Which innovations will genuinely change underwater photography in 2027? Computational sensors, embedded AI, drones, augmented reality: an honest analysis.
In 2027, underwater photography will change on three simultaneous fronts: computational sensors will reach accessible housings, consumer-grade underwater drones will become viable, and augmented reality will start assisting real-time species spotting.
I read the announcement of the second-generation Sony IMX989 sensor in March 2026. I was in a café in Brussels between two pool training sessions. The technical paper described a one-inch sensor with embedded multiframe processing, optimised for low-contrast environments. My first reaction: this is what I have been waiting for in underwater photography. My second reaction: how long before it is available in a reasonably priced housing?
That question drives this article. Not press releases, not hype. Just what is genuinely going to change for you when you dive with a camera.
The fundamental problem in underwater photography is not resolution. A well-exposed 12-megapixel sensor produces a usable photo. The problem is contrast (reduced by water diffusion), dynamic range (compressed at depth), and noise (amplified by the high ISO values needed in low light).
Current sensors address these problems after the fact, in software. The result is often acceptable, sometimes excellent, but always a compromise.
Computational sensors integrate the processing directly in the silicon, not in a separate processor. Concretely: multiframe fusion happens in a few milliseconds, denoising is applied before the data even leaves the sensor, and optimisation profiles can be specific to the environment.
For underwater photography, this opens a concrete possibility: an embedded water mode that adjusts processing in real time based on depth, turbidity and available light.
The comparison between one-inch sensors and 8K formats remains relevant for current equipment choices, but computational sensors will reshuffle these criteria significantly.
In 2026, the only real embedded AI in a consumer underwater product remains Samsung Ocean Mode (on S24 Ultra and S25 Ultra). Traditional housings (Nauticam, Aquatica, Ikelite) offer no software intelligence: they are waterproof tubes that pass the photographer's commands to the camera.
A few dedicated underwater camera manufacturers (Paralenz, notably) integrate intelligent automation, but without generative AI or advanced multiframe processing.
Two concrete use cases seem close.
Predictive autofocus. Current AF systems struggle with fast subjects underwater (turbidity, refraction). An AI model trained on thousands of underwater sequences could predict a fish's trajectory and anticipate focus.
Real-time colour correction. Not just a profile like Ocean Mode, but dynamic correction that compensates for chromatic absorption based on the depth measured by the housing. Nauticam filed patents in this direction in 2025.
My honest assessment: embedded AI in housings will arrive first in the professional segment (€10,000 and above), then trickle down to the consumer market over 3 to 5 years. In 2027, expect announcements in the premium segment, not accessible products.
Underwater drones like the FIFISH V-EVO have existed for several years. Video quality is good. The current problem remains manoeuvrability in current, battery life (45 to 90 minutes in practice), and above all a control interface that is not intuitive for a photographer.
!FIFISH underwater drone during a dive, autonomous descent to 30m depth, coral background
What changes in 2027: the latest ARM processors enable semi-autonomous piloting. The drone maintains its position facing the subject, compensates for current automatically, and can track an animal on a predefined vector. The photographer designates the subject, the drone handles stabilisation.
What this does not replace: your physical presence in the water, your relationship with marine wildlife, your real-time reading of the light. A drone does not know that the morning light on this particular reef is better at 7am than at 10am.
Scubapro Luna 2.0 and a few competitors already offer connected masks with information display (depth, time, heading). Integration with photography remains marginal.
What is in development: real-time species identification via the mask camera, with display of the animal's name and typical behaviour. iNaturalist and Wildbook both have APIs that could feed this type of interface.
For the photographer, this is potentially useful for spotting rare species before they disappear into the blue. Not for composing the image, but for not missing the opportunity.
This is the least glamorous but perhaps the most concrete trend for the average photographer.
Current housings are expensive because they are manufactured in series for specific camera models. When Sony releases a new body, the housing manufacturer starts from scratch for the port and button design. Lead time: 6 to 18 months. Price: €700 to €3,000 for a quality housing.
SLS nylon or high-precision resin 3D printing makes custom parts possible on demand. Several manufacturers (AOI, Ikelite) already use this technology for spare parts and accessories.
In 2027, we can expect hybrid housings to appear: a standard waterproof body (traditionally injection-moulded for mechanical strength) with button panels and interfaces 3D-printed and adapted to each camera body. Potential cost: 40 to 60% less than a fully integrated housing.
I will be direct about what I personally expect from these developments.
Predictive underwater autofocus interests me a great deal. It is the technical limit I encounter most often: fast fish (amberjacks, barracudas, fleeing groupers) are difficult to pin sharp at wide aperture with current systems.
Generative AI in post-processing interests me for denoising, not for image manipulation. The question of authenticity in underwater photography is central to AquaExposure, and I have developed my position on marine deepfakes in a separate article.
Modular housings interest me a great deal for accessibility. Currently, the entry cost into underwater mirrorless photography remains prohibitive for many divers. Anything that reduces this cost without sacrificing quality moves in the right direction.
What does not interest me: features that automate the photographer's eye. Choosing the angle, waiting for the right moment, deciding whether to shoot or not: that is where the art lives. AI can help me execute technically, but it cannot decide on my behalf what is worth photographing.
AquaExposure receives no affiliate commissions on any products mentioned in this article.
Mastering the fundamentals remains the foundation, whatever the equipment generation. The AquaExposure underwater photography course prepares you to get the best from your current gear and whatever comes next.
A few prototypes exist (notably Paralenz), but "smart" housings in the full sense with embedded AI and real-time connectivity do not yet exist in a consumer version. The first viable models are expected during 2027.
No, and that is not their purpose. Underwater drones like the FIFISH V-EVO open angles impossible for a diver (descending to 100m, holding position in current), but they do not have the creative judgement or the ability to interact with marine life.
Not yet in any practical way. A few experiments exist with connected masks (Scubapro Luna 2.0), but integration with photography remains marginal. Real-time species identification via AR is the most promising use case for 2027.
No. Good current equipment you have mastered is worth more than future gear you are waiting for. 2027 innovations will first appear in premium versions at high prices. The fundamentals of underwater image-making remain unchanged regardless of equipment generation.
AI can automate colour correction, cropping and denoising. It cannot decide to get up at 5am for morning light, stay still for 20 minutes while an octopus gets used to you, or choose the angle that tells a story. AI amplifies the artist, it does not replace them.
Significant but gradual. Manufacturers like Ikelite and AOI already use 3D printing for certain components. Over time, modular and custom-on-demand housings could reduce prices by 30 to 50% in the mid-range segment.
Sensors continue to improve, but the real leap will come from specific underwater algorithms. Samsung Ocean Mode opened the way. When Apple and Google offer equivalent underwater profiles, smartphone underwater photography will cross a real threshold.
