Converting a PDF to an image (JPG, PNG)
JPG or PNG, 150 or 300 DPI, text that turns into pixels: the settings to get right before you export the pages of a PDF.
A PDF page you want to drop into a slide deck, a contract preview to display on a website, a chart to send to someone who will reply ‘I can't open it': in these situations PDF is the wrong format, and an image does the job. What remains is deciding what comes out — JPG or PNG, at what resolution, and what you agree to lose.
What the conversion destroys along the way
A PDF stores text as vectors: letters described by curves, crisp at any zoom level, selectable, copyable, indexable and readable by a screen reader. An image contains nothing but pixels. The conversion only goes one way, and here is what disappears:
- Ctrl+F no longer finds anything in the document;
- no text can be copied — you will need OCR to get it back;
- clickable links, bookmarks and form fields are lost;
- sharpness is frozen at the resolution you chose: zooming past it only enlarges the pixels.
That is only a problem if you wanted to keep a document. For a visual — a thumbnail, a screen grab, an article illustration, a quick attachment — it is precisely the point.
JPG or PNG: look at what is on the page
The choice depends not on habit but on what the page contains.
| Page content | Format | Why |
|---|---|---|
| Text, tables, diagrams, flat-colour charts | PNG | Lossless compression: no artefacts around the letters. On a mostly white page, PNG is often smaller than JPG. |
| Photos, gradients, 3D renders, colour scans | JPG | Lossy compression is built for continuous tones. A PNG of the same page is easily five to ten times bigger. |
| An element to sit on a coloured background | PNG | JPEG has no alpha channel: transparency turns white. |
| A page to attach to an email, under 1 MB | JPG at quality 80 | Best size-to-legibility ratio when the page mixes text and visuals. |
The classic JPEG trap on text comes down to chroma subsampling (4:2:0): colour is encoded at half resolution in both dimensions, so a quarter of the luminance samples. On red, blue or green text sitting on white, that produces a smeared halo visible as soon as you zoom to 100%. Black on white holds up better, but ringing — the greyish halo that ripples along the edges — stays noticeable below quality 85.
WebP, for its part, does both — lossy or lossless, transparency included — for 25 to 35% less than an equivalent JPEG. Excellent for the web, best avoided the moment the file has to be opened by someone else in older software.
DPI, the one setting that genuinely changes the result
A PDF page has no dimensions in pixels: it has physical dimensions. A4 measures 210 × 297 mm, or 8.27 × 11.69 inches. It is the export resolution, in dots per inch, that decides how many pixels you end up with:
- 72 DPI → 595 × 842 px. Thumbnails and list previews only; body text becomes unreadable.
- 150 DPI → 1,240 × 1,754 px. The sensible default for a screen: text at 10–11 pt stays legible and the page comes to a few hundred kilobytes.
- 300 DPI → 2,480 × 3,508 px. For print, or for re-reading small-print clauses. Expect 1 to 3 MB per page.
- 600 DPI → 4,960 × 7,016 px. Reserved for very fine line work and for OCR on difficult documents. Past that, there is nothing left to gain.
Two important caveats. If the page came from a scan, it already holds an image frozen at a given resolution, often 200 or 300 DPI: exporting at 600 DPI merely interpolates the pixels that are already there — four times as many pixels as at 300 DPI, a much bigger file, and not one extra detail. A fully vector page, by contrast — text and graphics generated by software — really does gain sharpness at every step up.
Converting in practice
The PDF to JPG converter handles the whole document: you drop in the PDF (up to 100 MB, the limit shown on the tool page), every page is rendered, and you get back a ZIP archive holding one image per page, numbered in order. A 40-page PDF gives 40 files; simply pull out the one or two you need.
For PNG, run the resulting images back through the image format converter, which handles JPG, PNG, WebP, GIF and BMP. Bear in mind that a JPG converted to PNG does not recover what the compression threw away: it freezes the artefacts into a lossless format. Converting to PNG is therefore mainly justified when the destination demands it (a tool or CMS that refuses JPEG, or the need for a lossless format for later editing) — otherwise you are better off aiming for a higher DPI at export, so the artefacts stay invisible at display size.
Two finishing touches come up almost every time:
- You only want one extract — a table, a signature, a chart: crop the image rather than publishing the whole page.
- The image is going on the web or out by email: the image compressor brings an A4 page at 150 DPI under 300 KB at quality 80, with no visible difference on screen.
The pitfalls that crop up most often
Fonts that are not embedded
A PDF can reference a font without carrying it. The rendering engine then swaps in a lookalike whose character widths differ: lines that overflow, line breaks in the wrong place, table columns knocked out of alignment. The document looked right on your own machine simply because the font is installed there. Always check the resulting images before you send them out.
Colours in print-ready PDFs
A file prepared for a commercial printer is in CMYK. Moving to RGB shifts the saturated shades: oranges and deep blues go flat, and a ‘rich black' built from all four inks turns greyish-brown. For a visual sign-off, warn whoever is looking at it; for a file that has to go to press, do not convert.
Forms, annotations and protected documents
Values typed into a form that has not been flattened, along with certain annotations and highlights, may not show up in the render. Flatten the PDF before converting. An encrypted document, meanwhile, will be rejected until its password has been removed.
Metadata
Title, author, the software that created it and creation dates all stay in the PDF; the image does not carry them across. That is rather good news when you are sending a page outside your organisation — but do not count on conversion to anonymise anything: whatever is written on the page stays perfectly visible.
When you are better off not converting
- You only want to pull out a few pages: split the PDF and keep it as a PDF. The text stays selectable and the file stays small.
- The PDF is too big to send: PDF compression shrinks the images inside without turning the text into pixels. Converting every page to an image and then rebuilding a PDF from those images almost always gives a bigger file, one that is unsearchable and looks scanned.
- You need the text back: aim for a conversion to an editable document instead. Going via an image forces you to run OCR afterwards, with the errors that implies.
If the result is not what you expected
- Blurry images: DPI too low, or a page already scanned at low resolution. Zoom to 400% on the original PDF — if it is blurry there too, no export setting will rescue it.
- Coloured, smeared text edges: JPEG artefacts. Push the quality above 85 or switch to PNG.
- An outsized archive: 300 DPI across 80 pages easily tops 150 MB. Drop back to 150 DPI if the destination is a screen.
- A blank or entirely black page: transparency or unusual layers the renderer has misread. Re-print the PDF ‘to PDF' from your PDF reader to flatten it, then convert again.
- File rejected: the PDF is over the size limit shown on the tool page, or it is protected by a password.
Rendering vector pages calls for a library the browser does not ship with: PDF to image conversion therefore runs on our servers, and the files you upload are deleted there after an hour. One conversion a day is free with no sign-up; beyond that, the single subscription at €7 a year opens up every server-side tool. The touch-ups that follow — cropping, resizing, compression, changing format — still run entirely on your own device.