How to Screenshot a Page That Keeps Loading More

A feed, a search results page, a product listing. You start a full-page capture and it never finishes, or it finishes and the image has a visible join in it where the content jumps.

The page has no bottom. Every time something reaches the end, the page quietly adds another screenful, and the end moves.

What is actually happening

A full-page capture begins by asking the page how tall it is. It uses that number to work out how many screens it needs to photograph and where each one starts. On an ordinary page that number is settled before anything begins, and the plan holds.

On an infinite-scroll page the number is a guess with an expiry date. The act of scrolling toward the bottom is precisely what triggers more content to load, so by the time the capture arrives at the position it calculated as the end, the end is somewhere else and there is fresh material in between.

Two things can result, and which one you get depends on the tool.

• A seam. The capture sticks to its original plan, photographs the last position it intended to, and stitches. Because the page grew underneath it, the content at that join does not line up with what came before — a row is repeated, or a chunk is missing.
• A capture that will not end. The capture re-measures as it goes, finds the page is taller every time, and keeps going. On a busy feed that can run until the tab runs out of memory.

How to recognize it

The join is usually obvious once you look for it: a post cut in half, a repeated thumbnail, a sudden change of alignment partway down. It sits at a single horizontal line rather than affecting the whole image.

If the problem is instead that the capture ran and ran, you already know. The thing worth checking is whether the page is genuinely endless or merely long — a page that loads more on a button press rather than on scroll is not infinite, and captures normally up to whatever you have already loaded.

What you can do about it

1. Decide where the bottom is, and make it real. Scroll until you have the material you actually want, then stop and wait a few seconds for the last batch to settle. Capture from there. You are choosing an end rather than hoping the page has one.
2. Look for the non-infinite version of the page. Many sites have paginated URLs still working underneath the feed, or a print view, or an older interface. A page that ends is worth hunting for; it makes every other step unnecessary.
3. Capture in sections. Two or three on-screen captures as you move down are more reliable than one attempt at the whole thing, and on a feed they are often more readable as well.
4. Do not expect print to PDF to help. Print rendering asks the page for its height the same way a capture does, and gets the same moving answer. On most feeds it produces the first screen or two and stops.

The thing most people get wrong

The instinct is to scroll all the way to the bottom first so that everything is loaded. On a genuinely infinite feed there is no bottom to reach, and the attempt loads thousands of items into a tab that then struggles to render, let alone photograph itself. A page heavy enough will fail the capture for reasons that have nothing to do with scrolling.

Scrolling first is the right instinct on an ordinary long page with late-loading images — that case is covered in why images are blank or grey in your screenshot. It is the wrong instinct here.

What PageStill does

It measures the page before it starts and again when it finishes. If the height changed while it was working, it tells you the page grew during the capture and that the image may have a seam, rather than handing you a picture and letting you find the join yourself.

That is a warning rather than a fix. No capture can photograph a page that is still being written; what it can do is be honest about it, so you know to scroll to a stopping point and go again.