Visual testing: regression beyond pixel diffs
Visual testing compares what a page looks like against what it looked like before, and reports what moved. A pixel diff is the simplest form and the noisiest: an anti-aliasing change lights up a screenshot while a button that became unreachable does not. Useful visual testing combines a tolerant comparison with checks on layout, overflow and the elements a user has to reach.
The guide and the article below cover the baseline workflow and the limits of pixel comparison; the website analysis page reads performance, accessibility and security from one real page load; the UX testing page is for the layer above pixels, the journey.
What people search for
The literal queries this hub is written to answer; each leads to one of the pages below.
- visual regression testing
- visual testing tools
- screenshot comparison testing
- applitools alternatives
- ux testing tool
- website analysis tool
Learn
Definitions and the limits of a diff.
Visual regression testing
Baseline, threshold, highlighted pixels.
Beyond pixel diffs
What a diff misses and what to check instead.
Do it
The guides.
Visual regression guide
Baselines, thresholds and updating on purpose.
Website analysis
One URL, one real page load, three reports.
Core Web Vitals
The three numbers website analysis reads from the load: LCP, INP and CLS.
Cross-browser testing with Playwright
The engine differences that show up as visual diffs.
Prove it
The product pages.
Visual regression testing
The module, with its limits.
UX testing tool
The journey layer, as a lab proxy.
Questions people ask
What is visual regression testing?
Comparing a screenshot of a page against a stored baseline and reporting the difference: a percentage, a pass or fail against the threshold you set, and an image that highlights the pixels that changed. It catches a collapsed layout or a vanished button that a functional test walks straight past.
Why do visual tests produce so many false alarms?
Because pixels change for reasons that do not matter: font rendering, anti-aliasing, a timestamp, an animation caught mid-frame, a different engine. A tolerant threshold, masked regions, frozen time and a per-engine baseline remove most of them; the beyond-pixel-diffs article shows what to assert instead of pixels.
Is visual testing the same as UX testing?
No. Visual testing asks whether the page looks as it did; UX testing asks whether a person can complete the journey well. The UX testing page is explicit that it is a lab proxy run by your own testers, not a study with recruited users.
Try it on your own application
The Free plan asks for no card and runs for 14 days after the trial: record a flow, replay it, and read the evidence each step produced.