PDFWhy Is My PDF So Big? A Look Inside the File
A 40-page contract typed in Word can be 180 KB. Photograph those same 40 pages with a phone and you get 90 MB — five hundred times bigger, for the exact same words. Understanding why is the difference between blindly running a compressor and knowing what you can safely throw away.
A PDF is a container, not a picture
It helps to stop thinking of a PDF as a scan of a page and start thinking of it as a zip file with a page renderer attached. Inside are objects: text runs with positioning instructions, vector paths, embedded font programs, images, form fields, annotations, bookmarks, metadata and sometimes attachments or scripts. The viewer assembles those objects into what looks like paper.
Text is astonishingly cheap. A page of prose is a few kilobytes of characters plus coordinates. Everything expensive in a PDF is expensive because it is not text.
The four usual suspects
1. Scanned pages (by far the biggest culprit)
When you scan or photograph a page, the result is not text — it is a grid of coloured dots. An A4 page captured in full colour at 600 DPI is roughly 4,900 × 7,000 pixels. At three bytes per pixel that is about 100 MB of raw data for one page. JPEG compression pulls it down to a few megabytes, but the order of magnitude is set the moment you choose the scan settings.
Two settings dominate the result:
- Resolution. Halving DPI quarters the data. 600 → 300 DPI is a 75% reduction before any compression runs.
- Colour mode. Full colour uses three channels; greyscale uses one; black-and-white uses one bit. A text document scanned in colour is paying triple for paper that is beige.
This is why a scanned document is heavy even when it says almost nothing, and why the same document retyped is tiny.
2. Images dropped in at original size
Word, Google Docs, PowerPoint and design tools generally embed the image you gave them, not the image you see. Drag in a 12-megapixel phone photo, scale it down to a 5 cm thumbnail in the layout, and the full 12-megapixel original often still travels inside the file. A page will happily display a 4 MB photo in a space where 80 KB would look identical.
3. Embedded fonts
To guarantee the document looks the same everywhere, PDFs embed the fonts they use. A full font family with several weights, italics and an extended character set can add hundreds of kilobytes to a few megabytes. Well-produced files embed a subset — only the glyphs actually used on the pages. Files exported carelessly, or built by combining many documents from different sources, often carry several complete font programs, sometimes the same one more than once.
4. Leftovers you never see
PDFs support incremental saving: each save can append changes rather than rewriting the file. Edit and re-save a document a dozen times and it can carry a dozen layers of history. Add page thumbnails, unused form fields, orphaned objects from deleted pages, duplicated logos on every page stored as separate copies, and metadata from three different applications, and a modest document quietly puts on weight with nothing visible to show for it.
Quick diagnosis: can you select the text with your cursor? If yes, the file is text-based and the bulk is almost certainly images or fonts. If the text is not selectable, it is a scan — and the scan settings are your file size.
What compression actually does
"Compress PDF" is not one operation. A good compressor runs several passes, and they differ in how destructive they are:
- Downsampling images — reducing images to a resolution appropriate for the target use. Lossy, and the single largest saving in most files.
- Re-encoding images — recompressing pictures at a lower quality level, or converting them to a more efficient encoding. Lossy.
- Font subsetting — dropping glyphs the document never uses. Visually lossless.
- Object stream compression and deduplication — packing structural objects tightly and storing repeated resources (that logo on every page) once. Completely lossless.
- Discarding cruft — thumbnails, unused objects, old revisions. Lossless in appearance, though it removes edit history.
The practical consequence: a text-heavy PDF with a lot of leftovers can shrink by half with no visible change, while a scan-heavy PDF only gets meaningfully smaller when you accept lower image resolution.
How small should you actually go?
Pick the target from the destination, not from a slider:
| Destination | Sensible image resolution | Notes |
|---|---|---|
| Professional printing | 300 DPI | Going below this is visible in print |
| Office / home printer | 200–300 DPI | Rarely worth more than 300 |
| Screen reading, archiving | 150 DPI | Sharp on any normal display |
| Email, upload forms | 110–150 DPI | Readable; typically cuts a scan by 80–90% |
One rule saves a lot of regret: compression is one-way. Detail discarded to hit a size target does not come back, and re-compressing an already-compressed file mostly adds artefacts rather than kilobytes saved. Keep the original.
When compression is not the answer
Sometimes the file is simply large because it contains a lot. Then reach for a different tool:
- You only need part of it. Extract the pages you actually need, or split the document into sections. Sending chapter 4 beats sending a compressed copy of all twelve.
- It is a bundle of separate things. Split first, then compress the heavy part only.
- It is a scan you need to search or edit. Run OCR to add a text layer, or convert it to Word. A text document is orders of magnitude smaller than the images it came from.
- It is a stack of photos. Resize the images before building the PDF, not after.
A five-step routine that works on almost any file
- Try to select the text. Selectable → text-based; not selectable → a scan.
- Decide the destination and therefore the resolution you actually need.
- Remove what you do not need to send: extract or split first, compress second.
- Run compression, starting with a moderate setting rather than the most aggressive one.
- Open the result and check the pages you care about — small type, stamps and signatures degrade first.
Ready to try it? Compress a PDF now — free, no account, and on PDFMax the heavy lifting can run entirely inside your browser, so the file never has to leave your device.