Morris · check your swing
3D golf swing analysis explained
By Morris Golf · Updated

Why not try an AI golf caddie on WhatsApp?
Ask MorrisWhat 3D motion capture measures
A 3D system follows the pelvis and the chest as separate segments. For each one it typically reports three angles in degrees (rotation around a vertical axis, forward tilt and side tilt) and three positions in centimetres: towards or away from the target, towards or away from the ball, and up or down. The gap between chest and pelvis rotation is often called the X-factor. For a left-handed golfer every direction mirrors, so compare figures only with others recorded for the same hand.
Depth matters because the trunk turns while bent forward and tilted sideways. One study calculated the same golfers' trunk angles from vectors projected onto flat laboratory planes and as 3D segment angles. Trunk flexion at the top and side bend at impact differed significantly, and the X-factor by roughly 16°, largely because of the projection. Those were lab data: the gap came from flattening the motion, not from a poor camera.
How the movement is captured
Marker-based labs track reflective markers on bony landmarks with several cameras around the golfer, and research uses them as the reference. A 2023 study validated two inertial sensors, worn at the base of the neck and on the lower back, against a ten-camera marker system: agreement was strong for every variable, closest for upper-torso and pelvic rotation and lowest for pelvic side tilt, and sensor placement mattered.
Markerless capture uses pose estimation to find joints such as shoulders, hips and wrists in each frame. Several synchronised cameras can triangulate each point in 3D; a single camera cannot see depth, so a model has to infer it or leave it out. The authors of a 2023 single-camera golf study rate their 2D method as less accurate than 3D motion analysis, and they found the clubhead harder to locate near impact.
Frame rate and shutter limit what a camera sees around impact. The clubhead moves much further between frames than the pelvis, so ordinary video shows it as a blur or a jump; that study filmed at 240 frames per second. Many phone cameras read the image line by line (a rolling shutter), which can bend a fast shaft in the picture; a global shutter exposes every pixel at once.
What Morris would check on video
Morris's free web analyzer is not 3D motion capture. It reads one swing video, face-on or down-the-line and up to 20 seconds long, returns the result in about three minutes and deletes the video after 30 days. It finds setup, takeaway, top, impact and finish and comments on the visible movement alongside your description of the miss. From that single camera it estimates, each with a confidence level, shoulder turn, hip turn and the gap between them (face-on, from foreshortening), head sway and lift, hip drift and the backswing-to-downswing tempo ratio; from down-the-line, the hand path and the shaft angle at key positions.
Morris also says what 2D video cannot show: depth, the true rotation hidden behind the body, the clubface angle and ground forces, and on WhatsApp reads the TrackMan and Garmin data you connect. A launch monitor's parameters cover the club, such as path, face angle and attack angle, and the ball, such as speed, launch, spin and carry; none of them describes body movement.
One invented swing, three ways to record it
Synthetic example: invented numbers to show the arithmetic. Not a real player and not a target.
| What you want to know | 3D capture | Launch monitor | One phone video (2D) | Synthetic value |
|---|---|---|---|---|
| Chest (shoulder) rotation at the top | Measured | Not available | Estimated (face-on) | 84° |
| Pelvis (hip) rotation at the top | Measured | Not available | Estimated (face-on) | 41° |
| X-factor (chest minus pelvis) | Measured: 84 − 41 | Not available | Estimated (face-on) | 43° |
| Pelvis shift to target, setup to impact | Measured | Not available | Estimated, image plane only | 7 cm |
| Clubface angle at impact | Only if the club is tracked | Measured | Not available | 1.5° open |
The 43° gap is 84 − 41. A face-on video would estimate it from how chest and pelvis appear to narrow, so its figure could differ: compare it only with swings filmed from the same spot.
How to film a swing that analyzes well
Filming checklist proposed by Morris for the free analyzer; not an instructor transcript.
- 1
Fix the camera
Set the phone on a tripod or steady bag at about hip height, showing your head, feet and the club at the top. Face-on, aim it at your chest; down-the-line, put it behind you, parallel to the target line. No zoom.
- 2
Light and speed
Keep the light behind or beside the camera. Film at normal speed, not slow motion: the analysis counts on real speed. A higher frame rate adds frames near impact, not depth.
- 3
One swing, one sentence
Stop after the finish, inside the 20-second limit. Describe the miss in one sentence (start line, curve, strike) and film the next swing from the same spot.
Self-check: The clip works when your head, feet and clubhead stay in frame and the camera stays still. If Morris reports low confidence, change one filming condition and try again.
Apply it to your own game
A swing video and a description of your miss help turn a general idea into a specific check.
Sources
- Smith, Roberts, Wallace & Kong (2016) — Comparison of two- and three-dimensional methods for analysis of trunk kinematic variables in the golf swing, Journal of Applied Biomechanics — Projection onto laboratory planes versus 3D segment angles; X-factor gap of about 16° · Reviewed 20-11-2026.
- Kim et al. (2023) — Validation of inertial measurement units for analyzing golf swing rotational biomechanics, Sensors — Body-worn sensors against a ten-camera marker system; rotation, obliquity and X-factor · Reviewed 20-11-2026.
- Yamamoto et al. (2023) — Extracting proficiency differences and individual characteristics in golfers' swing using single-video markerless motion analysis, Frontiers in Sports and Active Living — 2D pose estimation from one camera at 240 Hz; accuracy limit versus 3D analysis; clubhead near impact · Reviewed 20-11-2026.
- TrackMan — 40+ TrackMan parameters — Club and ball parameters a launch monitor reports · Reviewed 20-11-2026.
Check your own swing for free
Drop a video of one swing here
One swing, face-on or down the line, your whole body in the frame. Up to 20 seconds.
Your video is used only for this analysis and is deleted after 30 days. No name, no email, no account.
Common questions
Can a phone app measure my swing in 3D?
Is a launch monitor a 3D swing analysis?
Do I need 3D motion capture to improve?
Can video measure my X-factor?

A golf caddie, not another app
No downloads, no passwords: you text him on WhatsApp and Morris helps with your game, using everything he knows about you.