Why extract frames instead of taking screenshots?
A screenshot captures your screen, not the video: you get the player controls, whatever resolution your display happens to be, and compression artifacts from the playback surface. Extracting the frame reads the decoded video data directly, so the still is the actual pixel content at native resolution, with no UI in the shot.
The frame-step buttons matter more than they look. The difference between a great thumbnail and a blinking one is often a single frame, a thirtieth of a second, which is impossible to hit reliably with a pause button.
What makes a frame worth extracting
For thumbnails and covers, look for peak expression and mid-motion: mouths open mid-word read as energy, gestures read as story. High contrast and a readable subject at small sizes beat busy wide shots. Grab three or four candidates and compare them at thumbnail scale before choosing.
Cover images decide clicks, but retention decides reach, and that is the edit itself. Wavcut automates the retention half on iPhone: tight pacing from silence cuts, burned-in captions for muted viewing, and b-roll at the right beats. Post the extracted frame as the cover, and let the edit hold the viewer.