The choice of cameras, housings, and mounts affects which subjects can be recorded, how clearly important features can be seen, and whether the relationship among camera views remains stable. Useful systems range from compact action cameras to interchangeable-lens cameras and tethered video systems. Evaluate the complete system under the conditions in which it will be used.
This page covers camera and mount selection. See Camera settings for resolution, frame rate, focus, exposure, stabilization, and file formats, and Lighting and cinematography for VidSync data for visibility, motion blur, glare, contrast, and pilot review.
Choose the camera as part of a system
A camera specification does not describe the finished recording system. The lens, housing port, recording mode, battery, storage medium, mount, stabilization and image-processing settings, and any external cables or recorders all affect the resulting footage. For underwater work, test the camera through the intended housing and port rather than evaluating it only in air.
Begin with the observations the study requires. A camera used to detect large animals has different demands from one used to locate small landmarks, distinguish individual markings, follow fast motion, or record for several unattended hours. Important selection criteria include:
- adequate detail throughout the intended field of view;
- progressive recording at a frame rate suited to the event;
- stable, repeatable focus and focal length;
- recording formats that work through the complete VidSync workflow;
- sufficient recording duration, power, storage, and thermal tolerance;
- controls that can be recorded, fixed, or disabled as required;
- a housing and port suitable for the environment; and
- a secure way to attach the camera to a rigid mount.
Matching camera models and recording modes can simplify a multi-camera system because the views have similar resolution, frame rate, color, delay, and field of view. Mixed cameras can also work, but each combination should be tested with a shared visible timing cue and representative reference footage.
Focus and optics
Cameras with adjustable lenses are most useful when focus and focal length can be set manually and repeated. Fixed-focus action cameras can also be effective because their lenses are commonly designed for a broad depth of field. In either case, inspect the nearest and farthest subject positions and the corners of every view.
Underwater housings change the optical system. Flat ports narrow the field of view through refraction and may reduce sharpness away from the center. Dome ports have different focusing and distortion behavior and can introduce reflections or lighting artifacts. Performance depends on the lens, port dimensions, camera position within the housing, subject distance, and water conditions. Test the exact combination rather than relying on a general claim that one port type is preferable.
The camera must remain fixed within the housing. A housing can be rigidly attached to a crossbar while the camera shifts slightly inside it. Use fitted trays, locating pins, firm clamps, or other repeatable supports when available. Check that closing the housing, connecting cables, or pressing external buttons does not move the camera or lens controls.
Recording duration, power, storage, and heat
Advertised battery life and storage capacity provide only starting estimates. Record continuously in the intended mode for at least as long as a real deployment. Confirm whether the camera stops, overheats, dims the display, changes recording mode, or divides the recording into multiple files. If files are segmented, verify that the boundary is gap-free and that frames and timing cues remain continuous.
Calculate storage from the actual bitrate of the chosen mode and include every camera, pilot recordings, repeated deployments, backups, and preserved originals. For reference, 4 MB/s (32 Mb/s) produces approximately 30 GB in two hours. HEVC at about 2.5 MB/s (20 Mb/s) produces approximately 18 GB in two hours and may preserve similar quality for comparable 1080p footage. These are starting points. Compression performance varies with the camera, encoder, motion, noise, and scene, so compare representative clips and confirm playback in VidSync before converting a dataset.
Long recordings can generate substantial heat. A waterproof housing may insulate the camera even in cold water, and direct sunlight can heat both the housing and the equipment inside it. Test the system at the expected air or water temperature and in the expected sun exposure. External power and recorders extend operating time but add cables, connectors, failure points, and their own thermal limits.
Self-contained and tethered systems
Self-contained cameras are compact, portable, and simple to deploy. Their battery, recording medium, display, and controls are enclosed with the camera. Underwater housings often prevent a live view once the camera is in position, so aim, focus, exposure, and recording status must be checked before deployment or by briefly retrieving the camera.
Tethered systems send a live view to an external recorder or computer. A live view can reveal poor aim, focus, exposure, glare, or an obstructed subject area while corrections are still possible. External power and storage can support long sessions. The cables and connectors make the system less portable and can tug on the mount, leak, disconnect, or introduce electrical and operating complications. Secure cables so their weight and movement do not change camera position.
Camera arrangement
A common VidSync system uses two cameras attached to a shared crossbar, although other arrangements are possible. The cameras must have overlapping views of the subject and any reference objects. Camera separation, viewing directions, subject distance, field of view, and the practical size of the calibration or reference object all influence the arrangement.
When a subject is far away compared with the separation between cameras, the viewing rays from the cameras meet at a narrow angle. Small differences in the observed ray directions then correspond to larger differences in the reconstructed position. Increasing camera separation or moving the cameras closer to the subject generally increases the intersection angle, but it also reduces overlapping field of view and can make the system and its reference frame harder to handle. Laboratory arrangements can often use wider intersection angles than field systems.
Underwater systems usually benefit from short camera-to-subject distances because water and suspended material reduce contrast. Wide-angle optics help retain the needed field of view at short range, although their behavior through a housing port must be tested. Arrange the cameras around the expected subject volume rather than around a single central point, and confirm that important positions remain visible in every required view.
Use pilot recordings and known reference objects to evaluate a proposed arrangement before committing to a full study, especially when selecting a new geometry.
