From video

How to screenshot subtitles from a video

Pausing and capturing by hand is the slow way. There are three ways to get the lines out, and one of them skips the screenshots entirely.

Method 1 — pause and capture

Play the clip, pause on each line, take a screenshot, seek forward, repeat. On Windows that is Win + Shift + S; on macOS it is Cmd + Shift + 4; on Android and iOS it is the power-and-volume shortcut.

It works everywhere and needs no software. It is also the slowest option, and it is the one that produces mismatched frames: you resize the window once, or the player goes full-screen halfway through, and now the frames do not line up.

Two things make it less painful. Full-screen the player first so the chrome stays out of every capture, and lock the window size before the first screenshot so it does not change under you.

Method 2 — a screen recorder, then frames

Record the clip, then pull stills from the recording with a video editor or ffmpeg. You get consistent frame sizes for free, because every still comes from the same recording.

You still choose each frame by hand, and you still have to find the right moment for every line. For a long scene that is a lot of scrubbing.

Method 3 — let a tool extract them

QuoteStitch takes the video file directly. Drop in an MP4, MOV or WEBM and it samples up to 320 points across the clip, finds the rows carrying dense horizontal edges, and keeps one frame per line of dialogue. You never pause, never capture, and never align anything.

It works on the picture, not on the words. Burned-in subtitles are sharp near-vertical strokes, so the rows that carry them are detectable even without reading a character. Two different lines share a brightness envelope but not a glyph pattern, so the shape of the strokes decides where one line becomes the next. Camera motion does not confuse it.

• Video inMP4, MOV, WEBM — anything the browser can play
• Images inPNG, JPG, WEBP
• Long image outPNG at 1x or 2x, at original width or 720 / 1080 / 1440 px
• SamplingUp to 320 points across the clip, so seek time stays bounded
• LimitNo fixed image count. Practical limits come from device memory.
• NeedsSubtitles burned into the picture, not a separate .srt file

What this cannot do

• Subtitles in a separate file. If the subtitles are a .srt or .vtt file rather than burned into the picture, there is nothing on screen to detect. Use a subtitle file tool.
• Streaming services. A browser tool can only read a video file you give it. It cannot capture a protected stream playing in another tab, and it cannot bypass DRM.
• Text you can copy. The output is an image, not text. If you need the dialogue as a transcript, this produces the wrong artefact — you want speech-to-text, not frame extraction.
• Every line, exactly. Sampling 320 points across a clip means very fast dialogue can be missed, and two lines that appear in the same place with a similar glyph pattern can be read as one. Check the result against the clip.

Frequently asked