Skip to content

Lumens vs Candela: What Actually Makes a Flashlight Bright

Updated 2026-08-16

Short answer

Lumens are the total light a torch produces; candela is how concentrated that light is in the centre of the beam. Lumens tell you how much you're lighting up, candela tells you how far you can see. A 1,200-lumen flood light and a 1,200-lumen thrower make the same amount of light — the thrower just aims almost all of it at one small spot, which is why it reaches 700 metres and the flood light reaches 90.

Three units, three different questions

Almost every argument about flashlight brightness is really people using one word for three different measurements. Lumens are luminous flux: the total quantity of visible light leaving the torch, in every direction it goes. Candela is luminous intensity: how much light is travelling in one particular direction, which for a flashlight means the middle of the hotspot. Lux is illuminance: how much light actually lands on a surface, which depends on both the candela and how far away you are standing.

The relationship between them is the whole game. A torch spreads its lumens over a cone. Squeeze the cone narrower with a deeper reflector or a tighter optic and the same lumens produce far more candela, because the light is concentrated. Widen the cone and candela collapses while the lumen figure on the box stays exactly the same. Nothing was added or removed; the light was simply redistributed.

This is why buying on lumens alone goes wrong so often. Somebody wants to identify an animal at the far end of a paddock, buys the highest-lumen light they can find, and discovers it is a 4,000-lumen flood light that lights up the fog between them and the paddock beautifully and shows them nothing at 200 metres.

How beam distance is calculated (and why it's optimistic)

The beam distance printed on a box comes from the ANSI/PLATO FL1 standard, and it is a pure calculation from candela rather than a field measurement. The standard defines a light's throw as the distance at which the beam has fallen to 0.25 lux, which is roughly the illumination of a clear night under a full moon. That works out to distance in metres equals twice the square root of the candela.

So a light rated 10,000 candela throws 200 metres, and one rated 250,000 candela throws 1,000 metres. Because it is a square-root relationship, quadrupling candela only doubles the rated distance — a brutal law of diminishing returns that explains why dedicated throwers get so large. The reflector has to get physically deeper and wider to keep buying reach.

Treat that number as a ceiling, not a promise. Full-moon illumination is enough to notice that something is there, not enough to identify it. A practical rule from people who use throwers is that usable identification distance is roughly a third to a half of the FL1 figure. A light rated at 1,000 metres is a confident few-hundred-metre light in real use.

Reading the two numbers together

Dividing lumens by candela gives you the solid angle the light is spread across, and that single ratio tells you the beam's character faster than any marketing copy. A high ratio means a wide, even beam; a low ratio means a tight column. Every light in this catalog has its beam drawn from exactly that calculation, so you can compare shapes at a glance instead of reverse-engineering them from two numbers.

Some rough anchors. Under about eight degrees you have a pencil-beam thrower that is genuinely awkward indoors. Eight to sixteen degrees is a classic tactical or search beam. Sixteen to thirty is the balanced zone most good EDC lights live in — a defined hotspot with enough spill to see your feet. Thirty to sixty is flood territory, ideal for camp chores, working under a car, or a headlamp. Above sixty degrees you are essentially holding a lantern.

The other thing worth checking is whether the two numbers were measured at the same moment. Manufacturers usually quote both at turbo, thirty seconds after switch-on, which is the FL1 convention. Small lights cannot hold turbo for long, so those figures describe the first half-minute rather than the light you will actually be using five minutes into a walk.


Common questions

Is more lumens always brighter?
It is always more total light, but not necessarily brighter where you are looking. If the extra lumens go into spill rather than the hotspot, the scene in front of you can look no brighter at all — and slightly worse, because more spill means more light bouncing back off nearby objects and dust, which hurts your night vision and contrast.
What is a good candela figure?
It depends entirely on the job. Around 2,000-6,000 candela is plenty for a general-purpose pocket light. 10,000-30,000 suits a light you want to reach across a car park. Serious throwers start around 100,000 and dedicated long-range lights run 300,000 to 700,000, which is where the physically large single-emitter designs live.
Why do two lights with the same lumens cost so differently?
Lumens are the cheapest specification to hit. What costs money is holding those lumens without overheating, delivering them in a beam with no artefacts, rendering colour honestly, and surviving being dropped. Sustained output, driver quality, emitter binning and machining tolerances are where the money goes, and none of them appear in the headline number.
Does the ANSI FL1 rating get audited?
No. FL1 defines how to measure, but nobody polices whether a manufacturer followed it. Reputable brands measure honestly and are close to their claims; some budget listings quote emitter-datasheet maximums that the light never produces. Independent measurements from reviewers with integrating spheres are the only real check.
What is 'turbo' and why can't a light hold it?
Turbo is the highest output step, and it is limited by heat rather than battery. A small aluminium body can only shed so many watts, so the driver steps output down — often within 30 to 90 seconds — to stop the light cooking itself and its cell. Sustained output, not turbo, is the number that describes what a light does on a long task.

Keep reading