A stamped signature is an image, not a certificate
This tool draws a picture of your signature onto a page. That is the whole mechanism, and being clear about it up front saves a lot of confusion later.
There is a second, entirely separate thing also called signing a PDF: a cryptographic digital signature, where a certificate issued to you is used to compute a hash of the document and store it in a signature field, so that any later change breaks the seal and a reader can display who signed and when. This tool does none of that. It adds no certificate, no hash, no timestamp and no tamper detection. Somebody with a PDF editor can move your signature image, copy it onto another document or delete it, exactly as they could with a signature you wrote in ink and scanned.
That is not a criticism of either approach, because they answer different questions. A great deal of routine paperwork simply needs to show that a named person applied their mark: internal approvals, delivery notes, school forms, freelance agreements, rental paperwork. Some filings genuinely require a certified or qualified electronic signature and will not accept an image. The practical rule is to find out which one the other side expects before you send anything, rather than after.
How the document and the signature are told apart
The upload control here accepts a mixed set of extensions, which is unusual: .pdf plus .png, .jpg and .jpeg. That is deliberate, because one signing job needs two different kinds of file at once.
Sorting them out is done purely by file extension. Every file whose name ends in .pdf is treated as the document, and every file whose name ends in .png, .jpg or .jpeg is treated as the signature. There must be exactly one of each. Two PDFs, two images, a missing image or a missing document all stop the run before anything is drawn.
Because the split is by extension and not by inspecting the bytes, order does not matter at all, and you can drop both files at once or add them one at a time. It also means the extension has to be honest: a PNG that somebody renamed to .jpg will be handed to the JPEG decoder and fail.
Where the signature lands on the page
Three options control the placement, and each maps onto a specific piece of arithmetic.
Page to sign (0 = last page) counts from 1, and the default of 0 means the final page, because that is where a signature block usually sits. A number larger than the document is quietly clamped to the last page instead of erroring.
Signature position offers five anchors: bottom right (the default), bottom center, bottom left, top left and top right. Each one keeps a fixed margin of 24 points, a third of an inch, from whichever edges it is anchored to. Bottom center is centred horizontally and sits the same distance up from the bottom edge as the others.
Signature width (% of page width) accepts 10 to 70 and defaults to 30, so on A4 the default signature is about 2.5 inches wide. Height is derived from the imageβs own proportions, never squashed. The image is drawn on top of whatever was already on the page, so it will cover a printed signature line rather than sitting behind it.
Signing the document
- Have your signature ready as a PNG or JPG image, cropped closely around the strokes.
- Drop the PDF and the signature image into the box above together, or click Choose files. They can be added one after another.
- Set Page to sign (0 = last page), pick a Signature position and set Signature width (% of page width).
- Click Sign a PDF.
- Save the result, which comes back with
-signedadded to the original name.
Getting a signature image worth stamping
The quality of the finished document depends almost entirely on the image you feed it. Two routes work well.
The quickest is Signature Generator, where you draw with a mouse, a stylus or a fingertip, or type your name in a handwriting face, and download a transparent PNG ready to use here. Transparency is the reason to prefer it: the strokes sit directly on the page with no rectangle around them.
The traditional route is to sign a blank sheet in dark ink, photograph it in even light, and crop hard so the picture is almost all signature. Crop matters more than resolution, because the width percentage scales the whole image; leave half a page of white around your name and the name itself ends up a fraction of the size you asked for.
What the signed PDF does not contain
Interactive form fields are untouched. The stamp is drawn onto the page content, not into a field, so a document you filled in with Fill PDF Form keeps its fields live and editable after signing. When you are finished with the whole document, Flatten PDF burns those field values into the page so nobody can alter them afterwards.
Pages carrying a rotation flag are the one placement case that misbehaves. The drawing uses the pageβs underlying coordinate box, which ignores that flag, so a scan that a viewer displays upright can put your signature along an unexpected edge. Rotate PDF adds to that flag rather than replacing it and never rewrites the page content, so bring the flag back to zero there by picking the complementary angle, 270 for a page marked 90, and sign the result.
Everything runs inside the tab. That is worth more here than on most tools, because signing means handing over two things at once: the document and an image of your actual signature, which is the pair somebody would need to reproduce your mark on paperwork you never saw. Neither file is part of any network request. Further reading: the best PDF tools, compared.