Loading corpus...
About & Help
About: This tool searches transcripts from Swedish historical newsreels (SF Veckorevy). The corpus contains two types of transcripts: Speech (automatic speech recognition) and Intertitle (OCR from silent film text cards).
Search: Enter keywords to search across all transcripts. Enable Fuzzy search to find approximate matches (useful for OCR/ASR errors like "Stockho1m" matching "Stockholm").
Filters: Narrow results by transcript type, collection (sf, sj, kino, nuet), or year range. Videos with unknown years are listed under "XXXX".
Viewing: Click a result to see both transcripts side-by-side with timestamps. Links to the original source files on GitHub are provided in the viewer header.
Sharing: The URL updates as you search, so you can bookmark or share any search or specific video.