Comparing today's lift against an older one is the fastest way to see whether a technique change actually took. It is also easy to do badly, in ways that make two different lifts look identical or two identical lifts look different.
Sync on an event, not on the start of the clip
This is the mistake that ruins most comparisons. If you start both videos at frame zero, they drift apart immediately, because you did not start recording at exactly the same moment relative to the lift and the two reps did not take the same amount of time.
Instead, pick a single unambiguous event in both clips and align on that. Good anchor points:
- The moment the bar leaves the rack or the floor
- The lowest point of the descent
- The moment of lockout
SetSync handles this with sync anchor points, which is what lets two clips of different lengths be compared honestly.
Compare like for like
A comparison only means something if the only variable that changed is the one you are investigating. Before you read anything into a side by side, check that both clips share:
- The same camera position, height and distance
- A similar load, or at least a known difference in load
- The same rep in the set, since rep one and rep five of the same set look different for reasons that have nothing to do with technique
- The same footwear, since a change of shoe heel changes squat mechanics
If the camera moved between sessions, the comparison shows you a camera change. That is the single most common false positive.
Side by side or overlay
Both views answer different questions.
Side by side keeps each video readable and is better for comparing timing, rhythm and overall shape. It is the right default, and the one to use when showing a lifter their own video.
Overlay, where one clip is made semi transparent on top of the other, is better for spotting small positional differences such as bar position over mid-foot or depth. It only works if the camera was in the same place for both clips, otherwise you are overlaying two different perspectives and the differences you see are artefacts.
What is actually worth comparing
Vague comparisons produce vague conclusions. Decide what you are looking at before you press play:
- Depth, judged at the lowest frame of each clip
- Bar path shape and horizontal drift
- The timing of the hip rise relative to the chest out of the bottom
- Where in the rep the bar slows down
- Joint angles at a matched point in the lift
Pick one. A comparison that tries to look at everything tends to confirm whatever you already believed.
Use more than two when it helps
Two clips show you a difference. Three or four show you a trend, which is usually the more useful thing. Lining up the same lift at four points across a training block makes it obvious whether a change is real or whether you cherry picked two reps that happened to differ.
SetSync supports comparing up to four videos at once for exactly this reason.