Why aren't LED lights revolutionising film and TV?

LEDs (Light Emitting Diodes) are often described as efficient: a one-hundred-watt LED light will be a lot brighter than a one-hundred watt tungsten-filament domestic light bulb. Perhaps four times brighter. We need a quarter the light, and four times the light is emitted. That's a total of one-sixteenth the power required in our lighting. Shouldn't lighting budgets have dropped off a cliff?

Well, sort of. If we need, say, 75% less light, that alone won't reduce the lighting budget by 75%, because a hundred-watt light isn't a quarter the price of a 400-watt light. It can save some productions the expensive need to bring in generator trucks and it allow others to run entirely from batteries, and smaller, handier lights are often cheaper to transport and faster to rig with smaller crews. Still, our assumption of a four-to-one efficiency saving in LEDs only applies when they're compared to old-style tungsten lights. Fluorescent lights and HMIs (an abbreviation of hydrargyrum medium-arc iodide), the modern equivalent of an arc light, have been available for decades.

Surely, modern LED technology must produce more light per watt than a mid-20th-century arc light? Perhaps by a little, and this is mainly because of mechanical considerations. An HMI bulb emits light in all directions and engineers must include reflectors and lenses to throw as much of it as possible in a useful direction.

The arrangements for LED emitters can be simpler and the results can sometimes be a bit more efficient than HMI. The difference is perhaps 20% or 25%, depending heavily on the sort of beam that's being produced. That's less than one f-stop, so the efficiency advantage of LED over HMI doesn't often make a noticeable difference to production budgets in terms of power consumption or generator rental.

The main reason that LED hasn't replaced everything is that HMI is available at far higher power levels. It’s a field that’s moving fast, but until the end of 2016 the most common LEDs were around a hundred watts each. LEDs rated at 400 to 500 watts have since become more common, but it's still only similar to the very smallest common HMIs at 400W or 575W. HMIs, on the other hand, are available up to 24,000W for upscale productions.

In parts of the world with 220-volt mains power, it's common to use 2,500W HMIs even on very modest productions, since they can be run from a wall socket. No LED available in early 2017 approaches a 2,500W HMI for sheer power. Mole-Richardson's Tener LED, at 1,600W, is probably the closest, though it's really intended to equal a 10,000W tungsten light.

So, the new technology is great and promises much, but it isn't a universal replacement for the old guard quite yet. Light colour is also an issue, a bugbear of early LEDs that could seem cold and cyan and make healthy actors look freshly exhumed. This is largely a solved problem, especially if we overlook unreliable measures of colour quality such as the Colour Rendering Index (CRI).

CRI wasn't designed to work with things like LEDs, and specifications according to the Television Lighting Consistency Index (TLCI) are better. LEDs with a TLCI above 90 should show no serious problems with colour, but often that doesn't matter: they still won't have the power to replace any but the smallest HMIs.

Within that lower power range, let's say up to 500W, LED is actually very competitive on price. HMI lights of 575W HMI lights are a workhorse. A lamp head alone, without a bulb, sells for $2,500 at specialist US store B&H Photo Video, with the required control electronics – the ballast – another $2,500.

It's difficult to find exact equivalents (there is no LED at exactly 575W) but a Mole-Richardson 400W Junior LED is far cheaper, at around $2,500. With B&H's discount at the time of writing, the LED is getting on for half the price, making for a cheaper rental even if it doesn't save much power over the HMI.

Another factor in favour of LED is that HMIs can't really be dimmed – well, they can, but only to about 50% of maximum, and they go blue as they get dimmer. Fluorescents go purple. LEDs can fade to black without any colour shift at all, which is unique among lighting technologies, and this is one reason why they're popular for studio installations.

Saving even small amounts of power in a news studio can be very worthwhile, because the studio may run long hours and because the energy that goes into heating up the air in the studio must be removed with power-hungry air conditioning. Dimming concerns mean that neither HMI nor fluorescent have ever been contenders in that environment.

Portability matters, too. LEDs also don't generally require the ballast of an HMI, which can be big, heavy, and, as we've seen, expensive (they may sometimes require an external power supply, though this tends to be small and light). They're more likely to run from batteries, although it's worth bearing in mind that increasingly powerful LEDs will be just as difficult to battery power as any HMI or fluorescent of the same wattage.

A minority of LEDs even provide variable colour, so they can match either daylight or artificial light, or be tuned to match the pre-existing lighting of a location, or produce special saturated colour effects. These effects trade off brightness for flexibility, so they rob LED of its ability to compete on sheer power and generally increase prices significantly.

There are many things that LEDs do well – when out shooting news, when working somewhere without power, or when carrying gear to the top of a mountain. There is no category of production which needs to discount LEDs entirely. It's just a matter of understanding that they don't do much more, watt for watt, than pre-existing technology, and that even largest LEDs are currently fairly small lights in the grand scheme of things.

This will change, given that LED development is driven by the domestic lighting market, and the film industry is likely to continue to benefit from the huge amount of research and development that's being poured into the field.

Ultimately, when a new technology is seen to fail – as LED sometimes does when it's asked to do a job it can't handle – resistance to that technology grows. That's a shame, because the future is, literally, bright, but the only way to get there is patience. In the meantime, we can't assist with the march of progress if we ask LEDs to do too much.

Images: Phil Rhodes