Recording underwater requires solving two problems that don’t exist on land: keeping your equipment dry and compensating for how water transforms both light and sound. A waterproof housing protects the camera, but the port you shoot through changes image quality in measurable ways. Color shifts dramatically within a few meters of depth, demanding either in-camera corrections or careful post-processing. And sound behaves so differently in water that a standard microphone is useless; you need a hydrophone, a fundamentally different sensor, to capture anything meaningful. The good news is that each of these challenges has practical, well-understood solutions.
Dome Ports Versus Flat Ports
The window between your camera lens and the water is one of the most consequential decisions in underwater imaging, and it gets surprisingly little attention from casual shooters. Underwater housings use one of two port types: a flat port (a simple pane of optical glass or acrylic) or a dome port (a curved hemispherical or near-hemispherical window). Both keep water out, but they behave very differently optically.
A flat port introduces refraction at the water-glass and glass-air interfaces. Light bends when it passes through a flat surface at an angle, and the wider the field of view, the worse the distortion at the edges. Flat ports also magnify the scene by roughly 25 to 33 percent, which narrows your effective field of view and shifts your apparent focus distance. For macro work with a narrow-angle lens, that magnification can actually be useful. For wide-angle shooting, it’s a liability.
A dome port, by contrast, is designed so that light enters the curved surface close to perpendicular across the entire field. This eliminates most of the refraction-related distortion and preserves the lens’s native field of view. Testing in semi-submerged industrial structures has shown that a dome port produces lower image residuals and higher precision and accuracy in three-dimensional object measurements compared to a flat port, with a measurable difference in quality across individual color channels as well.1Copernicus Publications. FLAT VERSUS HEMISPHERICAL DOME PORTS IN UNDERWATER PHOTOGRAMMETRY If you’re doing any kind of photogrammetry or scientific documentation underwater, a dome port isn’t just a nice-to-have; it measurably changes the quality of your results.
The tradeoff is cost and complexity. Dome ports are larger, more fragile, and more expensive. They also create a virtual image very close to the front of the dome, which means your lens needs to be able to focus at short distances, often requiring a diopter or close-up lens behind the dome. Flat ports are compact, cheap, and work well enough for narrow-angle and macro shooting where edge distortion is minimal. For someone building a simple monitoring rig to drop on a reef, a flat port may be the practical choice. For a filmmaker or scientist who needs accurate color and geometry across a wide frame, a dome port pays for itself.
Housings on a Budget
Professional underwater camera housings from manufacturers like Nauticam or Ikelite can cost as much as the camera itself, sometimes more. That price barrier has pushed researchers and independent filmmakers toward do-it-yourself solutions, and some of them perform remarkably well. A published design for a low-cost housing built from commonly available materials demonstrated the ability to record full high-definition video for up to 13 hours at depths reaching 100 meters, using tools available in most workshops and requiring limited technical background to assemble.2FACETS. How to build a low-cost underwater camera housing for aquatic research
The key to any DIY housing is the seal. O-rings are the standard approach, and they need to be free of debris and properly greased every time you close the housing. A single hair or grain of sand across an O-ring can cause a flood. Many DIY designs use PVC pipe or acrylic tube as the pressure vessel, with a flat or dome port on one end and an O-ring-sealed cap on the other. The electronics stay dry inside, and controls are routed through sealed switches or magnetic couplers if you need to adjust settings at depth. For autonomous monitoring setups where you just press record and deploy, you can skip the controls entirely and rely on an intervalometer or software timer.
Depth rating is another consideration. Every sealed housing has a limit set by the weakest point in the enclosure, usually the port or a cable penetrator. Standard recreational diving depths, around 30 to 40 meters, are achievable with relatively simple construction. Going deeper than that requires thicker walls, higher-quality seals, and ideally a pressure test before deployment.
Camera Settings That Actually Matter
Underwater photography doesn’t require a fundamentally different camera, but it does demand you use your camera differently than you would on land. A few settings make an outsized difference.
- White balance: Water absorbs warm wavelengths first, so everything looks blue-green within a few meters of depth. Setting a custom white balance underwater, or shooting in RAW so you can adjust it later, is the single most impactful thing you can do for color. Auto white balance generally overcompensates or gets confused by the monochromatic light.
