Why DOCX extraction preserves more
Copying or screenshotting an image from Word can change dimensions, color or compression. A DOCX package normally stores embedded media under word/media, so extracting those entries can preserve the exact bytes Word retained.
This is not always the file originally selected by the author. Word may have converted, resized or renamed it before storing it. The accurate claim is that extraction preserves the bytes currently inside the DOCX.
Use the local extractor
Choose the DOCX, decide whether to remove exact duplicates, then create a ZIP. The tool hashes each supported media item and writes a UTF-8 CSV manifest with internal path, candidate name, format, byte size and SHA-256 value.
- Select a document copy.
- Leave deduplication off when every occurrence matters.
- Extract and download the ZIP.
- Check manifest.csv before renaming or archiving files.
Filename limits
Word often keeps only generic names such as image1.jpg. Descriptions and drawing attributes can sometimes preserve a useful candidate, but no program can reconstruct a filename that is absent from the package. Linked web images are also different from embedded media.