The Phone Video Cheat Sheet
Most filming advice is written as prose, which is the wrong shape for it. You do not read a paragraph while you are standing in a cafe deciding which lens to be on. You want a table you can glance at.
So this is that. Four tables and a checklist, no story around them. Everything here is the stuff I actually reach for, and it is all technique rather than gear, so none of it expires when a new phone comes out.
Which lens and frame rate, by situation
The single most common reason footage looks cheap is being on the wrong lens. The ultrawide has the smallest sensor on the phone and bends everything near the edges, which is unflattering on a face and unforgiving in low light. If you are unsure, you are on 1x.
| Situation | What to use |
|---|---|
| Filming yourself talking | 1x · 30fps · lock exposure |
| A room or a whole space | 0.5x · 30fps · hold still |
| Anything you will slow down | 1x · 60fps · bright light only |
| Indoors under ceiling lights | 1x · 25fps if you see banding |
| Hands, food, close detail | 1x · step closer, do not zoom |
| Walking with the phone | 1x · elbows in · 30fps |
| Golden hour outside | 1x · exposure down a notch |
| A dark room or a bar | 1x only · never 0.5x |
| Something far away | The telephoto lens, not pinching |
| Posting to more than one platform | 4K 30, then crop for each |
On frame rate: 30 is the safe default everywhere. 24 gives you slightly more motion blur between frames, which is the look people mean when they say cinematic, but it also means you have to pan more slowly. 60 is for footage you intend to slow down later. Played at normal speed it reads like a home camcorder, which is the opposite of what most people expect when they turn it on.
The banding one catches people out. If you film indoors and see faint stripes rolling up through the frame, that is your lights flickering at mains frequency and your shutter not matching it. Switching between 25 and 30fps usually kills it.
Where to stand, in any light
Light position does more for a shot than any setting, and it costs nothing. Straight-on light flattens a face because it removes every shadow that gives it shape. Move so the window is off to one side and the shape comes back.
| The light | Where you go |
|---|---|
| A big window, daytime | Side on to it, about a metre back |
| Harsh midday sun | Find shade, or put it behind you |
| Golden hour outside | Face into it, exposure down |
| Overcast, flat grey | Get closer to anything bright |
| One lamp in a dark room | Lamp beside you, never behind |
| Overhead ceiling lights | Turn them off, use a window |
| A cafe with mixed light | Face the window, not the room |
| Street lights at night | Under one, not between two |
| A white wall nearby | Face it. It bounces light back at you |
| No good light at all | Film your hands, not your face |
That last row is not a joke. A hands shot in bad light is usable. A face in bad light is not, and no amount of grading afterwards will fix it.
What each setting costs you when it is wrong
Every one of these is a switch you can find in under a minute, and every one of them quietly degrades everything you shoot until you change it.
| Setting | What it costs you |
|---|---|
| HDR video on | Looks right on your phone, unpredictable once a platform re-encodes it |
| High-efficiency format | Smaller files, and apps that refuse to open them without converting |
| 60fps by default | Reads like a home camcorder at normal speed |
| Beauty filter on | Skin texture removed at capture. It does not come back |
| Auto exposure | The shot brightens and darkens every time you turn |
| View outside the frame | You spend the take watching the part you are not recording |
| 0.5x indoors | Soft, noisy, and bent at the edges |
| Capturing at 1080 | No crop room and no room to stabilise afterwards |
| Text in the bottom quarter | Covered by the caption and the handle |
| The front camera | Smaller sensor and a softer lens than the back |
The capture-resolution one is worth spelling out, because it gets stated backwards constantly. Filming in 4K does not mean your post is delivered in 4K, and on most platforms it is not. What the extra pixels buy you is room: room to crop in, room to reframe, room for software stabilisation to shave the edges off without softening what is left. Shoot big, finish small.
What you see, and what to do about it
This is the one to keep. It is indexed by the symptom, which is how you actually arrive at the problem: you look at a clip, something is wrong, and you do not have the vocabulary to name it.
| What you see | What to do |
|---|---|
| Your face dark, the window blown out | Tap and hold on your face to lock exposure there |
| Everything looks flat | Move so the light is off to one side |
| Harsh shadows under the eyes | The light is too high. Get lower or diffuse it |
| Orange face, blue background | Two light sources of different colour. Pick one |
| Grainy and soft indoors | More light, or drop back to 1x |
| The shot brightens as you turn | Lock exposure before you move |
| Stripes rolling through the frame | Switch between 25 and 30fps |
| Blown-out white sky | Drag the exposure slider down |
| Skin looks plastic | A beauty filter is on. Turn it off |
| Washed out after posting | HDR video is on. Turn it off |
The checklist, before you record
Nine things, about forty seconds. This is the part that actually saves footage, because every item on it is something you cannot fix afterwards.
- Lens on 1x, not 0.5x
- Tap and hold until focus and exposure lock
- Exposure dragged down a touch
- Grid on, horizon level
- Nothing important in the bottom quarter of the frame
- The light is off to one side of you
- Something close to the lens, for depth
- Two seconds of handles at each end of the take
- The background is still, not busy
The handles one is the cheapest habit on the list. Start recording two seconds before anything happens and stop two seconds after it finishes. Almost every clip that has to be binned is missing those four seconds.
And before you post
- Exported once, not twice
- Text clear of the caption band
- The first frame works as a cover on its own
- Sound checked on speakers, not headphones
- It still looks right on a screen that did not film it
- You held the last shot one beat longer than felt right
- The file came from your edit, not from downloading your own post
That last one matters more than it sounds. Saving your own video back off a platform and re-uploading it somewhere else puts the footage through a second round of compression, on top of a file that was already compressed once on the way up. Always go back to the export.
The one thing not on any of these tables
Every table above assumes you are shooting one framing and living with it. If you post to more than one place, you are either cropping a vertical video into a horizontal one, which never looks right because the composition was not built for it, or you are filming the same thing twice.
That is the problem ContentCam exists for: it records vertical and horizontal at the same time, from one take, so the wide shot is a real wide shot rather than a crop. There is more on how that works here, and on why cropping between the two rarely survives.
If there is a row you would add to any of these, email me and I will put it in.
Common questions
What frame rate should I use for social video? 30fps for almost everything. 24fps when you want a filmic cadence, 60fps only when you intend to slow the footage down.
Which lens should I use? The main one. It has the largest sensor and widest aperture. Keep faces off the ultra-wide, and never use digital zoom.
What is the 180 degree shutter rule? Set shutter speed to roughly double your frame rate: 1/60 at 30fps, 1/48 at 24fps. It gives motion the amount of blur the eye expects.
Why does my video look grainy? ISO is too high because there is not enough light. Add or move a light rather than raising gain further.
Do I need different settings for TikTok and YouTube? The settings barely change. The framing does, because a widescreen crop of a vertical frame keeps only about 32% of the height.