Converting PDF Pages to JPG or PNG

When an Image of the Page Is the Right Choice

A JPG or PNG of a page is useful where PDFs are not accepted or not shown directly: in a slide deck, a chat or forum post, an upload form that takes only images, or a document thumbnail on a web page.

It is the wrong tool for making a file smaller (see How to Compress a PDF) or for getting the words out (use PDF to Text). The image has no selectable text, links or bookmarks, so keep the original PDF.

How the Quality Levels Translate to Pixels

PDF pages are measured in points, 72 to the inch; an A4 page is about 595 × 842 points. PDF.js, the library the converter uses, draws a page at scale 1 with one pixel per point, so 72 pixels per inch of the page, and at scale 3 with 216. The three LeedyBox levels have base scales of 2, 3 and 4. The converter multiplies them by 1.5 when the browser reports a device pixel ratio of 1.5 or less (typical of monitors at Windows display scaling of 100 to 150%) and by 1.2 above that (most phones and tablets, Mac Retina screens and Windows scaling above 150%), so the same setting gives smaller images on a phone. Below, "desktop" is short for a pixel ratio up to 1.5 and "phone" for a higher one. Pixel sizes are rounded down to whole pixels:

A US Letter page (612 × 792 points) gives 2,754 × 3,564 pixels at High on a desktop. Landscape pages swap width and height, and the exact size can differ by a pixel or two between PDFs because programs store the A4 size with slightly different decimals. The pixel ratio is read again for every page you convert, and in some desktop browsers page zoom changes it, so keep the default zoom for predictable sizes.

Very high has four times the pixels of Standard. One A4 page at Very high on a desktop is about 18 million pixels; at 4 bytes per pixel, the drawing surface alone needs roughly 72 MB of memory before the image file is encoded. See also Resizing Images: Pixels, Aspect Ratio and DPI.

JPG or PNG for Page Images

JPG is the default and is saved at a fixed quality of 0.95 (on the browser's 0 to 1 scale); the Clarity / Quality Level menu changes the pixel size, not the JPG compression. Sharp black-on-white letter edges are expensive for JPG and can leave faint speckles around the letters, while lossless PNG compresses the flat white areas of a page well.

In our test (two A4 pages rendered at 324 dpi with the Poppler library rather than PDF.js, and saved with Python's Pillow library: JPG at quality 95, PNG at default settings), a page of body text came to about 2 MB as JPG and 600 KB as PNG, and a page with a bar chart and a table to about 700 KB and 270 KB. Browsers render and encode differently, and their PNG encoders may favour speed over file size, so your files will differ; even so, for text and chart pages PNG is often both cleaner and smaller.

Choose JPG for photos, scans and gradients, which make PNG files large. Both formats get a white page background, not transparency. The settings are read at each click, so with the Download button under each preview one PDF can give a chart page as PNG and a photo page as JPG. For the formats in general, see JPG vs PNG vs WebP vs AVIF.

Step by Step in LeedyBox

1. Open PDF to JPG and choose or drop one PDF in Choose PDF File or Drag & Drop Here. If you drop several, only the first is used, and other file types are refused.
2. Wait while a preview of every page is drawn; the progress bar shows "Analyzing PDF document..." until all previews are ready, so long documents take longer. The file name, size and page count then appear.
3. Set Image Format: JPG (Recommended) or PNG (Lossless format).
4. Set Clarity / Quality Level: Standard (smaller files), High (recommended), the default, or Very high (largest files).
5. Click Convert All Pages & Download ZIP, or Download under one preview. From report.pdf you get report_images.zip containing report_page_1.jpg, report_page_2.jpg and so on; a single page downloads as report_page_3.jpg, for example.

The previews are drawn at 36 pixels per inch and shown as compressed JPGs, so judge sharpness on a downloaded image. To start over with another PDF, click Change File.

Choosing Settings for Common Jobs

A worked example for slides: a 16:9 slide shown on a Full HD screen is 1,920 × 1,080 pixels. A whole A4 page at full slide height needs only 1,080 pixels of height, but a chart cropped out of the page and stretched across the slide needs about 1,920 pixels of its own width. A chart that fills the text width of an A4 page with 2.5 cm margins covers 16 of the page's 21 cm, so it comes out about 1,360 pixels wide at Standard on a desktop (too narrow) and about 2,040 at High (enough). On a phone, High gives about 1,630, so choose Very high (about 2,180).

For print, A4 at 300 pixels per inch needs about 2,480 × 3,508 pixels and US Letter about 2,550 × 3,300. High on a desktop covers both; on a phone, choose Very high (about 346 dpi). A page scanned at 300 dpi, by contrast, already has its final detail, and Very high only enlarges its pixels.

Common Problems and Fixes

• Small text looks soft. Check the image at actual size and try a higher level. If the letters already look pixelated when you zoom into the PDF, the page is probably a scan or an image of text, and no setting can add detail that is not there.
• The images are too large. Try PNG for text pages or choose Standard, then use the Image Resizer or Image Compressor to meet an upload limit.
• You need only some pages. Use Download under each preview, or extract a range with Split PDF first: pages 5–12 of an 80-page report then give 8 images instead of 80. Split PDF saves the range as report_extracted.pdf, whose pages start again at 1, so the images are named report_extracted_page_1.jpg to report_extracted_page_8.jpg.
• "… cannot be read by the browser or is corrupted". The file may be damaged or need a password to open, and the converter has no password field. If you are allowed to, save a copy without the password in an application where you can enter it.
• A page is sideways. The image follows the page's rotation setting, and a page that was scanned or created sideways stays sideways. Fix it with Rotate PDF (see How to Reorder, Rotate and Delete PDF Pages) or turn the image with Rotate & Flip Image.
• The conversion stops or the tab reloads. Large pages at high levels can exceed a browser's memory or canvas limits, especially on phones. The ZIP is also assembled in the tab's memory, so every page image is held until the download starts. Use a lower level, download pages one at a time, or split a long PDF into parts first.
• Wrong order after unzipping. The names have no leading zeros, so some programs list page_10 before page_2. Sort numerically or rename the files.
• The image looks different from your PDF viewer. Fonts that are not embedded in the PDF are drawn with substitute fonts, so letter shapes and spacing can look different, and print colours (CMYK) are converted to screen colours by PDF.js, which can differ slightly from your viewer's conversion.

What to Check Afterwards

• The number of images matches the page count shown when the file loaded.
• The smallest text and thin lines are readable at actual size.
• Orientation is right and nothing is cut off at the edges.
• Comments, stamps, signatures and form fields appear as in your PDF viewer.

Limits of the LeedyBox Converter

• It converts one PDF at a time and renders whole pages; photos embedded in a PDF are not saved separately at their original resolution.
• There is no page-range option and no separate JPG quality setting.
• The converter does not write the rendering resolution into the files, so an image editor may show a default such as 72 or 96 dpi. The pixel count is what matters for print size.
• The ZIP does not compress the images further, so it is about as large as all of them together.

Frequently Asked Questions