Mount rigidity
The mount must hold the cameras steady relative to one another while the associated observations and reference material are recorded. Movement can occur at tripod heads, action-camera joints, quick-release plates, clamps, crossbar connections, housing attachments, or inside the housings. Cables, current, buoyancy, people, and animals can apply forces that were absent during setup.
Use the fewest adjustable joints practical. Tighten each connection with an appropriate tool, within the limits of the hardware, and prevent controls from slipping. Lock washers, keyed or pinned connections, fitted camera trays, safety lines, and visible witness marks across joints can make movement less likely or easier to detect.
Rigidity checks before and after recording
- Assemble the complete system. Include housings, cameras, batteries, cables, external recorders, and every mounting part used in the field.
- Apply realistic forces. Gently test each camera and joint in the directions likely to be loaded by handling, cable tension, current, or buoyancy. Watch for rotation, flex, clicking, or a camera shifting inside its housing.
- Record a reference scene before deployment. A rigid target with fine stationary features can reveal movement or slipping when views are compared.
- Inspect the system after it is positioned. Check witness marks, fasteners, focus and zoom controls, recording status, and cable strain before collecting the primary observations.
- Complete associated reference footage before moving the cameras. If the system is bumped or adjusted, document the time and collect the material required for the changed setup.
- Repeat the inspection after recording. Record any movement, loose fastener, shifted control, water intrusion, or uncertainty while the setup and events are still fresh.
Opening a housing, changing a battery, removing a camera, or adjusting the mount can alter the relationship among views. Very repeatable hardware may allow a setup to be reproduced, but that should be demonstrated with reference recordings rather than assumed.
Underwater housings and condensation
Inspect and clean housing seals before every deployment. Follow the housing manufacturer’s instructions for O-ring cleaning, lubrication, replacement, and pressure testing. Keep hair, sand, fibers, and excess lubricant away from sealing surfaces. A leak test without the camera can be useful after a housing has been serviced or reconfigured.
Warm, moist air sealed inside a housing can condense against a cold lens port. A dry desiccant packet or purpose-made insert can reduce fogging when it fits without interfering with seals, buttons, heat dissipation, or the camera. Store housings open in a dry environment before use, close them in dry air when possible, and inspect the port for condensation during pilot recordings.
Clean the inside and outside of the port. Water spots, residue, fingerprints, scratches, and droplets close to the lens can obscure detail or create glare. Verify that the housing does not vignette the chosen field of view and that buttons can be operated without shifting the camera.
Historical example: DSLR cameras with external recorders
During one Alaska research project, we used Nikon D5300 DSLR cameras connected to Atomos Ninja recorders on shore. The cameras available to us stopped internal recording after approximately 30 minutes, so the system used external recording, external power, modified hardware, heat sinks, and a particular startup procedure. It produced high-quality recordings more than five hours long, but the system was complicated and temperamental. I would not recommend reproducing it when a simpler camera can meet the study requirements.
Heat was a recurring problem. The underwater housings insulated the cameras, and sunlight added heat even in cold water. In our particular setup, an opaque coating reduced solar heating inside the camera housings. The external recorders then became the limiting component, so we placed them between aluminum plates in shallow cold water as improvised heat sinks. This is retained as an example of the system-level problems that can appear during long recordings, rather than as a general cooling method.
Historical notes and checks for action cameras
GoPro and other action cameras have been used successfully with VidSync. Their low cost, small size, wide field of view, fixed-focus operation, and simple housings can make them practical for tight spaces and portable field systems. Older models also illustrated several limitations: short battery life, segmented files, limited control over exposure and focus, and reduced underwater sharpness near the corners. Current models and recording modes vary too much for those observations to serve as purchasing advice.
Use the following checks for any action camera:
- Disable electronic stabilization and horizon leveling for a rigid measurement setup unless the complete workflow has been tested with the feature enabled.
- Identify the real crop and field of view for the selected resolution, aspect ratio, lens mode, and frame rate. Labels such as Wide, Linear, Medium, and Narrow do not have consistent optical or quality implications across models.
- Test file segmentation. Record through at least one file boundary and verify that no frame or timing interval is lost when the clips are used in VidSync.
- Check timing. Confirm constant frame intervals and compare a shared visible timing cue across all cameras.
- Inspect the full underwater image. Check focus, distortion, vignetting, reflections, and sharpness at the center and corners through the intended housing port.
- Tighten mounting joints with appropriate tools. Hand-tightened action-camera thumbscrews can rotate when bumped. Apply only the force the hardware is designed to tolerate.
- Verify power, storage, and heat behavior over the full intended recording duration.
- Confirm orientation before deployment. If the camera is mounted upside down, verify that the recorded orientation remains fixed and is handled as expected throughout the workflow.
- Use live preview when available to check aim and exposure before the camera is submerged. Confirm whether preview remains available in the deployed configuration.
System verification checklist
- Record continuously for the full expected deployment time.
- Verify progressive scan, frame timing, file boundaries, codec compatibility, and simultaneous playback.
- Inspect focus, field of view, corner sharpness, motion blur, exposure, and compression in realistic footage.
- Confirm that stabilization, crop, lens correction, focus, and zoom remain fixed.
- Test the mount for movement with housings, cameras, cables, and power installed.
- Record reference material before and after a representative deployment.
- Document the verified setup in the field protocol.