Converting PDF Pages to Images Without Uploading Anything
Why DPI matters more than the file format you pick, when JPG beats PNG for a scan, and how a browser rasterises a PDF entirely on its own.
Most reasons to turn a PDF into images are mundane. A form only accepts JPG uploads. You want one page in a slide deck. A client asks for “just a picture of the invoice” because opening PDFs on their phone is a hassle. The PDF to JPG Converter does it in the browser tab you already have open.
The part worth knowing is that the usual sites for this take your document, upload it, process it on a server, and promise to delete it in an hour. For a takeaway menu, fine. For a signed contract, a medical result, or a scan of your passport, that promise is the entire security model.
DPI matters more than the format
Nearly every complaint about converted pages — blurry text, jagged table borders, unreadable footnotes — comes from resolution, not from JPG versus PNG.
PDF pages are measured in points, 72 to the inch. A standard A4 page is 595×842 points. Convert at 72 DPI and you get a 595×842 pixel image, which is smaller than a phone screen. The layout is recognisable; the 9pt fine print is mush.
Three settings cover almost everything:
- 72 DPI for a thumbnail or a preview grid
- 150 DPI for anything read on a screen — A4 becomes about 1240×1754, and small type stays legible
- 300 DPI when the image will be printed or fed to OCR software
300 DPI on a long document is heavy. Twenty A4 pages at 300 DPI is roughly 2480×3508 each, and each one has to be drawn, encoded, and held in memory. On a phone that is where things get slow.
JPG or PNG
For a scanned page — anything that started life on a flatbed or a phone camera — JPG wins easily. The source already has sensor noise and paper texture, so lossy compression has plenty of room to work. Files come out three to five times smaller than PNG with no visible difference at 150 DPI.
For pages that are mostly vector content — text, tables, charts, diagrams — PNG is worth the size. JPG compresses in 8×8 blocks, and sharp black-on-white edges are exactly what that hurts most. You see faint grey halos around letters. At 300 DPI it stops mattering; at 72 DPI it looks cheap.
The quality slider only affects JPG. Below about 70% you start seeing the artifacts; above 90% the file grows fast for very little visible gain. 85% is a reasonable default.
How a browser draws a PDF page
A PDF page isn’t a picture. It’s a small program: move here, set the fill to this colour, draw a curve, show this string in that font at that size. Rendering means running the program.
The converter reads the page’s content stream, tracks the transformation matrix as it changes, and turns every instruction into a drawing command on a canvas. Paths, clipping regions, dash patterns, transparency, embedded images in half a dozen colour spaces — they all come through.
Text is the one place where a browser has to compromise. Rasterising an embedded font program means interpreting glyph outlines, and no browser exposes that. So each character is drawn with the closest system font and positioned individually using the width table stored in the PDF. Line breaks, column edges, and page geometry land exactly where the document put them; on pages set in an unusual typeface, letter shapes can look a little different. Scanned pages sidestep this entirely — they’re just images, and they come out pixel for pixel.
What it won’t do
Encrypted PDFs are rejected outright rather than producing a stack of blank pages. Two image formats also stay undecoded: JPEG 2000 and CCITT fax encoding, which turn up in old scanner output. If every page comes back blank, that’s almost certainly why.
Everything else — the page count, the render, the JPEG encoding, the ZIP of all pages — happens on your machine. Load the page, disconnect from the network, and it still works.
Drop a file into the PDF to JPG Converter and pick 150 DPI unless you know you need more.