Scanned and digital PDFs store different kinds of information
A digital PDF may store compact text, vector shapes, fonts, and a few images. A scanned PDF usually stores a full-page photograph for every page, so its size is driven largely by pixel dimensions, color depth, and image compression.
Two documents with the same page count can therefore differ by hundreds of megabytes. Page count alone is not a useful compression estimate.
- Digital text can remain sharp at any zoom level.
- Scanned pages need enough pixels to keep small type legible.
Resolution, image encoding, and fonts drive file size
Downsampling embedded images reduces the number of pixels. JPEG compression can reduce photographic scan data further, but stronger settings may add halos or block artifacts around text. Embedded font subsets and repeated images also affect bytes, while simply resaving an already efficient document may change very little.
- Use grayscale only when color carries no meaning.
- Avoid repeatedly recompressing an already lossy scan.
- Keep the original before experimenting.
Rasterizing can shrink a file while removing useful structure
Rendering each page as one image can simplify a complex PDF and sometimes reduce its size. The tradeoff is substantial: selectable text, search, links, form fields, accessibility structure, and vector detail can disappear.
For archival, legal, or accessibility-sensitive documents, preserving document structure is usually more important than reaching an arbitrary target size.
Choose a quality target based on how the PDF will be used
A screen-only handout can tolerate lower image resolution than a document that will be printed or enlarged. Compare small text, diagrams, photographs, and the actual byte size after each change instead of assuming the strongest setting is best.
- Check several representative pages, not only page one.
- Confirm that text remains readable at the intended viewing size.
- Use tools such as PDF to JPG only when a visual page image is the intended output.