Convert DOC to RTF

This conversion is a response to a rule rather than a preference, and the rule is usually invisible until it fires. Word says the file type is blocked by a policy setting. An email arrives with the attachment removed and a note from a security appliance. An upload form lists its accepted formats and `.doc` is not among them. An e-discovery or records intake process normalises everything it receives and specifies rich text.

None of that is a judgement about your document. The binary DOC format is an OLE compound file — a small filesystem inside a file, with named streams that can hold a VBA project, embedded objects and a good deal else — and for two decades it was the most reliable way to deliver something unwanted into an organisation. Blocking it wholesale is a coarse defence that works, and it is not going to be lifted for one submission. RTF is on the other side of that line because it is characters, and characters cannot execute.

It is worth being clear that this is not modernisation. The registry dates both formats to 1987 and marks both legacy, and Microsoft stopped developing the RTF specification in 2008. You are trading a binary container for a text one of exactly the same vintage, and the trade only makes sense because the destination cares about the container.

If nothing is refusing the file, this is the wrong page. A document you intend to keep working on should become a DOCX, which is a live format with a validated structure; a document that just has to be read should become a PDF, which fixes the layout and cannot be altered by accident. Choose RTF when a specific system has named it, or when the recipient has no office suite at all and WordPad is what they have. Outside those two cases the other targets are better and this site has pages for them.

The two formats fail in opposite directions and both failures are worth knowing about. A binary DOC is a structured storage file: streams, storages and a header, with the text held in a piece table that reconstructs the document from fragments. It is dense, efficient for 1987 hardware, and completely opaque without a library that understands it.

An RTF is the opposite in every respect. Open one in a text editor and the prose is legible, wrapped in backslash-prefixed control words for the font table, the colour table, the style sheet and each run of formatting. That transparency is why an RTF survives being pushed through systems that mangle binary content, and why a broken one can often be repaired by hand. It is also why nothing checks it: a DOC with a damaged stream refuses to open, while an RTF that is slightly wrong opens slightly wrong and says nothing. A clean-looking render is not evidence of a clean file.

Most of it, because RTF was built to carry exactly this. Named paragraph styles come across as styles rather than as manual formatting, since RTF has a real style sheet. Tables keep their cells, borders and merged regions; lists keep their numbering; headers, footers, page numbering, footnotes, endnotes, hyperlinks, bookmarks and page setup all have direct representations on both sides.

The conversion runs through LibreOffice, which reads the binary DOC format and writes RTF as first-class filters, and the machine it runs on carries metric-compatible substitutes for Arial, Times New Roman, Courier New, Calibri and Cambria. Those have the same character widths as the originals, so line breaks land in the same place and a document laid out to fit a page still fits it. What neither format can do is carry a font with the file — both name fonts and stop there — so anything drawn in an unusual face was already being substituted by whatever opened the DOC.

Old Word files accumulate things their authors never intended to send. The document properties hold the original author, the company the licence was registered to, the total editing time and often a chain of previous file paths. Older versions of Word wrote fragments of deleted text into the file, and a document that has been through fifteen revisions can contain sentences nobody meant to keep. Embedded objects — a linked spreadsheet, a chart pasted from Excel — can carry data well beyond what is visible on the page.

A trip through RTF strips a good deal of that incidentally, because there is simply nowhere in the syntax to put an OLE stream or a compound-file property set. That is a reason some intake processes ask for rich text in the first place. It is not a reason to trust the result blindly: read the converted document before you send it, particularly if the DOC is old and its history is unknown to you.

The RTF syntax has no place for a VBA project, so any automation in the document is left behind by definition rather than by policy. On a 1990s DOC that is more likely than people expect: macro-bearing templates were normal office practice, and a document created from one inherits whatever the template carried.

If Word has ever shown you a security warning when opening this file, that is what it was about, and converting is the cleanest possible answer to it. The words, the tables and the layout arrive at the other end; the code does not, and cannot. For a filing or intake system that accepts inert documents and rejects containers that can execute anything, this conversion is the response that respects the requirement instead of arguing with it.

RTF has no compression whatsoever and stores an embedded picture as hexadecimal text — roughly two characters in the file for every byte of image. A DOC that is mostly prose grows modestly. A DOC with a letterhead logo on every page, a scanned signature and two photographs can grow several times over.

That matters more here than on most pages, because the reason for converting is often an attachment or upload limit in the first place. Clearing a policy block and then failing a size check is a poor trade. If the document is illustrated, reduce the images in Word before converting rather than after — the cost of a picture is paid in full, and a decorative letterhead repeated on every page is the first thing to remove. A scanned page pasted into a document is usually the single largest item in it.

On our server, not in your browser. Reading a binary Word file and writing rich text needs an office engine, and this page says so plainly rather than implying that everything on this site runs locally. The file goes over an encrypted connection to a container running LibreOffice headless, and the container has no outbound internet access of its own.

Each job gets its own scratch directory and its own LibreOffice profile — two `soffice` processes sharing one profile corrupt it, and the symptom is a conversion that never returns rather than an error — and the directory is deleted when the job ends regardless of the outcome. Anything still running after sixty seconds is terminated, which matters for old documents in particular: LibreOffice tends to stop rather than fail on a malformed binary file. A document that asks for a password fails with a message instead.

Open the result and check it before submitting, and open it in whatever most resembles the destination rather than in whatever is convenient. A rich text file rendered by Word and the same file rendered by an intake system’s own parser can differ, and the parser is the one that decides whether the submission is accepted. WordPad is a useful proxy for a minimal renderer precisely because it supports so little.

Look at the tables first, then at anything that used to be a field — a date that updated itself, a page count, a cross-reference — because each of those resolves to the value it held at the moment of conversion and stops being live. Then check the first and last page for the ordinary damage a thirty-year-old file accumulates. Thirty seconds there is cheaper than a rejected filing and a resubmission window, and it is the only check that tells you anything the conversion itself cannot.