Choose what the screenshot should demonstrate
Decide whether the task is showing the video frame alone, the subtitle presentation or the visible player state including an on-screen overlay. Those are different evidence needs. Pause at a permitted non-sensitive sample so that comparisons concern the same moment.
The stable manual distinguishes original-resolution video capture from scaled playback capture and window capture. It also qualifies exact behaviour by video output. A saved image with different dimensions or overlays should therefore be interpreted with its mode and output context, not assumed to be a literal operating-system screenshot.
Inspect the actual binding and command
mpv documents separate default screenshot shortcuts, including a version without subtitles and a window-style capture including subtitles and OSD. Its screenshot command also has explicit video, subtitle, scaled and window-related flags. Check custom input configuration before assuming the key pressed used the default.
Keep rendering state separate from capture selection. If subtitles are visible in playback but absent only in a video-only capture, changing their timing or reinstalling fonts does not address the selected output mode. If the visible window itself lacks text, that is an earlier track or rendering question.
Open and inspect the saved file
Check the saved image directly for text, overlays and dimensions. For a script-driven capture, mpv’s successful command result can provide the saved filename. The screenshot directory can also differ between a command-line launch and pseudo-GUI mode, so verify the actual file rather than a similarly named older image.
Before sharing, inspect the complete image for private filenames or overlays. For help, report mpv version, video output, active screenshot binding or flags, visible subtitle state and saved dimensions. Those details distinguish deliberate capture composition, a renderer-dependent difference and a recording written somewhere other than the expected location.