How to build a properly useful small underwater camera

Jules Charlton has spent many a dive fantasising about the ideal underwater camera and complaining about the real world alternatives. This is his view from the dive boat…

I have been fond of pocketable, waterproof cameras ever since I bought a Kodak Playsport HD camcorder in Currys for the bargain price of £50. It may be trite to say that the idea of a small, go-anywhere snapper is so compelling that it resulted in competitive, lucrative businesses for GoPro, DJI, Insta360 and so on but — I would argue — if you are a recreational scuba diver, as I am, you are remarkably ill served by current market offerings and thus make do in a way that you would not tolerate in respect of any other piece of your kit, such as your torch or dive computer.

To demonstrate what I mean and to explore what I think a proper, submersible pocket camera should look like, I would like to set out some key, minimum features for such a lenser.

These features may be grouped under a discussion of:

1. Housing
2. Image Quality
3. Image Stabilisation

The Question of Housing

Let's start with housings and that all-important specification of depth rating. This may be considered to be not only a rating of resistance to water ingress in respect of water pressure equivalent to depth, but also an indicator that a camera may operate normally for an extended period up to that rated depth without substantial risk of malfunction or system failure. You can have all the advanced software driven features you want, but if your camera cannot perform at your preferred submersion it's useless for your needs.

The current limit for open water divers on air in the PADI system is 40 metres [131 feet] so any camera that a sports diver may take underwater should ideally have a depth rating equal to or greater than that. Having said that, the practical limit for most scooby doos is 25 to 30 metres.

The great and much missed Nikon W300

Before it was discontinued by Nikon, the Nikon W300 was considered a preferential choice by recreational divers primarily because, apart from being shock resistant, its casing was water resistant to 30 metres without the need for additional housing. It may not have RAW capture like the OM TG-7 or a large sensor but you can jam it into the pocket of a BCD or clip it onto a shockline and you are good to go. You don’t need to have it handed to you over the side of the boat, you aren't worried if it gets banged about a bit, and it leaves your hands free if required to do other tasks such as check your dive computer, check your air gauge, or carry out buddy breathing if that was necessary.

Another option might be the Sealife Micro 3.0 which is permanently sealed — without a removable battery or media card — and has a depth rating of 60 metres but is bulkier and generally intended to be used in conjunction with a tray and additional LED lights.

By contrast, current market offerings such as the OM TG-7 and the Ricoh WG-6 are only water resistant to 15 metres and 20 metres respectively, far short of the PADI limit. 

Action camera limitations

The situation is equally unsatisfactory when it comes to action cameras. The DJI Osmo Action 4 is water resistant to 18 metres and both the Insta360 X3 and the GoPro Hero 12 Black are water resistant to a mere 10 metres. These cameras all need additional protection at a further cost to operate at greater depth and this brings up the issue of the need for an external housing for your device.

If you operate at the high-end of underwater imaging making then you will be familiar with companies such as Gates and Nauticam who make housings not just for DSLRs and mirrorless cameras - such as the Canon 5D MK IV and Nikon D850 - but also the ARRI Alexa Mini. Their housings cost, in some cases, the same or more than cameras they are designed to protect, are model specific, and require targeted training to rig up and use properly because if you flood an ARRI…

This stunningly engineered Nauticam housing for the ARRI Alexa Mini costs over $31,000: protecting expensive kit costs!

However, we are focused here on the average diver who wants to take a few snaps or capture short video clips, not the camera team of Blue Planet, so what are the options for us? 

OM Systems offers an external housing for the TG-7 that extends its operable depth to 45 metres but this adds a significant cost to your additional outlay and also means the camera can no longer be tucked into a BCD. DJI and GoPro respectively offer housings that operate to 60 metres, as does Insta360 whose Invisible Dive Case for the X3 has a depth rating of 50 metres. 

So, what's the issue?

In a word, plastic. These housings are made with some form of acrylic which means that they are fragile and may be prone to heat expansion resulting in water ingress. They should be transported on a dive boat — where space is at a premium — in towels and/or protective boxes, and once in use they may fog, especially in hot climates. I have had two DJI housings in a row fail on me as well as an Insta360 case on loan from a friend that developed a crack in its dome and flooded at 25 metres. They all fogged.

There may be an expanding market for smartphone housings such as the Oceanic+, Sealife SportDiver or Divevolk but I, for one, don’t fancy taking the chance of a financial and data hit if my shiny iPhone gets fritzed. Furthermore, a number of such housings rely on some form of touch sensitive panel to interact with the screen, and that is of limited use at greater depths, in cold water, or where you need to wear diving gloves.

Image quality

You wouldn’t guess it by watching advertorial videos - sorry, product reviews - shot just below the surface under ideal conditions at midday in places like Hawaii and Tahiti, but water does funny things to light. The deeper you go the greater the loss of luminance and chrominance you encounter, not to mention issues such as suspended particles in the water column that cause backscatter. 

This is why sports divers prize cameras that can shoot RAW files over JPEGS even when they bring additional lighting because, at just two metres of depth, colour shift is already significant. 10 bit 4.2.2 log files for video are equally desirable because, as with RAW files, they allow for significant post-processing correction back home. 

As we have seen, modern action cameras by DJI and GoPro - as well as the TG-7 - all offer RAW file capture, but we are stuck with those potentially leaky housings. 

A larger sensor, such as may be found in the Sony R100, range may ameliorate low light levels and extend the flexibility of RAW/LOG files, but would only make sense for our purposes if it meant that such a camera could still be slipped into your BCD.

Another desirable feature for video, apart from 10-bit log, is high resolutions and frame rates. One of the secrets of good underwater footage is to be able to shoot in 4K at 100 fps (or higher) and then slow your footage down to 25fps in the edit. Those fish swim with such languid grace, they stay in frame for more than a fraction of a second, you see their scales in sharp detail and your own dodgy, buoyancy dependent framing appears smoother! Pity that neither the TG-7 or WG-6 shoot at those frame rates in 4K.

The OM Systems TG-7 has many great features, but falls short on high frame rates

Finally - unless you are addicted to licking your lens - some form of scratch resistant hydrophobic lens coating that reduces or eliminates splash distortion is a must.

Image stabilisation

Recent advances in image stabilisation, as may be seen with the current generation of DJI and GoPro cameras, is nothing short of excellent. Features like horizon lock,  that gives video footage a gimbal-like feel of stability, is desirable when filming underwater, especially if you consider that a diver is subject to currents and other mobility factors such as finning and buoyancy. This level of image stabilisation should be a standard feature on any small underwater camera in 2024 but, while the TG-7 and the WG-6 both have some form of EIS, it falls short of that standard, as does the Sealife Micro 3.0. 

Underwater camera conclusions

From what we have seen above, a number of cameras available to buy may have some of those features but none is a complete package. So what might a small underwater camera really suitable for sport diving look like in 2024?

• Waterproof to 60 metres without the need for an external housing
• Shock and freeze proof
• RAW capture for stills
• 10-bit 4.2.2 LOG for video
• Video capture equal to or higher than 4K at 100fps (in all image profiles including LOG)
• Horizon lock and stabilisation equal to or better than the GoPro Hero 12 Black or DJI Osmo 4 Action
• A scratch-proof, hydrophobic lens coating
• A bright screen that can be read in direct sunlight or at depth
• Size and weight that enables it to fit neatly into the side pocket of a BCD 
• Reliable WiFi image transfer capability
• Image sensor size of at least 1” and at least 4K compatible that still allows for some form of optical zoom and macro capability

Do I think there is a possibility that a manufacturer may come up with a camera like this? 

I freedive as well as scuba but, frankly, I am not holding my breath.