Add a Signature to an Image or PDF

What this tool does, and what it deliberately does not

This is a way to put your signature onto a document without printing it, signing it by hand, and scanning it back in. You make a signature once — by drawing it or by uploading a photo of one you already signed on paper — and then drop it onto an image or a PDF, exactly where it needs to go. The result is the same as a signed printout: your mark is visibly on the page.

It is worth being clear about the ceiling. A signature placed this way is a picture of your signature on a document. It is not a certificate-based electronic signature — there is no cryptographic key tying it to your identity, no tamper-evident seal that breaks if the file is edited afterwards, and no audit log of who signed and when. For the large majority of forms, consent slips, letters, and internal sign-offs, a visible signature is exactly what is being asked for. For a contract that explicitly demands a qualified or notarised e-signature, you want a service built around identity verification instead. Saying so up front is more useful than implying a level of legal weight the tool cannot provide.

Drawing versus uploading a photo of your signature

Drawing is the clean path. The pad captures your strokes on a transparent canvas, so what you hand to the document is ink and nothing else — no background to remove, no paper texture, no shadow. On a phone or tablet you are writing directly with a finger or stylus, which usually produces a better signature than fighting a laptop trackpad. Pick a pen colour and thickness before you start; black reads as the most neutral, but blue is the classic "this is an original, not a photocopy" choice.

Uploading a photo is the path when you already have a signature on paper. Here the tool has to separate ink from paper, and it does that with a brightness test: it walks every pixel, and any pixel brighter than a fixed threshold is made fully transparent while darker pixels are kept as ink. That is a deliberately simple rule, and its limits follow directly from how it works. Dark ink on bright white paper separates cleanly. A dim photo, grey or cream paper, or uneven lighting muddies the gap between paper and ink, and some paper survives as a faint haze or some ink gets eaten. The fix is on the input side: a brighter, higher-contrast, evenly lit photo. The crop step exists for the same reason — trim the box down to just the signature so the hand, the pen, and the edge of the desk never reach the brightness test at all.

The threshold is a hard cutoff, not a gradient, and that explains the two ways it disappoints. Because a pixel is either kept whole or dropped whole, there is no partial transparency to soften the stroke edges, so a heavily zoomed signature can look slightly jagged where ink meets paper. And because the same single cutoff is applied to the entire image, a photo that is bright in one corner and shadowed in another cannot be satisfied everywhere at once — lift the cutoff to clear the shadowed paper and you start eating the faint tail of the ink; lower it to preserve the ink and the shadow survives as a grey smear. This is the built-in limit of a global threshold, and it is exactly why an evenly lit source beats a powerful setting.

Images flatten to PNG, PDFs stay PDFs

The two document types are handled differently on purpose. When you sign an image, the document and the signature are composited onto one canvas at the image's full resolution and exported as a PNG. PNG is lossless, so the signature edges stay crisp and the page text is not chewed up by compression the way a JPG re-save would do it.

When you sign a PDF, the tool keeps the original PDF and embeds your signature into it as an image on the pages you signed. Every page is preserved — a five-page contract comes back as a five-page PDF, with the signature stamped only where you placed it. It is not rasterised into a single flat picture, so selectable text and the document's page structure survive. Each page carries its own signature placement, which is why you can sign the signature block on page one and an initials line on page four independently, and leave everything else untouched.

To render the PDF for you to look at, the tool uses a PDF engine that is served from this site itself, not fetched from an outside CDN. That is a deliberate choice: it means opening a PDF here makes no third-party request, so the "nothing leaves your browser" claim holds for PDFs and not just images.

Placing it so it looks signed, not pasted

Drag the signature to move it, pull any corner to resize it, and use the handle on the stalk above it to rotate. A real signature is rarely dead level and rarely huge, so a slight rotation and a size that matches the surrounding text usually reads as genuine, while a giant perfectly-horizontal signature reads as pasted-on. Position matters too — most forms have a signature line, and sitting the ink just above the line rather than centred in the box is what a hand would actually do.

Two practical notes. If part of the sensitive area of a document needs to stay hidden rather than signed, that is a job for a redaction tool, not this one. And if your signature photo has a wide messy border you could not crop away here, the crop tool gives you a full-screen crop before you ever bring it in.

Before you send the signed file

Open the download and look at it once. Check that the signature is on the right page and over the right line, that resizing did not push part of it off the edge, and that an uploaded signature's background actually cleared instead of leaving a pale rectangle. If a signed PDF looks larger than you expected, that is the embedded signature image adding weight; running it through the image compressor first — before signing, on the source image — keeps the final file lean. The document you send is the export, not the preview, so it is the export that deserves the last look.