Why a scanned PDF becomes unexpectedly large

A scanned PDF can be much larger than a longer text document because a scan is usually a stack of page images. The PDF container adds relatively little; the pixels inside each page carry most of the weight. That is why a four-page contract photographed at maximum camera resolution can be larger than a hundred-page report exported directly from a word processor.

The practical fix is not to drag a quality slider to its lowest setting. First work out whether the excess size came from the way the pages were captured, the way they were assembled, or images that can be reduced safely.

Diagnose the source of the size

Start by asking whether the document contains selectable text or only page images. If selecting a sentence does nothing, the PDF is probably image-only. Each page may contain millions of pixels even when the visible content is only black text on white paper.

Several capture choices make those page images larger than necessary:

• Maximum camera resolution. A phone can capture far more detail than an A4 page needs for ordinary screen reading.
• Full colour for monochrome material. Colour channels preserve information even when the page is mostly black ink and white paper.
• Loose framing. A large desk border, fingers, or objects around the page become part of the image.
• Shadows and uneven paper tone. Visual noise is harder to encode efficiently than a clean, consistent background.
• One photograph per page without correction. Perspective and background detail remain in every image instead of being cropped away.

File size is therefore not just a page-count problem. Two scans with the same number of pages can differ substantially because their pixel dimensions, colour content, and background complexity are different.

Decide whether to rescan or compress

Rescan when the source pages are blurry, heavily shadowed, badly framed, or captured at an obviously excessive resolution. Compression cannot restore clarity that was never captured, and aggressive settings can make already weak text harder to read.

The document scanning workspace lets you review page corners, correct perspective, crop away the surroundings, adjust the page treatment, choose an output quality, and validate the resulting image-only PDF. Automatic corners are suggestions rather than guarantees, so inspect every page and correct weak boundaries manually.

Compress the existing PDF when the pages are already clean and readable but their stored images are larger than the page actually needs. The PDF compression workspace analyses the document before changing it, identifies safe oversized image candidates, and preserves native text, vectors, page geometry, and unsupported images instead of using hidden whole-page rasterisation.

Some images are intentionally skipped, including images with transparency, unusual colour handling, unsafe dimensions, or unsupported encoding. A file that does not shrink is not necessarily broken; it may already be efficient, or its largest contributors may not be safe to recompress.

Use a quality target that matches the document

The right output depends on what the recipient must do with it. A reference copy read on a screen can tolerate more image reduction than a document containing small print, stamps, signatures, diagrams, or evidence that may be printed.

After reducing the file, inspect the smallest and most consequential details:

1. Zoom into fine text, dates, reference numbers, and handwritten marks.
2. Check pages with faint printing, coloured highlights, seals, or photographs.
3. Confirm that page count, order, dimensions, and orientation remain correct.
4. Compare the new size with the original rather than assuming the operation helped.
5. Keep the original file because discarded image detail cannot be recovered from the smaller copy.

If compression produces little improvement, reconsider the scope. The recipient may need only a page range rather than the entire scan. For recurring work, fix the capture settings so the next document starts smaller instead of repeatedly compressing oversized source images.

A repeatable scanned-PDF size workflow

Use this sequence whenever a scan is unexpectedly large:

1. Preserve the original.
2. Confirm whether the pages are images and identify the pages that look unusually detailed or noisy.
3. Rescan poor pages with even light, tighter framing, and an appropriate output quality.
4. Analyse clean existing pages before applying compression.
5. Review the output at normal reading size and at close zoom.
6. Send the smaller version only after checking that the important content remains legible.

This separates two different jobs: creating a clean scan and reducing an already usable one. Treating them separately avoids spending bytes on bad capture data and avoids sacrificing readability when the real problem should have been fixed at the camera.