Pick a random, company, or custom question and set the prep and answer seconds; the clock runs as in the real thing, with no extension.
Use the framing guides to put the lens at eye level with the light in front of you; the preview is mirrored, the replay is what others see.
Recording starts when the prep countdown ends and cannot be paused or retaken; recovering after you stall is the part this trains.
Watch it back from the interviewer's side, score five items, and change only the lowest one; recordings stay on this device and nothing is uploaded.
Start the camera to switch devices
Recorded interviews reject people over gaze, pace, and posture — things you cannot see in yourself. Run one take under real conditions, then check the replay from the interviewer's side; the gap only shows on playback.
Watching yourself on screen reads as downcast eyes on their side. Treat the lens as their eyes — you can hide your self view once recording starts.
The preview is mirrored; the recording is not. Left and right are only true in playback — always review from the replay.
The real thing has no pause and no retake. Let practice run to the end too — recovering mid-stumble is exactly the skill to train.
A high angle reads arrogant, a low one cramped, and backlight sinks your face. Line up with the framing guides once before recording — three seconds.
Pick your lowest self-rating and fix only that on the next run. One change per take shows a curve in three takes; changing everything changes nothing.