If text turns blurry after converting a PDF to JPG, the JPG label is not the only possible cause. A PDF viewer can redraw text and lines for the display, while an exported page image has a fixed number of pixels. The export dimensions matter first.
We rendered one document page at three sizes and saved each as three image types. The aim is to choose for the image's destination, not to declare that the largest file is always best.
The practical choices
Our A4 page measured 595×842px at 1× and 1190×1684px at 2×. Inspect small text at the image's real display size before accepting the smaller output.
At 1.5×, this text-and-vector page produced WebP 77.6KiB, JPG 140.8KiB, and PNG 172.5KiB. Those rankings are not guaranteed for photos or scans.
A page image records its appearance, not the original PDF's selectable text layer. Preserve the source PDF when text search or reuse matters.
One page, nine image outputs
- 1
We generated a synthetic A4 page (595×842 PDF points) with 8-, 10-, and 12-point text, 22 body lines, a vector color gradient, and fine lines. PDF text extraction returned 30 text items before export. This is a document-style test page, not a photographed scan.
- 2
We used the same PDF.js renderer as Toolit's PDF-to-images tool at 1×, 1.5×, and 2×. Each rendered page was encoded as JPG, PNG, and WebP. JPG and WebP used quality 90; PNG was lossless. Sizes use 1KiB = 1,024 bytes.
- 3
The reproducible script is scripts/measure-pdf-image-exports.mjs. Its Node canvas encoder (@napi-rs/canvas) can differ from a browser's image encoder. These byte counts describe this controlled run, not a guaranteed download size in every browser.
More resolution meant more pixels and more bytes
Doubling both width and height gives four times as many pixels. Compressed file size did not increase by exactly four times because page content and the encoder matter too.
| Render scale | Image pixels | JPG 90 | PNG | WebP 90 |
|---|---|---|---|---|
| 1× | 595×842 | 74.4KiB | 99.2KiB | 42.1KiB |
| 1.5× | 892×1263 | 140.8KiB | 172.5KiB | 77.6KiB |
| 2× | 1190×1684 | 205.9KiB | 245.1KiB | 110.9KiB |
Why can the letters look blurry?
A PDF viewer can draw source text again at a different zoom level. Once the page is saved as a 595px-wide image, small letters have only those pixels. Enlarging the image cannot recreate strokes that were not captured.
Increasing output resolution also cannot restore detail absent from a blurry scan. Lossy JPG may add artifacts around fine text; compare the result at its actual display size. We measured pixels and bytes here, not human reading accuracy.
Choosing JPG, PNG, or WebP
WebP was smallest for this synthetic page, but a photo-heavy PDF, a different color pattern, or another encoder may change the ranking. Compare formats at the same dimensions.
Start with JPG if the destination accepts only JPG. Try PNG when preserving sharp edges without lossy image compression matters most. Compare WebP if the destination supports it and file size matters. Resolution and format are separate decisions.
| Destination | Try first | Inspect |
|---|---|---|
| Chat preview or small screen | 1× or 1.5× | Small text on a phone |
| Large display or zooming | 1.5× or 2× | Letters and lines at final size |
| Searchable, reusable document | Keep the PDF | Text selection and search |
What a page image does not preserve
Saving a whole page as JPG, PNG, or WebP captures its visible appearance. It does not carry the PDF text layer or form fields in the same way. Putting the image into a new PDF does not automatically restore selectable words. OCR is a separate recognition step and should be checked for errors.
Keep the original PDF and make image copies only for the pages that need them. For formal submissions, check the receiving site's format, dimensions, and size requirements first.
Limits of this comparison
This was one synthetic text-and-vector page encoded in Node. It cannot represent real photographed scans, complex images, other browsers, or print quality. Neither 2× nor WebP is a universal answer.
Official documentation used for this analysis
Choose display size first, then image format
The first defense against blurry PDF page images is checking text at the size where the result will actually appear. Our 1.5× result was 892×1263px, not a universal recommendation.
Start smaller for a compact preview and test a higher scale for zoomable documents. Keep the source PDF separately whenever searching or copying its text matters.