Topaz were the first company to bring machine learning to photography when they released their DeNoise, Sharpen and Upscale apps six years ago. For a while they were unquestionably the industry leaders in AI post-processing tools and I was a happy customer, using their products on a daily basis.
Then they decided to ditch the individual apps and release an all-in-one AI powerhouse instead - an app that could take care of all your AI post-processing needs in one central AI location. That app was PhotoAI and, let's be honest, it's been sub-standard, under-powered and over-priced from day one.
So it's fair to say that I wasn't expecting very much from the recently released version 4 and, with impressive consistency, TopazLabs certainly didn't let me down.
Once a serious player, TopazLabs now feel like a legacy brand coasting on old hype, and PhotoAI 4 does nothing to suggest a course correction. The app is officially marketed as a "powerful photo restoration and enhancement tool" that will enable you to "achieve studio-level focus, detail, and clarity anywhere you shoot." Colour me surprised that a studio would waste their time or money on an app that somehow manages to get worse and worse with each iteration.
In this new version 4 release, the big changes are a new dust and scratch removal tool, super-focus v2 and 'active autopilot'. The dust and scratch removal tool which they somewhat bizarrely describe as "the first tool of its kind" can allegedly be used on damaged prints and film scans and for "cleaning up archival or scanned work." Super Focus meanwhile has an 'upgraded' deblurring model which enables users to 'refocus blurry images'. And finally the new active autopilot will suggest which of Photo AI 4's impossibly useless 'enhancements' will best 'fix' your photo.
That's the marketing bollocks - here's what you're actually getting.
Keep reading with a 7-day free trial
Subscribe to RAW & Unfiltered - Photography in the Real World to keep reading this post and get 7 days of free access to the full post archives.