Speaking to camera

People who are entirely comfortable in front of two hundred people frequently freeze in front of a camera, and the reason is that everything they use to calibrate has been removed.

What makes it difficult

The specific absences.

No reactions, so no way of knowing whether anything is landing.

No energy from the room to feed off.

No permitted mistakes, or at least the sense of that, since it is recorded.

Self-consciousness about appearance, which is heightened by seeing yourself.

And the unnatural act of speaking warmly to a piece of glass.

Giving yourself a person

The most effective fix.

Speak to one specific person — someone you know, imagined just behind the lens.

Which changes the register from broadcasting to talking, and it is audible.

Some presenters put a photograph next to the lens for exactly this reason.

Better still, have an actual person standing beside the camera to speak to, with instructions to react visibly.

And address the viewer as "you" rather than as an audience.

Energy

The calibration problem.

What feels like normal energy reads as flat on camera, for the same reason it does on audio: the room is not supplying any.

Which means deliberately raising it, which feels excessive and looks normal.

The only way to calibrate is to record, watch, and adjust — repeatedly, until you know what your own "too much" looks like on screen.

And smiling, genuinely, at the start, which sets the tone for everything after.

Eyes

Where most of it is won.

Look at the lens, not at the screen showing yourself, which is the single most common error and reads as looking away.

Cover the self-view if the software allows, since watching yourself is fatal to natural delivery.

Blink normally, which sounds absurd as advice and is affected by concentration.

And when you look away to think, look to the side rather than down, which reads better.

Framing and setup

The technical minimum.

Lens at eye level, which usually means raising a laptop on books.

Head and shoulders in frame, with a little space above the head.

Light on the face — a window in front of you, or a lamp — and nothing bright behind you.

A background that is tidy and uninteresting.

And a microphone closer than the camera, since audio quality affects perceived credibility more than video quality does.

Scripted or not

The decision.

Fully scripted and read from a prompter works if you can read aloud well, and looks stilted if you cannot.

Unscripted with bullet points is more natural and requires more takes.

Half-scripted is generally the worst, since the transitions between memorised and improvised are visible.

Which means choosing one and committing.

And if scripted, writing for the ear rather than the page, which is a different discipline entirely.

Working in takes

The advantage the medium gives you.

Record more than you need.

When you fumble, pause for two seconds and restart the sentence rather than stopping, which makes the edit obvious to find.

Record the opening several times and choose.

Record a short version and a long version.

And do the whole thing once badly before trying to do it well, which removes the pressure of the first attempt.

Watching yourself back

Necessary and unpleasant.

Almost everyone dislikes seeing themselves on video, and the discomfort reduces with exposure.

Watch it three times: once for content, once for delivery, once with the sound off to see what the body is doing.

Note two things to change, not ten.

And separate "that is bad" from "that is unfamiliar", which is most of the initial reaction.

Live video

Where the takes are not available.

Test everything twenty minutes early.

Have a backup route in — a phone, a dial-in number.

Have your notes on paper rather than on the screen you are presenting from.

Assume something will fail and decide in advance what you will do.

And keep going when it does, since recovering calmly is more impressive than a flawless run.