PDFをJPGに変換
PDF to JPG saves every page of a document as a picture, at whatever size you choose. The drawing is done by pdf.js — Mozilla's own PDF engine, the same code Firefox uses to show PDFs — so pages come out looking the way a real reader draws them, not the way a simplified parser guesses.
このツールの画面は英語表記です。
Drop a PDF here
or drop anywhere on this page · Choose a file · 100 MB
以下の解説は英語のみでご覧いただけます。
How does PDF to JPG work?
Converting a PDF page to an image means rendering it: running the page's content stream — its text drawing operators, its vector paths, its embedded images, its transparency groups — and painting the result onto a canvas. That is a large piece of software, and writing a second-rate one is how pages come out with the wrong fonts.
The renderer is the one in Firefox
pdf.js is Mozilla's PDF implementation, and it is what Firefox uses to display every PDF you open in it. Using it means the page you get is the page a real reader draws, including the parts that are easy to get wrong: embedded font subsets, Type 3 fonts, blend modes, clipping paths and CJK text.
DPI is a real number, not a quality slider
A PDF page is measured in points — 1/72 inch — so the scale factor for a target resolution is exactly DPI/72. At 72 DPI an A4 page is 595 × 842 pixels; at 150 it is 1240 × 1754, which is readable in print; at 300 it is 2480 × 3508 and four times the memory of 150. That last point is the practical ceiling: at 300 DPI a single A4 page is about 34 megabytes of pixel data before compression, and a browser canvas has a maximum size it will silently refuse to exceed.
Each canvas is released as soon as its image is encoded, which is what lets a fifty-page document convert at 300 DPI without the tab being killed.
Why JPEG output needs a white page painted first
A PDF page has no background; the white you see is the reader drawing one. A fresh canvas is transparent, and JPEG has no alpha channel, so an unpainted canvas encodes as black. This is the cause of almost every “my converted pages came out black” report, and the fix is one fillRect before rendering — done here for JPEG and WebP, and deliberately not for PNG, where transparency is usually the reason someone picked PNG.
Choosing a format
JPEG for scanned or photographic pages, where its lossy compression is efficient and its artefacts are invisible. PNG for pages that are mostly text or diagrams: lossless, so text stays crisp, and often smaller than JPEG for large flat areas of white. WebP is usually 25–35% smaller than either at the same visual quality and is supported everywhere that matters now.
What is lost, unavoidably
An image of a page is not a page. Text is no longer selectable, searchable or readable by a screen reader; links do not work; and the file is typically several times larger. That is inherent to rasterising, not a limitation of this tool — so convert to images when you need an image, and keep the PDF when you need a document.
Input
report.pdf (12 pages) format: JPEG, 150 DPI, quality 0.9
Output
report-1.jpg … report-12.jpg 1240 × 1754 each white painted before render, canvas released per page
What options and edge cases does PDF to JPG support?
| Parameter | Type | Default | Behaviour & edge cases |
|---|---|---|---|
| JPEG | format | default | Lossy, efficient on scans and photographs. No alpha, which is why the page is painted white before rendering. |
| PNG | format | — | Lossless, so text stays crisp, and often smaller than JPEG on a page that is mostly white. Transparency is preserved. |
| WebP | format | — | Typically 25–35% smaller than JPEG at the same visual quality. |
| 72 DPI | resolution | — | Screen size: an A4 page is 595 × 842 pixels. For a thumbnail or a preview. |
| 150 DPI | resolution | default | 1240 × 1754 for A4. Readable when printed; the sensible default. |
| 300 DPI | resolution | — | 2480 × 3508 for A4, and about 34 MB of pixel data per page before compression. Print quality, and the practical ceiling in a browser. |
| Quality | 0.1 – 1.0 | 0.9 | JPEG and WebP only. Above 0.95 the file grows fast for no visible gain; below 0.7 artefacts show around text. |
| Page range | 1-3, 7 | all pages | Print-dialogue syntax. Leave empty to convert the whole document. |
| Memory | released per page | — | Each canvas is freed as soon as its image is encoded, which is what lets a long document convert at high DPI. |
Frequently asked questions
Why did my converted pages come out black?
That is what happens when a tool renders onto a transparent canvas and encodes it as JPEG, which has no alpha channel — transparent becomes black. A PDF page has no background of its own; the white you see is the reader painting one. This tool fills the canvas white before rendering for JPEG and WebP, and deliberately does not for PNG, where you probably chose PNG because you wanted the transparency.
What DPI should I use?
150 for almost everything — readable when printed, and a manageable file size. 72 for a thumbnail or web preview. 300 only if the result is going to a printer that will do it justice, and be aware that at 300 DPI a single A4 page is about 34 megabytes of pixel data, so a long document at that setting is genuinely close to what a browser tab can hold.
JPG or PNG?
JPG for scanned pages and anything photographic: its compression is built for continuous tone and the artefacts hide in the noise. PNG for pages that are mostly text, tables or line drawings — it is lossless so text edges stay sharp, and on a page that is largely white it is frequently smaller than the JPEG as well. WebP beats both on size if the destination accepts it.
Can I get the text back out of the images?
Not without OCR, and that is not a step you want to need. Rasterising throws away the text layer completely — the result is a picture of words, not words. If you need the text, use the PDF to Text tool on the original document, which reads the text layer directly and is both exact and instant.
Why is the image file bigger than the whole PDF?
Because a PDF stores instructions and an image stores pixels. A page of text is a few kilobytes of drawing commands referencing a font that is embedded once for the whole document; the same page as an image at 150 DPI is two million pixels that have to be compressed individually. A 200 KB PDF becoming 8 MB of images is entirely normal, and it is a good reason to keep the PDF.
Is my document uploaded?
No. pdf.js runs in this tab — it is the same engine Firefox uses to display PDFs, running as JavaScript on your machine. Rendering happens on a canvas in your browser and the images are created locally. The page's Content-Security-Policy sets connect-src 'self', so an upload would be blocked by the browser.