assuming they should work in your room.
Test autofocus under the worst conditions, not the best ones
This is another piece I think matters.
An empty auditorium with one person standing under good lighting and a static background is probably the easiest autofocus test you can give a camera.
That isn’t Sunday morning.
If you’re trying to make autofocus work in front of an LED wall, test it with the LED wall doing the things that normally cause problems. Put lyrics on it. Run bright graphics. Use motion backgrounds. Switch suddenly from a dark slide to a bright slide.
Then put somebody on stage and have them move. Have them walk toward the camera, away from it, and from one side of the stage to the other. Zoom in. Zoom out. Change the composition. Have someone cross through the frame.
Try to make the autofocus fail.
That will tell you far more than watching it lock onto one person in an empty room.
The goal is not to prove the autofocus works under perfect conditions. The goal is to understand where it fails and then adjust the camera until those failure points become manageable.
The shot itself still matters
Settings are only part of the equation.
If your subject is tiny in the frame and surrounded by a giant, bright LED wall, you are giving the autofocus system a much harder decision to make. If you’re on a reasonably tight shot and the person occupies a meaningful portion of the autofocus region, you’re giving the camera much better information.
That doesn’t mean every shot needs to become a close-up. It just means we still need to think about what we are asking the autofocus system to see.
Good autofocus settings cannot completely overcome bad composition.
What we’re really doing is trying to stack the odds in favor of the camera making the same decision we would make manually.
For us, that meant slowing it down, making it less willing to switch targets, and limiting the part of the image it was allowed to evaluate.
Autofocus isn’t about making the cameras completely automatic
I still don’t want the cameras making every decision. Framing matters. Shot selection matters. Zoom speed matters. Following the right person matters. Knowing when to move and when to leave a shot alone matters.
Those are creative decisions, and I absolutely want a person making them.
But whether the pastor’s eyes are sharp should not need to be one of those decisions every five seconds.
That’s the distinction I keep coming back to. Good automation should remove repetitive technical work without removing human judgment from the production.
If autofocus can take care of focus while the operator concentrates on the shot, that is not giving up control. It is using the technology for the part of the job it can handle well.
Sharpness matters more than we sometimes admit
One thing this process reminded me of is how much focus affects the overall quality of a broadcast.
A slightly soft shot is not always obvious enough to trigger an immediate reaction from the viewer. Most people watching from home are not going to stop and say, “The focal plane is wrong.”
They just see an image that feels a little less professional.
When the eyes are sharp and the face has good detail, the entire image feels better. The camera looks better. The lighting looks better. Even the overall production feels more intentional.
That’s why I don’t think focus is a small detail. It’s one of those things that quietly affects everything else on screen.
Technology should reduce volunteer workload
There is a broader lesson here that applies to more than cameras.
Sometimes we keep systems manual because manual feels more professional. A professional camera operator rides focus. A professional engineer makes every adjustment. A professional production person controls everything.
That philosophy makes sense in environments staffed entirely by professionals. Church production is often different.
If a piece of technology can reliably handle a repetitive technical task, there is no prize for making a volunteer do it manually just because that is how it has traditionally been done.
The goal should be to make the system easier to operate without lowering the quality of the result. In the best cases, good automation actually does both: it lowers the burden on the operator and improves consistency.
That is exactly what we are trying to accomplish with focus.
The bigger lesson
For a long time, manual focus felt safer because autofocus had burned us before. The LED wall would grab attention, the camera would hunt, and a perfectly usable shot could suddenly become soft because the background became more interesting than the person on stage.
Turning autofocus off fixed that problem, but it did not really solve focus. It simply moved the responsibility from the camera to the volunteer.
Once our stage got wide enough, our subjects moved enough, and our volunteer operators became part of the equation, that tradeoff stopped making sense.
So I don’t think the real question is whether manual focus or autofocus is better. The better question is:
How do we configure the camera so the easiest way to operate it is also the best way to operate it?
For our Panasonic AG-DVX200s, the answer has been Custom AF, a slower autofocus response, low sensitivity, and a restricted autofocus area. Your camera may give you completely different controls, but the goal can still be the same.
Make the camera reluctant to chase the LED wall and much more interested in staying with the person you actually care about.
If we can do that reliably, it’s one less thing a volunteer has to fight on Sunday morning—and one more thing we can make better behind Sunday.