แก้ไข PDF

เครื่องมือนี้ทำอะไร

This adds new content to a page: a text box, an image, a rectangle, an ellipse, a line or a white patch, placed where you put it. What it cannot do is re-type the words already there. A PDF stores glyphs at coordinates in a font usually subset to the characters it already uses, so turning March into April may need a letter A that is not in the file at all. Anything you add is written into the page's content stream.

Reach for it to date and initial a printed form, to drop a logo onto a cover page, to write a correction beside a wrong figure, or to cover a phone number before a document goes further round the office.

วิธีทำงาน

1. Drop the PDF onto this page. It is rendered for you and you work on the page you can see.
2. Choose what to place - text, image, rectangle, ellipse, line or whiteout - and drag out the area it should occupy.
3. Set what that item needs: size, colour, opacity and rotation for a shape; family, weight and alignment for text.
4. Carry on across as many pages as you need. One pass takes up to 2,000 items.
5. Press Apply edits and the download starts on its own.

Every tool that appears to edit existing PDF text does one of two things: it rebuilds the page from a guess at its structure, which moves the layout, or it covers the old text and draws new text over it. This tool does the second, openly. For genuine paragraph editing, convert with PDF to Word, edit there and convert back - a route that re-flows the document rather than reproducing its pagination.

A whiteout patch covers text. It does not delete it. The words underneath stay in the content stream and can still be selected, copied and searched, so a covered bank detail is a bank detail you have published with a rectangle over it. Redact PDF rebuilds the page from pixels, which is the only version of this that removes anything.

Embed the font decides which faces your text is drawn in. On, it uses Noto Sans and Noto Sans Mono - about 2 MB, served from this site's own origin and cached after first use - which covers accented Latin, Greek and Cyrillic. Off, it uses the standard faces every reader already has, which adds no bytes but can only encode WinAnsi. Neither set covers CJK, Arabic or Indic, and characters outside the set are folded to the nearest the face can draw rather than aborting the job.

Images are placed at the size of the box you draw, so a 2,000-pixel logo does not arrive 2,000 points wide. PNG and JPEG are accepted and nothing else, and transparency is kept, which is what makes a PNG signature sit on the page rather than in a white square. The image is read from your disk locally and embedded in the browser; it is not posted anywhere first.