Turning PDF pages into images looks like one task but is really two, and picking the wrong one is the most common reason people are unhappy with the result.
Rendering draws a page — text, lines, images, all of it — into a fresh grid of pixels at a resolution you pick. You get a picture of the page as it appears.
Extracting pulls out the photographs already stored inside the PDF, at whatever resolution they were embedded with, without re-rendering anything.
If you want the page, render it: PDF to JPG. If you want the photo that is on the page, extract it: Extract images from PDF. Rendering a page to get a photo out of it degrades the photo twice — once by rasterising it into the page grid, once by compressing the output.
DPI is the whole decision
A PDF page has a size in inches, not in pixels. DPI — dots per inch — is what converts one to the other.
An A4 page is 8.27 inches wide. So:
| DPI | Pixel width of an A4 page | Sensible for |
|---|---|---|
| 72 | ~595 px | Thumbnails, contact sheets |
| 150 | ~1,240 px | Screen reading, email, web |
| 300 | ~2,480 px | Printing, OCR input |
| 600 | ~4,960 px | Archival capture, fine line art |
PDF to JPG offers 72, 150 and 300 DPI; the 600 row is there for comparison.
Two things follow. Higher is not free — pixel count grows with the square of DPI, so 600 DPI is four times the memory of 300, and a long document at 600 DPI will exhaust a phone browser. And you cannot exceed the source: if the page is a 150 DPI scan, rendering at 600 does not recover detail that was never captured. It produces a larger file of the same blur.
Choosing the format
PNG is lossless. Every pixel survives exactly, which matters for text and line art — JPEG compression produces faint halos around high-contrast edges, and a page of black text on white is nothing but high-contrast edges. Choose PNG for documents, diagrams, screenshots and anything destined for OCR.
JPG is lossy and dramatically smaller on photographic content, where the discarded detail is genuinely invisible. Choose it for pages that are mostly photographs, and for any case where file size matters more than perfect fidelity.
A rough rule: if you would describe the page as a document, use PNG. If you would describe it as a picture, use JPG.
Step by step
- Decide what the images are for, and pick 72, 150 or 300 DPI from the table above before you start.
- Open PDF to JPG and choose the file. Pages are rendered in your browser; nothing is uploaded.
- Set the resolution and format.
- Run it, then open one full-size image and look at the smallest text you actually need. That is the only real test — a thumbnail always looks fine.
- If the set is larger than you expected, re-export at a lower DPI rather than compressing the images afterwards. Rendering once at the right size beats rendering big and squeezing.
Related jobs, done properly
- Feeding OCR: export at 300 DPI as PNG if another OCR program will read the images. OCR PDF does its own rendering, at 200 DPI (less on very large pages), so it does not need this step.
- The pictures inside the document: Extract images from PDF returns them at their own resolution, not screenshots of the page.
- The reverse direction: PNG to PDF puts exported pages, screenshots or graphics back into a document with each 8-bit PNG stored without loss, and WebP to PDF handles the format websites hand you when you save an image.
- Shrinking the PDF instead: if the goal was a smaller file rather than images, Compress PDF keeps the document a document — text stays selectable, which converting to images destroys.
That last point is worth stating plainly: converting a PDF to images to make it smaller throws away every searchable word on every page. It is occasionally what you want — flattening a document so nothing can be copied — but it is rarely the right answer to “this file is too big”.