An automatic camera that frames whoever is speaking detects who has the floor and switches the shot to show them. In a video podcast, it does the job a director would do in a TV studio: cutting between cameras so the audience always sees the person talking. The difference is that here you don’t need anyone running a control board.
Here’s how different systems pull it off, which shots are worth combining, and what you can do to make the result look smooth.
The problem it solves
If you record a conversation with a single fixed wide camera, everyone looks small and viewers have to guess who’s talking. If you record with several cameras, someone has to decide when to switch from one to the other, either live or in the edit. That’s work, and a lot of it, when an episode runs for an hour.
An auto-framing camera takes care of that decision.
How it knows who’s speaking
There are several ways, depending on the setup:
By audio
The system measures the volume of each microphone or participant. Whoever sounds loudest for a sustained moment is treated as the speaker. It’s the most common method in calls and virtual studios, because each person already has their own separate audio channel.
By image
With physical cameras, some devices analyze the image to detect faces and lip movement. They work, but they depend on lighting and angle.
By sound direction
Some room microphones estimate where a voice is coming from and point the camera that way. They’re common in meeting rooms.
In a virtual 3D studio, since each person has their own avatar and their own audio, the system knows exactly who’s speaking and can move a virtual camera without the limitations of a physical one.
The hard part: not making viewers dizzy
Detecting who’s speaking is the easy bit. The hard part is deciding when to change the shot. A good system follows rules similar to a human director’s:
- It waits a moment before cutting. If someone says “uh-huh” or “right,” there’s no point in switching shots. It waits until the person has been talking for a little while.
- It doesn’t cut too often. Switching every second is exhausting. Each shot has a minimum duration.
- It goes back to the wide shot when several people talk at once or when there’s silence, so the whole scene is visible.
- It smooths out the movement. Instead of abrupt jumps, some cameras make a short, cinematic move.
Shots worth combining
| Shot | What it shows | When to use it |
|---|---|---|
| Wide | The whole table and the set | Openings, closings, group laughter, crossfire debates |
| Medium | One person from the waist up | Regular contributions |
| Close-up | Face and shoulders | Emotional moments or key ideas |
| Two-shot | Two people at once | One-on-one exchanges within the group |
A good mix keeps the pace up: the wide shot orients the viewer and the closer shots build connection.
How to talk so the auto camera works better
Even if the system does the work, you can help a lot:
- Don’t talk over other people. If two people speak at once, the camera hesitates and ends up on the wide shot or bouncing between them.
- Leave a beat before you jump in. Half a second is enough for a clean cut.
- Mute your mic if there’s noise at home. A barking dog can make the camera focus on you for no reason.
- Wear headphones. Otherwise your mic picks up other people’s voices and the system may think you’re the one talking.
- Avoid long “hmm”s while someone else is speaking. To the camera, that’s voice.
If you’re moderating the conversation, it’s worth explaining these ground rules before you start. You’ll find more ideas in how to moderate a panel discussion.
Advantages and limits
Advantages
- It saves hours of editing: the episode comes out with the cuts already made.
- It makes a format that would feel flat with a fixed camera much more dynamic.
- It lets you record or stream with several people without a technical crew.
Limits
- It doesn’t understand the content. If someone makes a visual joke while another person is talking, the camera won’t go looking for it.
- In very heated debates with lots of interruptions, the cuts can look jittery.
- It doesn’t replace a good editor’s storytelling intent in large productions.
For most conversation podcasts, the advantages win out. And if you later want to pull short clips, having the cuts already made makes the job a lot easier. See how in vertical clips from your podcast.
In a virtual studio
In remocast, the cinematic camera frames whoever is speaking inside the 3D set, and since the avatars are seated around the table, the shot changes feel similar to a TV show. With up to six people in the room, that’s the difference between a grid of faces and a conversation people actually want to watch.
If you’re still deciding whether your podcast needs visuals, check out video podcast or audio only.
Want to see how the camera moves around your own table? Create your studio for free.
Frequently asked questions
Does an auto-framing camera work with two people?
Yes. With two people, it alternates between them depending on who’s talking and returns to the wide shot when there’s silence or both speak at once. With more participants, the effect is even more noticeable.
What happens if two people talk at the same time?
Usually the system holds the current shot or switches to a wide one to show both. That’s why it’s a good idea to agree on taking turns and not talk over each other.
Can I use automatic framing with real cameras?
Yes, there are webcams and room systems that follow the speaker. They tend to cost more and depend on lighting and space. A virtual studio doesn’t have those physical limitations.