Why Scanned PDFs Are So Much Bigger Than Typed Ones

A PDF that started as a Word document and a PDF that started as a scanned paper form can look almost identical on screen, and yet differ in file size by a hundred times or more. The reason comes down to what is actually stored inside each one.

A typed PDF stores instructions, not pixels

When a PDF is generated from a Word document or a webpage, what gets stored is closer to a set of drawing instructions: "draw the letter A at this position, in this font, at this size." Text stored this way is tiny. A whole page of prose is often just a few kilobytes, and it stays sharp at any zoom level because it is not made of pixels at all.

Fonts are usually embedded too, but typically as subsets containing only the characters the document uses, so even that overhead is small. A 10-page typed report can be a couple of hundred kilobytes, and most of that is fonts and any logos.

A scanned PDF stores a photograph of every page

A scanner does not know what a page says. It just takes a picture. So a scanned PDF is, page by page, a stored image, usually captured at a resolution designed for print (300 DPI or higher) even though it will mostly be read on a screen.

Some quick arithmetic shows why this gets big. A US Letter page is 8.5 by 11 inches. At 300 DPI that is 2550 by 3300 pixels, roughly 8.4 million pixels per page. A full-color image that size can easily be a few megabytes even after JPEG compression, and some scanners store pages in less efficient formats. Ten pages at 4MB each is a 40MB file, which is the kind of gap the title describes.

Color depth matters as well. A page scanned in full color carries far more data than the same page scanned in black and white or grayscale, even though most paper forms are really just black text on white.

What to do about it

You have a few realistic options, and the right one depends on what you need the file to do.

1. You just need it smaller. Reduce the image data. Our PDF compressor works by rendering each page as an image and re-saving it as a JPEG at a quality you choose with a slider. That suits scans well, because scans are already images. It does mean the output is image-only, so text in the result is not selectable, even if the original had selectable text.
2. You need the text, not a smaller file. Use OCR (optical character recognition), which reads the page image and recovers the characters. Our PDF OCR tool does this in your browser, and it recognizes English only. Accuracy depends heavily on scan quality, so crooked, faint, or low-resolution scans produce more mistakes.
3. You can still redo the scan. The best fix is at the source: scan at 150 DPI, in grayscale or black and white if color is not needed. Often that alone cuts the size dramatically with no visible loss for on-screen reading.

Common mistakes

• Compressing a typed PDF and expecting a big drop. If a file is mostly text and vector graphics, there is very little image data to squeeze. A tool that works by turning pages into pictures can even make a small typed PDF larger, and it will remove selectable text. If a typed PDF is large, look for oversized embedded photos instead.
• Scanning at the highest setting "just in case". Very high DPI is rarely needed unless you plan to print at large sizes or archive for detailed reproduction.
• Compressing too hard. Dropping quality to the minimum makes small text soft and hard to read. Check a page with small print after compressing, and raise the quality if it looks muddy.
• Overwriting the original. Always keep the original scan. Compression is lossy, and you cannot get detail back later.
• Merging before compressing without checking. If you combine several scans, the total size is roughly the sum of the parts. You can merge the files first and then compress once, or compress each part first, but check the final size either way.

How to decide

Ask what the file is for. If it is going into an email or an upload form with a size limit, you want smaller, so compress. If it needs to be searched, quoted, or copied into another document, you want text, so run OCR. If both matter, note that compressing by turning pages into images removes any text layer, so you would run OCR on the compressed result rather than the other way around.

Quick answers

Why is my scanned PDF so large even though it is only a few pages?

Because each page is a full-page image, often captured at 300 DPI or more. A few such pages can reach tens of megabytes. Typed documents store text as compact instructions, which is why they stay small.

Does compressing a scanned PDF reduce readability?

It can if you push it too far. Moderate settings generally keep ordinary text readable on screen, but there is no universal safe number. Compare a small-print area against the original and adjust. Our compressor's quality slider lets you try a few values and see the new size before you download.

Will OCR make my PDF smaller?

No. OCR recovers the text so that it is usable, but it does not by itself reduce the image data. Treat OCR and compression as two separate jobs.

Is it safe to use an online PDF tool for private documents?

That depends on the tool. Both of ours process files locally in your browser, so the document does not need to be uploaded to a server for these operations. For very sensitive documents, check any tool's behavior before you use it.