Session 3: Editing

Edit in a fixed order, because doing steps out of sequence means redoing them. Import and organise — put all the takes in one place and watch them once without cutting, noting the good moments. Rough cut — assemble the structure and cut the obvious waste. Fine cut — tighten the pacing. Then text, music and captions. Export last. The mistake is adding captions and music to a cut you then change, which means doing them twice.

A phone editor is genuinely sufficient: CapCut, InShot and VN all handle trimming, text, captions and music, and they are free. A laptop editor gives you a bigger timeline and faster scrubbing, which matters as videos get longer, but it is not necessary for short-form. Use what you will actually use — the best editor is the one open on the device you already have.

Then a habit that saves more time than any technique: keep your project organised. Name the video, keep the source files together, and export a copy of the finished file somewhere safe before publishing. Editors crash, phones fill up, and losing a finished edit to a full storage warning is a particular kind of avoidable pain.

Ninety per cent of editing is deletion. Cut the pauses, the breaths, the restarts, the 'so basically what I mean is', and any sentence that says what the last one said. Viewers do not notice that these are gone; they notice only that the video feels tight. If a moment does not earn its place, it goes — however much effort it took to shoot.

The practical technique is to watch your rough cut and mark every point where your attention dips. Those are the cuts, and they are almost always accurate, because you are the least interested viewer your video will ever have — you already know what happens. Where you get bored, a stranger leaves.

Then the discipline of length. Short-form rewards completion, and completion falls sharply as length rises. A sixty-second video that everyone finishes reaches further than a three-minute video that most abandon halfway, even though the second contains more. Cut until it hurts slightly, then cut one more thing — that is roughly the right length.

The best transition is a cut. Straight from one shot to the next, on a word or a movement, with no effect at all. It is invisible, it is fast, and it is what professional video mostly consists of. Everything else is a choice that should be justified by meaning, not by availability.

Effects transitions — spins, wipes, glitches, zooms — date a video quickly and draw attention to the edit rather than the content. Use them rarely and consistently: if you use one style, use that one style, because five different effects in sixty seconds reads as indecision. A jump cut (cutting within the same shot to remove time) is the most useful technique in short-form and needs no effect at all.

Pacing is the rhythm of the cuts. Vary the length of shots: a run of quick cuts creates energy, a longer shot lets a point land. Uniform shot length, whether fast or slow, becomes monotonous and the viewer drifts without knowing why. And cut on movement or on a word, not in the middle of neither — cuts that land on a natural beat feel intentional even when nobody can explain why.

Text should support the message, not decorate it. Use it for the hook in the first second, for key points a viewer might miss, and for anything numerical or named. Keep it short — a few words, not a paragraph — place it where it does not cover your face, and use one font at one or two sizes throughout. Text that moves, spins or fades in and out is harder to read and looks less confident than text that simply appears.

Music sets tone and covers small audio imperfections, but it must be licensed. Use your platform's built-in library, which is cleared for that platform, or genuinely royalty-free sources. Taking a popular track from a streaming service is copyright infringement and can get the video muted, blocked or your account penalised. Keep music low — around a fifth of your voice level — because it should sit underneath, not compete.

Captions are not optional. A large share of video is watched without sound, so an uncaptioned video is silent to most of its viewers, and captions also help in noisy places and for viewers who are deaf or hard of hearing. Auto-captioning is a starting point, never the finished product: it mishears Nigerian names, places and technical terms constantly. Read every caption and correct it, keep them to one or two lines, place them where they do not cover your face, and use a size readable on a small screen.

Export at the resolution and frame rate you shot in — usually 1080p, and 30fps unless you deliberately shot slow motion. Higher is not better if the source is lower; it only inflates the file. Export vertically for short-form and check the finished file on a phone before publishing, because the editor's preview is not what a viewer sees, and text that looked fine on a laptop can be unreadable or badly placed on a small screen.

Then review it honestly against three questions. Would I watch this if someone else made it? If not, cut more. Can I follow it with the sound off? If not, the captions are not doing their job. Does it end before I want it to? If it ends exactly when you get bored, it is probably slightly too long.

The habit that improves editing fastest is watching your own work after a day away. Immediately after editing you know what was meant, so you fill in the gaps; a day later you see what is actually there. It costs a day and it is worth more than any tutorial, because the errors it reveals are yours specifically rather than generic.