Most lifting videos are useless for form review. Not because the lifter did anything wrong, but because the camera was in the wrong place, too far away, or shooting at a frame rate that turns the bottom of a clean into a blur. Fixing that costs nothing and takes about thirty seconds of setup.
Put the camera at the height of the thing you are judging
The single biggest error is filming from standing eye level. Shooting down at a squat makes depth look shallower than it is, and shooting up at a bench press hides bar path entirely. Parallax is not a small effect: a camera one metre too high can make a legal squat look like it missed depth by several centimetres.
Set the phone at the height of the joint or bar position you care about:
- Squat depth: hip crease height, roughly mid-thigh
- Bench press: bar height at the chest, so around chest level of the lifter lying down
- Deadlift lockout and bar path: hip height
- Snatch and clean: mid-torso, so the whole pull stays in frame
A phone leaned against a water bottle on the floor beats a hand-held shot from standing every time.
Get far enough back, then stop zooming
Frame the lifter with roughly a head's worth of space above the bar at its highest point and a little floor below. For most lifts that means standing about three to four metres away. Closer than that and the lens distortion at the edges starts bending the bar.
Do not use digital zoom. It crops pixels away and leaves you with a softer image, which matters a lot when you later step through frame by frame looking for the exact moment the hips shoot. If your phone has a 2x optical lens, that is fine. Pinch zoom is not.
Film perpendicular, or film deliberately off-axis
There are only two camera positions worth using, and the worst thing you can do is land between them.
A true side-on shot, with the camera square to the plane of the bar, is what you want for depth, bar path and joint angles. Any rotation away from square introduces error into every measurement you take off that video.
A deliberate 45 degree angle is useful for seeing knee tracking and hip shift at the same time as depth. It is a judgement shot, not a measurement shot.
What does not work is a camera that is almost square but not quite. It looks correct, so you trust it, and every angle you read off it is slightly wrong.
Frame rate matters more than resolution
For form review, frames per second beats pixels. A 1080p clip at 240fps is far more useful than a 4K clip at 30fps, because the thing you are trying to see is usually a transition that lasts under a tenth of a second.
At 30fps, the turnaround at the bottom of a squat might be captured in two frames. At 240fps it is sixteen. That is the difference between guessing and knowing.
SetSync captures at 120fps by default and 240fps optionally, which is the trade worth making: 240fps eats storage, so it is best saved for the lifts where the fast phase is the point.
Light the lifter, not the camera
Filming towards a window or a bright gym door turns the lifter into a silhouette, and a silhouette defeats automatic pose detection as thoroughly as it defeats your own eye. Put the light behind the camera where you can.
Plain backgrounds help too. A lifter in a black shirt against black gym flooring with black plates is a hard image for any tracking system, yours included.
Keep the setup identical between sessions
This is the step almost everyone skips, and it is the one that makes the video worth keeping. If you want to compare today's squat to the one from six weeks ago, the camera needs to have been in the same place both times.
Pick a spot, mark it with chalk or a plate, and film from there every time. Same height, same distance, same angle. Then a side by side comparison shows you a change in the lift rather than a change in the camera.