DxO Ships ViewPoint 6, Calls It Its Most Precise Geometry Fix Yet

DxO has announced ViewPoint 6, the newest version of its perspective correction software, and the French company is calling it its "most precise version yet." The claim rests on one thing: new AI masking technology built into this release.

The announcement, made on October 20, lands ViewPoint 6 as the latest iteration in a tool photographers have long used to straighten what lenses and viewpoints bend. DxO built ViewPoint's reputation on geometry — converging verticals, warped horizons, the leaning towers that plague architectural and street work alike.

The headline change in version 6 is the masking system. DxO says AI masking lets photographers apply geometric corrections with more precision than previous versions allowed. In practice, that means corrections can target what needs fixing while leaving the rest of the frame untouched. The company's own promotional imagery shows the point clearly: a woman in a hat and light coat standing by a river, with a digital grid overlay mapping the Eiffel Tower behind her. Dense correction mesh, human subject, one frame. That's the workflow ViewPoint 6 is selling.

What does this change for photographers?

Geometry correction has always carried a trade-off. Pull a building's verticals straight and you stretch something else — sky, foreground, the people in between. If the AI masking works as described, photographers get a way to isolate those distortions before correcting them. Architecture shooters working with wide lenses stand to gain most. So do travel and street photographers who mix human subjects with built structures, exactly the scenario DxO chose for its announcement visuals.

ViewPoint occupies a specific niche in DxO's lineup. It is not a raw editor and not a catalog. It does one job: fixing the geometry that lenses and angles impose on a scene. The software competes with the perspective tools baked into Adobe Lightroom and Photoshop, which have grown steadily more capable. DxO's answer has been depth — dedicated distortion controls rather than sliders bolted onto a broader editor.

Where the claim needs testing

"Most precise version yet" is DxO's phrasing, not an independent verdict. The software world has heard similar language with every major release from every vendor. What matters is whether the AI masking holds up on difficult frames — complex skylines, crowded streets, subjects overlapping the geometry that needs correcting. The Eiffel Tower demo is a controlled showcase. Real assignments rarely cooperate that neatly.

Still, the direction is clear. DxO has spent years pushing AI deeper into its products, from denoising to lens corrections. ViewPoint 6 continues that trajectory, applying machine intelligence to a craft problem — precise, selective correction — that manual tools have never fully solved.

For photographers already inside the DxO ecosystem, the upgrade question will come down to how often geometry correction drives their finished work. Heavy users of the previous version will want the masking on day one. Occasional correctors can wait for field reports.

DxO has not announced a follow-up event, but the company's release cadence suggests the AI masking introduced here will spread across its software line. Watch for it in future versions of PhotoLab — and expect competitors to answer with smarter selective corrections of their own.