One clip in. Everything below, out.

Thirty-three body points, from a video you already have

Upload one clip filmed on a phone and get the position of thirty-three body points in every frame. No markers, no suit, no capture volume, nothing to calibrate.

Fourteen joint angles, measured on every frame

Joint angles frame by frame across a movement, with the minimum, maximum and range for each joint, and a scrubbable skeleton lined up against the video.

Five ways to take the jitter out — and the research behind each

Butterworth, Savitzky–Golay, median, moving average or Kalman, applied to every chart, figure and download, with the exact settings recorded in the file.

Two measurements, lined up on the same moment

Put two measurements next to each other, scrub each until they sit on the same instant of the movement, save that as a compare point, and read the difference joint by joint from there.

Everyone you measure, session by session

File each analysis under the person in it and their profile lines every session up: range of motion per joint over time, left against right, with a gap between sides flagged when it opens up.

Pin what you noticed to the frame where you noticed it

Pause anywhere, tag the moment with a label and a comment, and it stays pinned to that timecode — on the charts, in the list, and in the downloads.

Send a report or an exercise to anyone — no account needed

Turn on a link for a range-of-motion report or an exercise and send it. Whoever opens it watches the video and reads the figures, with nothing to sign up for — and nothing they can change or download.

JSON and CSV for everything, with the provenance attached

Angles, landmarks and notes all download as JSON or CSV. Every file records where its numbers came from, what section of the video it covers, and what was done to it.

Draw the overlay to suit the footage

Choose how thick the skeleton lines and how large the angle labels are on the overlay video, with a live preview, each time you start an analysis.

Tag the stages of a rep, export logic your app can run

Scrub to each stage of a movement, tag it, and export a definition that counts reps on any pose engine — no labelled dataset, no model training.

If this, then that — cues that fire when the movement goes wrong

Tag a fault the way you tagged a stage — from a video of it going wrong — and attach the cue you want shown. It exports with the rep counter.