PDF to PNG Converter

PNG discards nothing, so there is no quality slider to trade against and the resolution is the whole of the decision. 72 dpi reproduces the page at roughly its natural size on an ordinary screen, which suits a thumbnail and looks soft under any zoom. 144 is the sensible middle: crisp on a high-density display and still legible when enlarged a little. 216 is for printing, or for a diagram you mean to enlarge. The cost climbs with the square of the scale, so tripling the setting multiplies the pixel count by nine. Choose for the largest use you can foresee — scaling an image down later is harmless, scaling it up is not.

It depends on what is printed on the page. A page of black text on white is mostly long runs of identical pixels, which is the case PNG's compression was built for; the same page as a JPEG is often no smaller and comes back with a faint haze around every letter. A page that is edge-to-edge photography goes the other way, and JPEG will usually be a fraction of the size at a quality nobody would complain about. Scans sit awkwardly between the two: they are photographs of text, so they compress like photographs but show the damage like text.

Not from a rendered page. A PDF page has no background of its own — it describes marks to be made on paper, and the paper is taken for granted — so anything drawing it has to decide what the unmarked areas become. Here they become white, painted in before the page is drawn on top. That is correct for a document and unhelpful if what you wanted was a logo floating over a coloured slide. Where transparency matters, cut the background out in an image editor afterwards, or redraw the artwork in a vector tool; a page render will not hand you an alpha channel.

No. The file you pick is read as bytes in the page, handed to a copy of pdf.js that loads into the tab, and drawn onto a canvas your own browser owns. The images that appear are held in memory, and each Download link points at that memory rather than at any address. There is no upload step in the code and no server behind the tool. That matters more here than on most converters, because the documents people turn into page images tend to be the sensitive ones — a signed contract, a payslip, a scan of a passport. Closing the tab throws all of it away.

Because browsers cap how large a canvas they are willing to encode, and a big page at 216 dpi can go past it — an A0 poster or a wide spreadsheet export are the usual offenders. Those pages are not dropped quietly: the ones that could not be encoded are named on screen, with the advice to drop the resolution and run the file again, so a ten-page document never comes back as seven images and leaves you to spot the gap. One step down usually clears it, since it removes more than half the pixels. The limit belongs to the browser, so a page that fails in one may go through in another.

No, and no tool can arrange otherwise — an image is pixels. Characters stored in the PDF as instructions for drawing letterforms stop being characters the moment they are drawn, and nothing in the resulting file records that a shape was once the letter g. A find command turns up nothing, no line can be selected, and a screen reader has nothing to read out. If the words are what you want, a text extractor lifts them straight out of the PDF; optical character recognition can guess them back out of an image, but that is a fresh reading of the picture rather than a recovery of what was lost.

Not as it stands. An encrypted document comes back with a message saying it could not be read and that it may be encrypted, password-protected or corrupted; there is no field to type a password into. The way round it is to remove the protection first, which whoever holds the password can do by opening the file in a reader that accepts it and exporting a fresh copy. The same message covers files that are damaged or only half-downloaded, so if you are confident the document has no password on it, fetching it again is the first thing worth trying.

Because the two store completely different things. A PDF page of text is a short set of instructions — use this font, put these characters at these coordinates — and it stays small however many words are on the page. A render throws all of that away and keeps a grid of pixels covering every part of the sheet, the blank parts included. A twenty-page report of a few hundred kilobytes can therefore produce twenty images that dwarf it. Nothing has gone wrong; size follows pixel count, so the resolution setting is the only lever you have.