How to convert a PDF to JPG

There is no Save As that does this, in any program, on any operating system. That is not an oversight, and knowing why takes about thirty seconds — after which most of the ways this goes wrong stop being possible.

Why can’t I save my PDF as a JPEG?

Because the two formats hold different kinds of thing. A PDF holds an ordered run of pages, each with its own physical dimensions. A JPEG holds exactly one raster image and cannot express a second. A forty-page contract therefore has no JPEG form at all, and a program asked for one would have to invent an answer. First page only? All forty stacked into an absurdly tall strip? Most decline rather than guess, and the refusal is correct.

What conversion means here is rendering. Something paints a page onto a grid of pixels — precisely what your reader does every time it puts the document on screen — and encodes that grid as a JPEG. Forty pages produce forty files. That is the procedure, not a workaround.

Two things follow, and between them they explain nearly every disappointing result. Somebody has to choose how many pixels the grid holds. And a painted page is a picture of words rather than words.

That second one matters before you go further. If what you want out of the document is the text, converting to JPG walks you away from it: a picture of a sentence is where a sentence stops being machine-readable. The characters are already sitting inside most PDFs and can be pulled out directly, and copying text out of a PDF covers why that misbehaves. Render to an image when you want the page’s appearance, not its contents.

Resolution is the setting that decides everything

Pages are measured in inches, images in pixels. Dots per inch is the exchange rate, and it is what you are really picking when a converter offers you Low, Medium and High.

Page, at this density	Output size	Megapixels
A4 at 72 DPI	595 × 842 px	0.5
A4 at 150 DPI	1240 × 1754 px	2.2
A4 at 300 DPI	2480 × 3508 px	8.7
US Letter at 300 DPI	2550 × 3300 px	8.4

Density doubles, area quadruples: at 300 DPI a sheet carries roughly seventeen times the pixels it does at 72, and nothing about the document changed. 72 matches a monitor showing the sheet at its true size — fine for a contact sheet, visibly soft anywhere else. 150 is the sensible middle: legible on paper, still sharp when someone zooms or opens it on a dense phone screen. 300 is print density, what a printer asks for and overkill for anything only ever seen on a screen.

File size tracks pixel count loosely rather than exactly, since JPEG spends its bytes on detail. Ignore the labels and work backwards from the destination: a slot on a web page 900 pixels wide does not need 2480.

JPEG blurs text, which is what the page is usually made of

JPEG’s compression was calibrated on photographs, and discards the fine variation eyes forgive in a picture of a face. A letterform is not that. The edge of a printed character jumps from black to white inside a single pixel, and JPEG represents each small block of an image as a sum of gentle waves. Gentle waves cannot produce an abrupt jump without overshooting on both sides of it, and the overshoot lands in the file as grubby speckling that clings to the outline of every character and every rule in a table. It has a name — ringing — and it is easy to miss at page-fit zoom and impossible to miss on paper.

The rule is therefore about content, not preference:

• Text, tables, diagrams, line art, screenshots of software → PNG. It loses nothing at all, and its compression is built for flat areas and hard edges — what a document page is made of.
• A page that is a photograph edge to edge → JPEG. Continuous tone is the case JPEG was built for, and PNG would produce a much larger file for no visible gain.

PDF to PNG is this same rendering step with a lossless encoder on the end, and for most documents it is the format you actually wanted. If something downstream insists on JPEG, push the quality setting high — and notice that where ringing around text finally disappears, the JPEG is often larger than the PNG would have been.

There is no transparency in a JPEG, and the fallback color matters

A JPEG stores color channels and nothing else — no alpha, no way to say “nothing here”. A PDF page frequently has no background drawn in it; the crisp white sheet in your reader is painted by the reader, not stored in the file.

Rendering therefore has to composite the page onto something, and which something is a choice made for you. White is nearly always right. Black is what a few tools and several command-line pipelines produce, and a page composited onto black is not subtly wrong: dark text on a dark field is an unreadable slab. Check the first file before converting the rest, and look at the margins, where nothing was drawn and the fallback shows.

If the PDF is a scan, converting gains you nothing

Plenty of PDFs are wrappers around photographs: a contract someone scanned, a receipt shot with a phone, a fax export. Each page there is already a compressed image, usually already a JPEG, and converting decodes it and compresses it again. That is generation loss, and worse than repeating the first pass’s damage, because the second encoding rarely aligns its blocks with the first one’s and fresh errors land on top of the inherited ones. Turning the density up rescues nothing — asking 600 DPI of a page scanned at 200 gives a larger file holding the same blur, enlarged.

Telling the two apart takes one gesture: zoom to 400 percent. Text that stays crisp is drawn from vectors and renders beautifully at any density. Text that dissolves into soft pixels is a photograph of text, and the original scan is the ceiling.

Where to actually do it

In a browser. Every major browser displays PDFs and prints them, and none of them exports a page as an image. What works is a converter running inside the page, doing the same rendering the viewer does — which also means the file never leaves your machine. The PDF to JPG converter here does exactly that, showing the pixel size of every page before it runs.

On a Mac. Preview exports: open the PDF, select the page in the thumbnail sidebar, then File, Export, and choose JPEG. The sheet offers a quality slider and — because the source is a page and not an existing bitmap — a resolution in pixels per inch, which is the decision from earlier. It handles one page at a time. For a whole document, Automator provides an action that renders PDF pages as images, with format and resolution settings, and saving it as a Quick Action puts it in the Finder’s right-click menu.

On Windows. Nothing built in does this. The Snipping Tool captures at screen resolution, so a page fitted to a window lands well below 150 DPI and prints badly. GIMP, which is free, imports PDF pages and asks for the resolution up front, which makes it a real option for a few pages. Otherwise most people use the browser route.

On a phone. Yes, this works on a phone, and the browser route behaves identically on iOS and Android without installing anything, precisely because the rendering happens in the browser rather than in an app. Screenshotting a page also produces a JPEG, but at screen resolution and with toolbars in shot — a way to share a glance, not a conversion.

What comes out is bigger than what went in

Forty pages at 300 DPI is forty multi-megabyte files, and mail servers reject attachments over limits nobody thinks about until the bounce arrives. Three ways out, in the order they are usually right: drop the density, since 150 cuts the total to about a quarter; send fewer pages, because the request that started this is usually about two of the forty; or squeeze what you have. Compress an image to a target size works backwards from a number you name, and how to reduce file size covers the same problem across formats.

Then consider not converting at all. If the recipient can open a PDF, the PDF is smaller than its own rendered pages, prints better, and still holds selectable text. Convert when the destination wants an image — and then pick the density on purpose.