How to Sign a PDF Without Printing It First

The default way people sign a PDF is to print it, sign the paper, and scan it back in. It works, and it is almost always the wrong choice. The scan turns your document into a photograph of a document: the text stops being text, the file gets larger, and the signature is now a picture of ink at whatever resolution the scanner managed.

Signing the file directly avoids all three problems. This guide explains what the tool actually writes into the PDF, what it costs in bytes, and how to check the result before you send it. It also covers what a visual signature is not — because that matters for some documents.

The measurements below were taken on 2026-10-20 in Chromium 133 on a Windows 11 machine with an AMD Ryzen 5 5600 and 16 GB of RAM, and are saved in docs/test-results/sign-pdf.json. Our tests used deliberately small synthetic PDFs, so treat the timings as a floor rather than a promise for a 200-page scan.

The three ways to sign, and what each costs

Print, sign, scan. You gain a wet signature and lose the text layer. A scanned page is an image, so anything that reads the document by extracting text — an applicant tracking system, a search function, a screen reader — may find nothing. The file also usually grows, because a 300 dpi colour scan of a page is several times larger than the original text.

Insert a signature image in a PDF editor. This keeps the original page content and adds one image object. Text stays selectable, page count stays the same, and the byte cost is small. That is what Sign PDF does.

Apply a cryptographic digital signature. A certificate-based signature can prove the document was not altered afterwards. That is a different feature with different requirements, and our tool does not do it. More on that below.

What gets written into the file

The tool renders each page with pdf.js so you can see where you are placing the signature, maps the screen position to PDF coordinates, and embeds your signature image as an object in the page’s content stream using @cantoo/pdf-lib. The rest of the document is carried over unchanged.

PDF coordinates are not screen pixels. A PDF page is measured in points, 72 to an inch, with the origin at the bottom-left corner. The tool converts your drag position into those units, which is why a signature placed near the bottom of the preview lands near the bottom of the printed page rather than at the top. If you drag a signature partly outside the visible page, the position is clamped to the page margins so it cannot be clipped off the edge.

The added bytes are the signature image plus a small amount of structure to reference it. In our three test runs, the overhead was:

• A one-page PDF at 679 bytes became 1,489 bytes — 810 bytes added, in 14 ms.
• A four-page PDF at 981 bytes became 1,794 bytes — 813 bytes added, in 3 ms.
• A 25-page, 52,440,219-byte PDF became 52,440,969 bytes — 750 bytes added, in 113 ms.

Two things are worth reading from that table. First, signing does not meaningfully change the file size; even on the 50 MB document the overhead is under a kilobyte. Second, the time depends on how much page content has to be parsed, not just on page count. The 50 MB file finished in 113 ms because its pages held simple content. A document full of scanned images will take longer.

The signature itself can be drawn with a mouse or finger, typed in a cursive style, or uploaded as a PNG or JPEG. A transparent PNG gives the cleanest edges if you already have a scanned signature; a JPEG will carry a white box around the strokes.

The workflow, step by step

1. Open Sign PDF and drop in the document. Password-protected files cannot be previewed or changed until they are unlocked; remove the password in the app that created the file, then sign.
2. Create the signature. Draw it, type it, or upload an image. Keep the box tight around the strokes so the signature does not take up half the page.
3. Choose the page. Use the page selector to move through a multi-page document. Each run places one signature image, so if you need to initial several pages, download the signed file and run it through again for the next one.
4. Position and resize it. Drag the signature box, or move it with the arrow keys for finer control. Check it sits on the signature line, not through the printed text.
5. Sign and download. The signature is embedded and the PDF is re-serialized in memory, then saved to your downloads.

If the document is several files that need to become one signed document, combine them first with Merge PDF, then sign the merged file. If the recipient’s upload form has a size limit, run Compress PDF after signing; it repacks the file structure without re-encoding your signature.

What this signature is not

This is a visual electronic signature — often called a simple electronic signature. It records that an image was placed on the page. It does not include a cryptographic certificate, a tamper-evident seal, or a trusted timestamp.

For most everyday agreements, that is exactly what is expected, and visual signatures are widely recognised under frameworks such as the US ESIGN Act and UETA and the EU’s eIDAS regulation when both parties intend to be bound. But some documents and recipients require more: court filings, notarised deeds, certain government forms, and some regulated industry paperwork may demand a qualified signature with a certificate, or a wet signature. Check the recipient’s instructions before you sign, and treat this paragraph as general information rather than legal advice.

The practical test is simple. If the other side says “sign and send it back”, a visual signature is normally enough. If they say “qualified electronic signature” or “digital certificate”, this tool is not the right instrument.

Check the result before you send it

• Reopen the downloaded file in a PDF reader and confirm the page count matches the original.
• Select a sentence. If the text highlights, the text layer survived and the document is still searchable.
• Zoom to 100% and look at the signature edges. A transparent PNG should sit cleanly on the page; a JPEG may show a pale rectangle.
• Check the placement against the printed signature line and any date field nearby.
• Open it in a second reader if the document matters. Two independent renderers catch more layout surprises than one.
• Confirm the file size fits the recipient’s upload limit, if there is one.

Why signing in the browser is the safer default

Contracts, lease forms, and permission slips are among the most sensitive documents people handle, and uploading them to an unknown signing service is a real disclosure. This tool does all the work in the tab: page rendering, coordinate mapping, image embedding, and re-serialisation. The saved test run records zero outbound network transmissions.

You can verify that yourself with the Network panel, and the stronger version is the offline test: load the signing page, disconnect the network, then sign a document. The network panel walkthrough shows the exact steps for Chrome, Firefox, and Safari.

Printing, signing, and scanning is not wrong because paper is bad. It is wrong because the scan throws away the structure of the document to capture a scribble. Signing the PDF keeps the document and adds the scribble.