Direct answer: choose PNG for pages dominated by small text, diagrams, screenshots, or line art; choose JPG for photo-heavy pages when smaller files matter. Start at 150 DPI for on-screen use and 300 DPI when the image will be printed, zoomed, or inspected closely. Use 72 DPI only for thumbnails. Convert one representative page first, then check the smallest text at the final display size before processing the rest.
Quick decision map: format and resolution
| Destination | Starting setting | Why | Verify |
|---|---|---|---|
| Thumbnail or contact sheet | PNG or JPG, 72 DPI | Smallest output; not intended for close reading | Page identity is clear |
| Website, chat, or slide preview with text | PNG, 150 DPI | Lossless encoding avoids extra text-edge artifacts | Footnotes are legible at the displayed width |
| Photo-heavy page for screen | JPG, 150 DPI | Usually smaller for continuous-tone images | Fine labels and color gradients still look acceptable |
| Print, evidence review, or close zoom | PNG, 300 DPI | More pixels preserve small detail during inspection | Measure print size and inspect stamps, tables, and signatures |
| Specialized high-detail print | PNG, 600 DPI only if needed | Very large output; gains may be invisible if the source is already low resolution | The extra pixels improve a real requirement |
These are starting points, not universal quality guarantees. Page size, source images, font size, final display width, printer settings, and later resampling all affect the result. Higher DPI creates more pixels; it cannot reconstruct detail that is absent from a scan or fix an out-of-focus source.
First choose the right job: render pages or extract images
Render every page
Use PDF to Image when you need one flat visual file per page, including its text, annotations, backgrounds, and layout. The result is a raster snapshot of the page.
Recover pictures inside the PDF
Use Extract Images when you want embedded photos, logos, or scans without the surrounding page. Vector art and flattened page scans may not exist as separate extractable image objects.
Rendering a page does not preserve selectable text, links, form fields, layers, tags, or PDF search. If the next person needs to search or copy text, keep the original PDF and use Extract Text for a separate text output instead.
PNG or JPG: what actually changes?
PNG stores the rendered pixels with lossless compression. That makes it a dependable choice for text edges, UI captures, diagrams, thin lines, and pages that may be edited or resaved. JPG uses lossy compression: it can reduce file size effectively for photographs, but compression can introduce ringing or block artifacts around letters and lines.
The MDN image-format guide recommends lossless formats for screenshots containing text because lossy compression can make it fuzzy. The W3C PNG specification defines PNG as a lossless raster format. Neither choice keeps the original PDF’s vector or text semantics once the whole page is rendered.
Step by step: convert PDF pages without avoidable blur
- Keep the source PDF. Work from the cleanest available original and save images as derivatives. Do not replace an accessible, signed, searchable, or archival PDF with page images.
- Open PDFCheck PDF to Image. Upload a non-sensitive file only when the workflow allows an online service.
- Choose PNG or JPG by page content. Default to PNG for text and diagrams. Choose JPG when photographs dominate and smaller files are worth a controlled quality trade-off.
- Choose the lowest DPI that passes the real test. Start at 150 for screen delivery or 300 for print/inspection. A US Letter page is roughly 1,275×1,650 pixels at 150 DPI and 2,550×3,300 at 300 DPI; an A4 page is roughly 1,240×1,754 and 2,480×3,508 respectively.
- Convert and inspect representative pages. Check the smallest footnote, a fine table line, a logo edge, a stamp, and one photo at 100% plus the final display width.
- Download individual pages or the ZIP. Keep page numbering intact. Record the format and DPI if another person must reproduce the output.
- Test the destination. Open the image on mobile, insert it into the actual slide or CMS, or print one page. Avoid a second lossy resize or JPG save after conversion.
PDFCheck capability boundary: PDF to Image renders each page as PNG or JPG at 72, 150, 300, or 600 DPI and packages the results for download. It does not OCR the images, preserve selectable text, links, form fields, layers, accessibility tags, attachments, or digital signatures in the output. It does not upscale missing source detail or certify print quality. Keep and validate the original PDF whenever those properties matter.
Failure states and recovery actions
| What went wrong | Likely cause | Recovery |
|---|---|---|
| Small text looks soft | Too few pixels at the final display size, JPG artifacts, or later scaling | Retry as PNG at the next DPI; remove any downstream resize and test at the final width |
| Output is extremely large | High DPI multiplied across many full-color pages | Test 150 instead of 300/600; use JPG for photo-heavy pages only when artifacts are acceptable |
| A 600 DPI image is still blurry | The PDF contains a low-resolution scan or already-compressed image | Find the original scan/source; extra output pixels cannot recreate missing detail |
| Text cannot be searched or copied | Whole-page rendering creates pixels, not PDF text objects | Keep the PDF, use text extraction, or run an authorized OCR workflow on the derivative |
| An expected photo is not separate | You rendered pages instead of extracting embedded assets | Use the Image Extractor and compare its output with the page render |
| The source PDF will not convert | The file may be malformed, encrypted, incomplete, or unsupported | Run the PDF Validator, obtain a clean original, and avoid bypassing access restrictions |
Realistic examples
A policy page for a website
The page is mostly 10–12 point text and a small table. Start with PNG at 150 DPI, place it at the final web width, and inspect the smallest row label on desktop and mobile. If people need accessible or searchable content, publish HTML or the original tagged PDF as well; the page image is not an accessible replacement.
A photo-led catalog page
Start with JPG at 150 DPI for a lightweight preview. Inspect product codes and price text because those edges reveal artifacts sooner than photographs do. If you need the original product photos, do not crop the page render—extract the embedded images instead.
A stamped form for evidence review
Start with PNG at 300 DPI and compare the image against the preserved PDF. Page rendering does not retain the PDF’s signature evidence or prove authenticity. Screen the original with the PDF Validator and follow any required institutional evidence workflow.
Final quality checklist
- □ The format matches the page: PNG for text/diagrams, JPG only for an acceptable photo/file-size trade-off.
- □ The smallest meaningful text is legible at the actual display or print size.
- □ Page count and order match the PDF, and filenames sort correctly.
- □ File dimensions and total ZIP size fit the destination limit.
- □ No second lossy conversion, automatic CMS resize, or chat compression damaged the result.
- □ The original PDF remains available for search, accessibility, links, forms, metadata, signatures, and validation.
Sources used for this guide
- MDN image file type and format guide — format trade-offs and lossless recommendations for screenshots and diagrams containing text.
- W3C PNG Third Edition — normative definition of PNG as a lossless compressed raster format.
- Ghostscript output-device documentation — rasterization, resolution, antialiasing, PNG, and JPEG device behavior.
- Adobe Acrobat export documentation — separates exporting PDF pages from exporting images or selections.
The safe default is simple: preserve the PDF, render a representative page as PNG at 150 DPI, test it in the real destination, and increase resolution only when a visible requirement fails. Choose JPG deliberately for photo-heavy output, not as an automatic shortcut, and never treat a page image as a replacement for the PDF’s searchable, accessible, interactive, or signed structure.