DxO PhotoLab 10 Can Remove Sensor Dust With AI: Why Product Photographers Should Care

Product photographers can spend an absurd amount of time removing the same tiny dust spot from image after image. If the mark is on the camera sensor rather than the product, it may appear in the same place across an entire shoot. DxO PhotoLab 10, released on 2 September 2026, adds an AI-powered Dust Removal tool designed to detect and remove those sensor spots automatically at the RAW stage.

For ecommerce, this is a more useful AI feature than it first sounds. It targets a repetitive photographic defect without asking a generative model to reinterpret the product itself. That distinction matters when the image contains exact labels, engravings, jewellery hardware or other details that should not be redrawn.

What PhotoLab 10 actually removes

DxO says the new tool automatically detects and removes sensor dust and small imperfections at the RAW level. The company positions it as a one-step way to eliminate one of photo editing's repetitive jobs while keeping the photographer in control of the result.

The important word is sensor. This is not the same job as cleaning fingerprints from a bracelet, removing lint from fabric or deleting a scratch from packaging. Sensor dust sits in the imaging system, so the same spot can repeat across many photographs, especially when shooting at narrower apertures where dust becomes easier to see.

That makes automatic detection particularly relevant to catalogue work. A photographer may shoot fifty rings, bottles or accessories against the same pale background and later discover a faint dark spot in the same part of every frame. Fixing it once as part of RAW processing is much more attractive than retouching fifty exported JPEGs individually.

Why this is a relatively low-risk use of AI

A lot of modern AI product editing is generative. Ask an image model to clean a product, and it may need to synthesize replacement pixels. That can work well, but on an exact SKU it creates a familiar risk: the cleanup can also change a real detail.

Sensor-dust correction is a narrower problem. The goal is to identify a small imaging defect and reconstruct the surrounding area, not to redesign the scene. For a product photographer, that is a sensible place to automate because the desired outcome is objective: the dust spot should disappear while the photograph otherwise remains the photograph.

It still deserves inspection. A spot that overlaps a thin chain, tiny engraving, stone edge or printed character should be checked at full size after correction. Automatic does not mean infallible.

Jewellery is a good stress test

Jewellery photography often uses small apertures to keep more of a ring, bracelet or pendant sharp. Those conditions can make sensor dust more visible, particularly on clean white or light-grey backgrounds. The category also contains exactly the kind of fine detail that makes careless retouching risky.

Imagine a dark sensor spot landing beside a fine silver chain. A manual healing brush can easily smear the chain edge if the selection is too large. An automated RAW-level cleanup could save time, but the finished frame should still be compared with the neighbouring links. The correction should remove the photographic defect, not simplify the jewellery.

The same rule applies around pavé stones, prongs, hallmarks and engraved text. Let automation handle obvious empty-background spots aggressively. Review corrections near SKU-critical detail more conservatively.

Do not confuse sensor dust with product cleanup

PhotoLab 10's new feature does not remove the need to prepare the physical product properly. Fingerprints on polished metal, loose fibres, packaging smudges and real scratches are different problems because they exist on the photographed subject.

Clean those before the shoot whenever possible. If a temporary product defect still needs retouching, make it a deliberate local edit and keep the untouched photograph available for comparison. A scratch that belongs to the actual item should not automatically disappear just because software can remove it.

A useful distinction is simple: sensor dust is a camera defect. Dust sitting on the product is a preparation problem. A permanent mark on the product may be product information.

PhotoLab 10 also adds more precise local editing

Dust Removal is only one part of the September release. PhotoLab 10 also introduces Depth Masks, expanded AI Masks and new U Point controls. Depth Masks let photographers select areas based on their distance from the camera, while AI selections can be refined using colour and luminosity information.

Those tools are less directly ecommerce-specific, but they reinforce the same useful editing direction: make a precise adjustment to the part of the photograph that needs it instead of globally regenerating the image.

For a lifestyle product photograph, depth-aware masking could help separate a foreground product from a distant environment for tonal adjustments. It is not a product-segmentation guarantee, and reflective edges still need checking, but targeted photographic editing is generally easier to verify than a full generative rebuild.

A sensible workflow for a catalogue shoot

Shoot RAW and keep the original files. Before detailed retouching, inspect a few frames at high zoom against plain areas where sensor dust is easiest to see. Run the automatic dust-removal pass, then check representative images from the start, middle and end of the shoot.

Pay particular attention to any correction that overlaps the product. If the same sensor spot appears across the set, verify the correction on several different SKUs because a safe repair over an empty background may behave differently when that coordinate crosses a chain, label or product edge in another frame.

After that, continue with normal exposure, white-balance and local adjustments. Only use generative retouching where the photograph genuinely needs reconstruction. This keeps the most factual version of the product intact for as long as possible.

Who should care about this release

PhotoLab 10 is most relevant here to sellers or photographers who already shoot RAW and process batches from a camera. It is not a reason for a phone-only seller to rebuild their workflow around desktop RAW software, and it does not replace product-focused generation tools when you need new scenes, models or colour variants.

But for brands producing their own catalogue photography, automated sensor-dust cleanup can remove a surprisingly annoying production step. DxO offers PhotoLab 10 for macOS and Windows with a 30-day free trial, so existing RAW workflows can be tested without committing first.

The takeaway

DxO PhotoLab 10 is a worthwhile launch for product photographers not because it generates a new product scene, but because it uses AI on a narrow problem that should be automated: repetitive sensor-dust cleanup.

For ecommerce, that is a useful model for AI editing in general. Automate defects that are clearly photographic, keep the physical SKU as the source of truth, and inspect any correction that touches commercially important detail. Saving time is valuable when the automation removes the dust without rewriting the product.

Sources: DxO, “Introducing DxO PhotoLab 10: a new dimension of creative control,” published 2 September 2026; DxO, “What's new in DxO PhotoLab 10?”, reviewed 4 September 2026.