- Aperture: Sharpness through a flat port degrades at the edges when you shoot wide open. Stopping down to f/8 or f/11 helps with both edge sharpness and depth of field. With dome ports the sweet spot is similar, though the optical penalty for shooting wide open is less severe.
- ISO: Water eats light quickly. Even in clear tropical water, you lose about a stop of light every five meters or so. Higher ISO lets you keep your shutter speed up, which matters because particles in the water (backscatter) become more visible with longer exposures. Use the lowest ISO you can get away with while maintaining a fast enough shutter speed for your subject.
- Shutter speed: Fast-moving marine life demands at least 1/125 second, and 1/250 or faster for darting fish. If you’re shooting video, your shutter speed is typically locked to double your frame rate, so compensate with aperture and ISO instead.
Shooting in RAW is nearly universal advice among experienced underwater photographers because the color shifts are so extreme that JPEG compression throws away exactly the color data you’ll need later. RAW files preserve the full range of captured information and let you pull back detail in the red channel that JPEG would have discarded.
Where Color Goes and How to Get It Back
Water is not a neutral medium for light. It absorbs different wavelengths at different rates, and red light is the first casualty. By about five meters deep, most red wavelengths have been stripped out. Orange follows by ten meters, yellow by fifteen to twenty. By the time you’re past twenty meters, you’re working with blues and greens only, and everything looks like it’s been shot through a cyan filter. This is true even in the clearest water on earth; it’s an inherent property of how water molecules interact with electromagnetic radiation.
Artificial light is one way to fight this. A strobe or video light carried alongside the camera reintroduces the full spectrum at close range, typically within a meter or two of the subject. This is why macro underwater photographers get vivid color while wide-angle shooters often end up with blue haze: a strobe can illuminate a nudibranch at arm’s length, but it can’t light an entire reef wall. For wide scenes, you’re stuck with whatever the ambient light provides, and that means color correction has to happen in post.
Red filters placed over the lens or behind a dome port can partially compensate in shallow water, roughly the top ten to fifteen meters. They work by blocking some blue-green light, which shifts the overall color balance back toward warm tones. The downside is they cost you about a stop of light, and they’re optimized for a specific depth range. Go shallower and everything turns pink; go deeper and they can’t overcome the loss. Filters are a compromise: useful when you can’t bring a strobe or a video light, but never as good as reintroducing the missing wavelengths artificially.
Post-Processing for Underwater Images
Even with good in-camera technique, most underwater images benefit from post-processing, and the science of doing this well has become increasingly sophisticated. The fundamental challenge is that light doesn’t reach every part of the scene equally: closer subjects get more illumination, deeper or farther subjects get less, and the color cast varies throughout the frame based on the water column between the camera and each object.
Recent work on underwater heritage documentation has produced algorithms that estimate the background light in different zones of the image separately, rather than applying a single correction across the whole frame. One method combines a segmented background light estimation with depth-aware transmission mapping and a color compensation model that adjusts corrections based on how far each part of the scene sits from the camera.3npj Heritage Science. Underwater heritage image enhancement and color restoration integrating partitioned background light estimation and deep fusion The practical upshot is that these approaches can recover natural-looking color even in images where different parts of the scene were at substantially different distances and depths.
For most photographers and videographers, the everyday version of this is simpler. Shooting in RAW and adjusting the white balance in software like Lightroom or DaVinci Resolve handles the bulk of the color shift. Boosting the red channel selectively, applying a graduated adjustment to compensate for the brightness falloff toward the bottom of the frame, and using a dehaze slider to cut through the scatter are standard moves. The key is to avoid over-correcting: pushing too much red back into the highlights can give skin tones and white sand an unnatural magenta cast. Correcting the midtones and shadows separately from the highlights helps keep things believable.
Recording Sound Underwater
Sound travels roughly four times faster in water than in air, and it travels farther too, because water is denser and absorbs less acoustic energy. But the mechanics of capturing that sound are completely different from anything a standard microphone can do. A conventional microphone uses a thin diaphragm that vibrates in response to air pressure changes. Submerge it, and the diaphragm can’t move properly because water’s impedance is vastly different from air’s. The result is either silence or a distorted, muffled mess.
