PDF workflow · 9 min read
Why scanned PDFs are so large
A PDF is a container, and its size depends entirely on what is inside it. A PDF generated from a word processor usually contains text and vector shapes, so a hundred pages might weigh only a few hundred kilobytes. A PDF created by scanning paper is a different animal: each page is a photograph, and a 300 DPI colour scan of an A4 page can be several megabytes on its own. Twenty pages at that quality will comfortably exceed 50 MB. This is why the same PDF can be tiny or enormous, and why the right compression strategy depends on which kind you have. Before you try anything, check whether the text in the PDF can be selected. If you can highlight a sentence, the file contains real text and is probably already efficient. If you cannot, the pages are images and the size is coming from those images.
The most effective lever: reduce the scan resolution
For image-based PDFs, resolution is the biggest factor in file size. Halving the resolution of a scan reduces the pixel count to a quarter, and the file size falls roughly in proportion. The question is how far you can go before the text stops being legible. For a document that will be read on screen and printed occasionally, 150 DPI is usually sufficient; 200 DPI is a safer choice for small print or handwriting; 300 DPI is really only needed for archival copies or documents that will be OCR'd and must be highly accurate. If you are re-scanning the original, change the setting at the scanner rather than compressing afterwards, because you get better quality at the same file size. If you only have the existing PDF, you can convert the pages to images at a lower resolution and rebuild the document from those images.
Colour versus grayscale versus black and white
Colour scans carry three channels of information per pixel; grayscale carries one. Converting a text document to grayscale can cut the file size substantially with almost no loss of usefulness, because most text documents contain no meaningful colour. Black and white, sometimes called bilevel, reduces each pixel to a single bit and produces very small files, but it can make photographs and light-coloured stamps look harsh and may thin out faint handwriting. Use grayscale for mixed documents and reserve black and white for clean, high-contrast text pages. If the document contains a photograph, a signature in blue ink or a coloured stamp that matters, keep it in colour and reduce the resolution instead, because colour is part of the information in that case.
Rebuild the PDF from compressed page images
If you do not have access to a dedicated PDF compressor, a reliable workaround is to convert the PDF pages to images, compress those images, and then combine them back into a PDF. The workflow is: render each page as a JPG or PNG at a moderate resolution, compress the images to a sensible quality, then use an image-to-PDF tool to rebuild the document in the correct page order. This gives you direct control over the two variables that matter, resolution and JPEG quality, and it is easy to check the result page by page. The downside is that the rebuilt PDF will not contain selectable text, links, form fields or bookmarks, because it is now a stack of pictures. For a document that only needs to be read or printed, that trade-off is usually acceptable; for a document that needs to be searched or filled in, it is not.
What not to do
Do not repeatedly compress an already-compressed PDF. Each round of lossy compression degrades the page images a little more, and after several passes the text develops visible ringing and blockiness. Compress once, from the best source you have, and check the result. Do not reduce the resolution so far that the text becomes fuzzy at normal reading size; a 400 KB file that nobody can read is worse than a 2 MB file that works. Do not assume that removing pages is a valid way to hit a size limit unless the portal genuinely does not need them. And do not upload a confidential document to an unknown compression service just because it promises a smaller file. If the document contains identity papers, financial records or medical information, use a tool that processes the file in your browser and check its privacy explanation first.
Check the result before you upload
Open the compressed PDF and look at the pages that matter most: the first page, any page with a photograph, any page with small print, and the last page. Zoom to 100 percent and confirm that the text is still crisp enough to read comfortably. Check that the page count matches the original and that no pages were dropped or reordered during the rebuild. Then check the file size against the portal's limit with a little margin, because some systems measure in a way that differs slightly from your file manager. Finally, keep the original uncompressed PDF in your own records. The compressed copy is for the upload; the original is what you will want if the document is ever needed for a legal, medical or official purpose where quality matters.
Do it now with a photoedit.in tool
Every article below is paired with the exact tool that finishes the job. Files are processed in your browser wherever the tool supports it, and no account is required.
- Compress an image-based PDF to a target size
- Convert pages to images and rebuild a smaller PDF
- Rebuild the PDF from compressed page images
Prefer a walkthrough first? Browse the step-by-step guides, compare every online tool, or send feedback about this article.