Pick how far back
Three to thirty seconds. Long enough to finish a rep and turn around.
Move. Turn around. See what you just did. PoseBack shows you a few seconds ago, continuously, so you never have to touch your phone mid-set.
The problem
A real mirror only works when you’re facing it. Filming means picking the phone up, stopping, scrubbing back, and losing your rhythm. PoseBack removes the loop entirely: keep moving, turn around whenever you want, and the last few seconds are already playing.
Three to thirty seconds. Long enough to finish a rep and turn around.
A countdown shows the live view while you walk into position.
The mirror keeps running. Say “replay that” to see it again.
Why it’s different
Say “replay that”, “slow motion”, “delay fifteen seconds”. Recognition runs on the phone, so it answers instantly and works in a basement gym with no signal.
Listening for a command does not interrupt whatever you’re playing. Most voice features take the audio session and silence everything else.
One purchase, yours for good. No subscription, no account, no sign-up.
The camera feed is held in memory for the few seconds of your delay and then discarded. It is never written to disk unless you save a frame yourself, and it is never uploaded. Speech recognition is pinned to on-device — the app refuses to fall back to a server. There is no account, no analytics and no third-party SDK.