Ten mistakes people make filming a lesson they teach

You have taught this lesson in a room, to a person, more times than you can count, and it works. Then a camera goes on a tripod and what comes back is a video of somebody talking, with the part people needed happening just outside the frame or buried under the sound of the room.

The short version: make the demonstration visible and the explanation audible. Add an angle or a microphone where a test shows it is needed, record enough detail for the intended output, and say where you are in the lesson as you film.

A general list of filming mistakes is shorter than this, because past a point it repeats itself. This one runs to ten because it covers one activity: a person filming a skill they teach.

Why does filming a lesson fail differently?

Because a lesson has two subjects and a talking head has one.

There is you, explaining, and the thing being demonstrated: the hands on the keyboard, the mouth making the sound, the cursor crossing the screen. Almost everything below is one of those two being served at the expense of the other. The camera, the light and the microphone all default to the person, because that is what they are sold to do, and the person is rarely what was searched for.

The mistakes post-production cannot fix sorts filming faults into the ones an edit can buy back and the ones that are simply absent. This list overlaps it and is wider, because much of what follows is recoverable and still not worth what recovering costs.

Ten checks before you film

You light the room instead of the thing you are teaching

A lamp pointed at the space makes everything equally visible and nothing important, so the eye settles on your face and misses the fingering, the stitch or the cursor. On a music education partnership the artificial light was balanced on the angle covering the hands, because the hands were what was being taught. Where to put your light when you only have one gives you the position; what changes here is which subject it serves.

One angle on a skill that needs two

Your face explains and your hands demonstrate, and one camera has to choose between them second by second. Cutting between the two is also the cleanest way to shorten a long lesson without the join showing, though not the only one: b-roll, an insert, a screen recording or a punch-in can each cover a trim, and the two-angle version is simply the one that hides it best. How many cameras do you need? works out whether a second earns its place at your scale, and Giving every camera angle a job stops you buying a third that never does.

You film the detail at the size you will show it

If the close angle is framed exactly as it will appear in the finished video, nothing is held in reserve, and punching in later, which means enlarging part of a frame in the edit so it fills the screen, throws away resolution and goes visibly soft. Frame that angle tighter than feels necessary, or record it larger than you publish. Zooming in without going soft explains how much room you have.

The demonstration and your voice arrive on one microphone

When voice and demonstration share a track, independent adjustment can be difficult. Separate microphones may give the mix more control, but each can still hear the other source. Test the placement before the lesson.

You film beside a classroom window across a long session

Daylight is the best free light you have and it refuses to hold still, so an hour of lesson filmed beside a window starts in one exposure and ends in another, and the halves stop cutting against each other. On that partnership the fix was not shutting the daylight out, it was adding a steady artificial source alongside it so the swing stopped mattering. Filming next to a window has the method.

Nothing is said out loud, so an hour of lesson cannot be searched

A long lesson arrives as one file named after the camera, and finding where you explained the third exercise means watching until you reach it. A cooking partner was asked to say the stage out loud while filming, on the recording, and hours of footage became scannable. What to say out loud while you film gives you the phrases, and the whole cost is remembering.

You teach over the demonstration at the same volume

When you talk across the thing you are demonstrating and both reach the viewer at the same level, neither lands: the words vanish into the sound of the work and the work vanishes under the words. Decide before you film which one carries each moment, then either stop talking over it or let the demonstration run alone. Mixing a lesson so people can follow it covers what a mix can do afterwards, which is less than deciding it at the shoot.

Wardrobe changes between sessions that belong to one series

Clothing changes are normal across separate lessons. Within a continuous sequence, photograph the first setup so pickups can match it. A series can still feel coherent through its framing, sound and recurring structure.

You open on your own name and logo

An animated title carrying your name spends the opening seconds telling the viewer something already on the channel, at the exact moment they are deciding whether this is the lesson they searched for. Start on the thing they came for, named or demonstrated, and put your identity somewhere it costs nothing. How long is a hook? sets the window, and Film three hooks, not one is what to do when you cannot tell which opening works.

The series has no repeatable form

A repeatable opening and structure can help viewers find their way through a series. Keep the form recognisable while allowing the lesson to run for as long as the teaching needs.

What do you change before the next lesson?

Not all ten, and not in one sitting. Take the two that match what you teach: if the skill lives in the hands, the light and the second angle carry the video, and if it lives in what you say, microphone separation matters more than any camera decision.

Three items here can cost money: a second camera, a second microphone, a lamp if you do not own one. The other seven are position, timing and the decision to speak while the camera is running. Do those first, film one lesson and watch it back before buying. The list of faults you thought were equipment problems tends to get shorter.

Sources and evidence

What here is backed by research on instructional video, and what is our own working defaults from specific partnerships.

What is established. Reviews of instructional video design, Cynthia Brame's summary of cognitive load in educational video and the later systematic review by Fyfield and colleagues, support the idea that asking a viewer to track an explanation and a demonstration at the same time asks more of their attention than either alone, and that giving each its own camera, light or microphone reduces avoidable extraneous load; neither review prescribes the specific setups below.

Our method. Lighting the thing being taught rather than the room, and separating the demonstration's microphone from the teacher's, are this studio's working defaults built from specific partnerships, not a fixed formula for every lesson.

References.

• Cynthia Brame. Effective Educational Videos. pmc.ncbi.nlm.nih.gov. Cognitive load, engagement and active learning should guide educational video design. Review/guideline, not a fixed-length law.
• Fyfield et al. Improving instructional video design: a systematic review. ajet.org.au. Segmenting and learner control are relatively robust; redundancy effects vary. Instructional contexts differ.