What reaching 1 MB changes in a PDF

Set 1 MB as the target and the engine works down to it in steps, stopping at the first result that fits rather than padding the file back out.

It is the pictures inside the PDF that shrink: they are downsampled a step at a time until the file fits. Text typed into the document stays text while that is enough to reach the target. Ask for a size downsampling cannot reach and the pages are rendered to images instead, which makes the size but ends the text layer.

Files are processed on our servers, not in your browser. The upload and the result are both permanently deleted within 1 hour, no backups and no copies kept.

No watermark, and no account needed to use it. 2 files a day at up to 50 MB each, sent singly or together; signing in is free and raises each file to 100 MB.

Getting a document ready to send

1. 1

   Decide whether you are clearing a cap or sending an email

   If a form on screen states a number, compress to that number instead. 1 MB is for the case where nothing is enforcing anything and you simply want the file to travel easily.

2. 2

   Aim at 1 MB and stop there

   One pass is enough. Choosing a tighter target out of caution spends image detail that a mail server was never going to ask you for, and nothing downstream rewards you for the smaller file.

3. 3

   Watch the whole message, not the one file

   Attachment limits apply to the message, and mail systems inflate files slightly while they are in transit. One compressed document is never the problem. Six of them in one reply can be.

Nobody imposes a 1 MB limit, which is why it is the right default

1 MB is the target you pick when the document is going to a person rather than to a validator: a signed agreement to a colleague, a scanned invoice to an accountant, a certificate attached to a reply. It clears the attachment allowance on every mail service in ordinary use, leaves room for two or three more files in the same message, and gives up very little of what the page looked like before.

What a twelvefold reduction actually looks like

A 13.2 MB three-page 300 DPI scan comes back at 724 KB against this target. An 8.9 MB file made mostly of full-page images comes back at 709 KB. Both are roughly twelve times smaller than they went in, and on screen the page detail is close to the original.

The reason the two land so near each other is that the budget is generous enough for neither to be straining. Below a few hundred kilobytes those same two files separate sharply, because the photographic one runs out of room first. At 1 MB neither has to make that choice yet.

The part that is never touched

Between 70 and 95 percent of what a PDF weighs is the images inside it. Those are what a size target works on. Text and vector drawing are never rasterised, so headings, body copy, tables and signature lines drawn as vectors come out exactly as sharp as they went in and remain selectable.

This matters more than it sounds for a document you are emailing. The recipient can still copy a clause out of it, search it for a name, or paste a figure into a spreadsheet. A compressed PDF is not a picture of your document, it is still the document.

Two situations where this is the wrong page

A PDF that is only text is already at its floor. A ten-page text- only document weighs 19 KB before anything is done to it, so running it through a 1 MB target achieves nothing at all. If the file properties already read a few hundred kilobytes, send the file you have.

A long scanned bundle is the opposite problem. Twenty or thirty scanned pages sharing one megabyte leaves each page very little, and the result shows it on anything carrying a photograph or a seal. For a set that size, the 2 MB target is the one to use.

Frequently asked questions