I have had this discussion with a few people since the start of 2014 already, and thought I might as well turn it into an informative post. So get some coffee quickly and then sit down for the read…
Most divers looking to get into shooting underwater video take an instant interest in GoPro. It’s powerful, compact, and although fairly expensive does not break the bank as badly as even entry level video cameras with similar capabilities and their respective underwater housings will. But there is a key element to keep in mind: Where some video cameras might have an underwater mode (especially those for which there are housings available), the GoPro does not. The specifications merely state that it CAN go to 40/60m under water and that the flat lens will allow you to focus whilst down there. It is however made primarily for shooting on land where light is ample and it’s software colour correction caters for this, which means the moment you hit the big blue, the image processor tries it’s very best to correct the now overwhelming blue colour by adding some yellow, and you are left with bright green video which is unappealing and pretty much useless.
There have been several manual fixes to counter this, the most popular being the red lens cover that manually adds the red that disappears at depth thus fooling the image processor into thinking everything is fine and dandy. The problem with this however is that the red lenses only tend to work as intended at a very specific depth, everything outside of which will again be susceptible to the colour correction and thereby lead to unnatural colouration in your video. This can be remedied to an extent in post processing, but it’s next to impossible not to mention a time consuming and intricate process.
The alternative is of course to go the extra mile and get video lights. Although more expensive, the immediate benefit is that they offer full spectrum light, restoring the natural colours of your subject thus giving you rich, colourful video at any depth with very little if any post processing needed. The downside unfortunately is that video lights are expensive, and entry level lights lack the punch to shoot from a distance whilst maintaining a decent amount of light on your subject, thus limiting your shooting distance to a mere meter or two. That being said, ANY video light is better than no light of all.
So what makes a video light different from a regular flashlight?
Well, first of all, all flashlights make use of some form of collimator (device that narrows the beam of waves) meaning that a fairly low power light source can be focussed into a usable light, thus keeping the cost of the hardware lower, in turn lowering the production and eventually retail price. Video lights however need a wide, even beam of light, almost 180 degrees, to ensure that the entire field of view is equally lit to avoid “hotspots” where image processors lower the ISO (light sensitivity) to normalise the colour of the subject, thus darkening the surrounding areas to a point where the are invisible. The result is a 007 intro style video which is completely useless.
Here’s a comparison I shot between using my LED Lenser D14 and using my video lights
Since the same amount of light which would under normal circumstances be adequate for a normal flash light is now dispersed, a higher power light source is required to provide a usable amount of light without collimating the beam. This means high end LED’s or HID lamps, meaning better hardware, and in turn higher production and retail cost. But how much higher..? Most high power LED lights make use of Cree LED’s, which are world renowned for having exceptional watt-to-lumen ratios. A very popular LED in modern mobile lighting is the Cree XM-L T6 which is a 3.7V 10W cool white LED with a maximum output of around 1000lm: Three times more than the advertised power of the LED Lenser D14, and all this from a single LED. The problem however is creating a portable power source capable of providing 3.7V at 10W for long enough to at least do a decent dive, which has led to manufacturers opting for more LED’s at a lower wattage instead. This is why you will find video lights sporting 3x XM-L LED’s, but only advertising a max power of 1000lm. Most of these XM-L lights use type 18650 batteries or a derivative thereof, which is a 3.7V battery meaning that very little current alteration is required since the voltage already matches that required by the LED, i.e. less electronics, lower production cost.
As I’ve bought all of the components separately, I have a fairly accurate idea of what the internals of these lights total and you’ll be surprised to know that the electronics for 2x of these lights, including a CREE XM-L T6, LED driver, reed switch and 18650 battery retails for little over $20… So where does the other $180 come from? The housing of course! As a manufacturer, you need to create something that is simple to use but also has the lowest possible chance of flooding which would ruin the internals and potentially lead to explosions, damage to property, claims and all of those nasty things. As such a lot of effort goes into making housings which are put through rigorous testing to ensure that their chances of flooding under normal use is minimal. The most popular way to achieve this is to hermetically seal the housings with merely a small, easily waterproofed charging port as a point of contact to the inside. The result is a shared peace of mind, where you as the customer don’t have to stress about flooding and incurring additional costs after the initial purchase, and they as the manufacturers don’t have to worry about any claims making their way back to them.
All that being said, if you are feeling handy, you might want to have a look at my little step by step on building your own, fairly fail-proof video lights to try out before making the purchase.
Here’s the instructional video
…and here’s the difference they make, shot at 20m