Hydrophones are the answer. These are transducers designed specifically to convert pressure changes in water into electrical signals. Most use a piezoelectric element, a material that generates a small voltage when compressed. Because the pressure fluctuations in water are small relative to the density of the medium, hydrophone sensitivity is measured on a different scale than microphones, typically in decibels referenced to one volt per micropascal. A recently developed MEMS piezoelectric hydrophone, for instance, achieved a receiving sensitivity of about −189 dB (re 1 V/µPa) at 920 Hz and operated effectively across a frequency range of 20 to 2,000 Hz.4Sensors and Actuators A: Physical. Design and fabrication of a novel MEMS piezoelectric hydrophone That frequency range covers much of what’s biologically and environmentally interesting in shallow coastal waters, including fish calls, invertebrate sounds, and boat noise.
For filmmakers, the practical setup is usually a hydrophone connected to a portable audio recorder in a dry box or on the surface, with the cable running down to the hydrophone element positioned near the subject. Some systems use self-contained underwater audio recorders that pair a hydrophone with a sealed recorder in one unit, which simplifies deployment but limits your control over gain and monitoring. If you’re syncing underwater audio with video, you’ll need a time-code reference or a clear clapboard-style event on both the audio and video tracks to align them in post.
Spatial Audio Beneath the Surface
Most underwater recordings are mono or stereo, but there’s growing interest in capturing full three-dimensional soundscapes, particularly for immersive media and scientific research. On land, ambisonic recording uses a compact array of microphones to capture sound from all directions simultaneously. Underwater, the same principle works, but the geometry has to change because of the faster speed of sound in water.
One approach uses four hydrophones arranged in a tetrahedral structure, with a side length of about 70 centimeters. That spacing is a compromise: large enough to capture pressure gradients at low frequencies accurately, but compact enough to avoid spatial aliasing at higher frequencies. The signals from the four hydrophones are then transformed into spherical harmonics, producing a first-order ambisonic recording that’s independent of the original microphone arrangement and can be played back in any spatial audio format.5Ingenta Connect. 3D audio recording of complex underwater sound fields
The result is an audio file where the listener can turn their head, virtually, and hear the sound field shift realistically. For virtual-reality films shot underwater, this is a significant step beyond stereo. For scientists studying how marine animals use sound to navigate or communicate, spatial recordings reveal information that a single hydrophone completely misses: the directionality of a sound source, the way reflections bounce off the seabed or a reef structure, and how multiple overlapping sources separate in space.
The Soundscape You’re Recording Into
One thing that surprises people new to underwater audio is how loud it can be. The ocean is not a quiet place. In many shallow coastal environments, the dominant sound above about 1.5 kHz comes from snapping shrimp: small crustaceans that produce sharp, impulsive clicks by rapidly closing a specialized claw. In oyster reef habitats along the mid-Atlantic coast, snap rates during summer can reach 1,500 to 2,000 snaps per minute, dropping to fewer than 100 per minute in winter. Sound pressure levels track snap rate closely, varying by about 15 decibels across the seasons in the 1.5 to 20 kHz range. Water temperature is the strongest predictor, with snap rate and temperature strongly correlated, and light availability adds a daily rhythm on top of the seasonal pattern, with many days showing a preference for nighttime or daytime snapping and additional bursts around dawn and dusk.6PubMed Central. The Curious Acoustic Behavior of Estuarine Snapping Shrimp: Temporal Patterns of Snapping Shrimp Sound in Sub-Tidal Oyster Reef Habitat
For practical recording purposes, this means your underwater audio will almost certainly contain biological noise that you didn’t plan for. In tropical and subtropical waters, snapping shrimp can dominate everything above a few kilohertz, making it difficult to isolate other sounds in that band. Fish choruses, particularly from species like toadfish and damselfish, add tonal components in the low hundreds of hertz. Boat engines produce broadband noise that can mask everything. If you’re recording a specific sound event, timing your session matters: early morning in winter, for instance, will give you a much quieter background than a summer evening on a tropical reef.
Filtering in post-production can help, but it has limits. A high-pass filter will clean up low-frequency flow noise from currents passing over the hydrophone, and notch filters can reduce specific tonal interference. But broadband biological noise like shrimp snaps is spectrally similar to many sounds of interest, which makes it hard to subtract cleanly without also removing the signal you want. The better strategy is to choose your recording location, depth, and time of day with the ambient soundscape in mind. Move away from hard substrates where shrimp concentrate. Record at depth rather than in the shallows where biological activity peaks. And be aware that the acoustic environment changes hour by hour in ways that have no direct parallel on land.