Why is my PDF so big?

First, a quick way to diagnose your file

Divide the file size by the number of pages. A text document exported from a word processor typically weighs somewhere between 10 and 100 KB per page. If yours is closer to 1 MB or more per page, the pages almost certainly contain large images, even if they look like plain text. That is the classic sign of a scan.

Another quick test: open the PDF and try to select a word with your cursor. If you can highlight individual words, the page holds real text. If the whole page highlights as one block, or nothing happens, the page is a picture of text, and pictures are where the weight lives.

High-resolution scans and photos

This is the most common reason by far. Phone cameras now take photos of 12 megapixels or more, and scanner apps often save at 300 or even 600 dpi. A single full-colour A4 page at 300 dpi holds around 8.7 million pixels. Even after JPEG compression, that can easily be 1 to 3 MB for one page.

How to tell: the size per page is high and fairly even across pages, and zooming in reveals paper texture or camera noise.

How to fix it: lower the resolution of the images. For documents meant to be read on a screen, 150 dpi is plenty. If you are scanning again, choose a lower dpi and grayscale for black-and-white pages. If you only have the PDF, a compressor that downsamples images will do the same job after the fact.

Images stored uncompressed or as PNG

Some programs embed images using lossless methods. That is ideal for screenshots and diagrams, but for a photo or a scan it can make each image several times larger than an equivalent JPEG. Documents assembled from screenshots of photos, or exported by older scanning software, often have this problem.

How to tell: a PDF with only a few pictures is surprisingly heavy, and those pictures are photographic rather than flat graphics.

How to fix it: re-encode the photos as JPEG at a moderate quality. For pictures of real-world scenes, a JPEG at 70 to 80 percent quality usually looks the same to the eye at a fraction of the size.

Embedded fonts

A PDF carries its own fonts so it looks the same on every device. Usually only the characters actually used are stored, which adds just a few kilobytes. Some tools, however, embed entire font families, and fonts covering Chinese, Japanese or Korean characters can be several megabytes each.

How to tell: a short, text-only document is larger than about 500 KB. In Adobe Acrobat Reader you can see the fonts under File, Properties, Fonts, though the list does not show their size.

How to fix it: re-export from the original application with font subsetting turned on, which is often the default in a Save as PDF dialog. Image compressors do not change fonts, so this is one case where you need the source file.

Merged files and duplicated resources

When you merge several PDFs, each one brings its own fonts, logos and background images. Ten invoices from the same company may each contain an identical copy of the letterhead and the same font. The merged file stores them all, so it is roughly as large as all the originals added together.

How to tell: the merged PDF is about the sum of its parts, and the parts share a common design.

How to fix it: this is normal and usually not a problem. If the total is too big for a portal, compress the merged file, which mainly helps when the repeated elements are images.

Incremental saves, hidden data and attachments

Many PDF editors save changes by appending them to the end of the file instead of rewriting it. That is quick, but the old versions of edited pages and images remain inside. A form filled in and saved a dozen times can carry a dozen layers of history. Other hidden extras include embedded page thumbnails, private data that editors leave behind, file attachments and optional content layers from design software.

How to tell: the file grew noticeably after small edits, or your PDF reader shows a paperclip or a layers panel.

How to fix it: use your editor's Save As option, which often writes a clean copy, or run a lossless clean-up.

• The Light level of the PDFHomebase Compress PDF tool removes page thumbnails, editor leftovers and objects no longer used by any page, such as old revisions.
• Attachments and layers are left alone by the compressor, so remove them in a full PDF editor if they are the problem.
• Light never changes how pages look, which makes it a safe first try.

What our compressor does, and when it will not help

The Compress PDF tool works on images. At the Recommended level it decodes each photo or scan inside the PDF, scales it down to at most 2000 pixels on the longest side and re-saves it as a JPEG at around 72 percent quality using your browser's built-in image engine. Text, fonts and vector drawings are not touched, so they stay sharp and searchable. Images in formats it cannot safely decode, such as CMYK or JPEG 2000, are kept as they are.

If the new file is not smaller than the original, the tool gives you back your original and says so. That is why a text-only PDF may come out unchanged: it is already about as small as it can be, and there are no images to shrink. Everything runs locally, so your file never leaves your device during the process.

Tools used in this guide

• Compress PDFShrink PDF file size while keeping it readable. Never makes a file bigger.
• Merge PDFCombine several PDFs into one document, in the order you choose.
• Split PDFSplit one PDF into several files by page range or every N pages.
• Delete pagesRemove unwanted pages from a PDF in a few clicks.

Frequently asked questions