
Use Cloudinary's density parameter, written as dn in delivery URLs, to control the resolution of a PDF page delivered as an image. The documented default is 150 DPI. Inspect that choice separately from the final image's displayed size, and first identify whether the soft content is vector lettering or an already-low-detail embedded photograph.
Key Takeaways
- PDF-to-image delivery has a documented density control.
- A large displayed preview is not proof of detailed source content.
- Compare the same page and final viewing size when evaluating a candidate.
Locate the softness in the source page
Open the original PDF in a capable viewer and identify a page by its visible heading or illustration. Compare a text edge, a vector shape and a photograph separately. Their detail can have different limits within the same document.
The Cloudinary paged-media reference documents conversion of a selected PDF page to an image and says dn changes the default 150-DPI resolution. It does not promise to reconstruct missing details in a photograph embedded in that PDF.
The print-PDF image-resolution guide covers that latter problem: putting a small raster inside a PDF does not turn it into a higher-detail original.
Make a controlled rasterization comparison
Keep the original PDF and existing preview address. Record the selected page, density, requested output size and output format. Change only the density for a non-sensitive candidate before comparing another setting.
Use the same page identity and final display size for both candidates. Otherwise, a different crop or zoom can look like an improvement unrelated to rasterization. Where the page contains both type and a photograph, record their results independently.
This is a proposed comparison, not a measured density recommendation. The cited page supplies the control and default, not one universal density suitable for every brochure, device or output size.
Inspect actual output rather than the container name
Download each preview and read its actual pixel dimensions. Compare its legibility at the intended display size, then inspect a close view to locate stair-stepping or soft source detail. A PDF filename, browser zoom percentage or large HTML display box does not establish those properties.
Do not describe the comparison as proof of an undocumented processing order. The useful decision is whether the documented density and requested delivery dimensions meet the specific preview requirement. If a source photograph remains soft while vector lettering improves, investigate the photograph separately rather than repeatedly enlarging the entire preview.
Accept detail for the intended page role
For a hypothetical brochure cover, the acceptance record could require a readable product name, an intact diagram and an adequately detailed hero photograph. Those are editorial requirements, not observed results.
If the next task needs an original embedded picture rather than the composed page, follow the render-versus-extract distinction. If only an editor's screen is soft, the display-preview diagnosis addresses a different stage. Accept the actual delivered page image when its detail is sufficient, while keeping unsupported source detail unresolved.
Related tasks
- Why Is My AI Artwork Still Blurry Inside a Print PDF?
- How Do I Take the Correct Scanned Page from a PDF for Image Editing?
- Why Does a Sharp Photo Look Pixelated Only Inside InDesign?
Sources and scope
- Cloudinary: Paged and layered media: PDF image delivery defaults to the first page and supports pg page selection and dn raster density, whose default is 150 DPI. For PSD delivery, pg_1 represents all layers and pg_2 the bottom layer; names are case-sensitive. Original layered files retain their layers, and incoming transformations can discard source layer data.
Documentation-based guidance from the cited official pages captured on October 25, 2026. Scenarios and acceptance procedures are proposed, not executed tests. No account operation, measured demand, AI-citation frequency, processing benchmark or search-indexing outcome is claimed.
Prepared by Imagild Editorial, the company editorial team responsible for these guides. AI assisted research and drafting. This is documentation-based guidance with proposed examples, not a claim of hands-on testing or an independent product review. Report a correction.
Cover: AI-generated conceptual editorial illustration; not a real application screenshot, observed output or before-and-after test.