Flashlights: Lumens, Candela and What FL1 Numbers Really Mean

Last verified: August 19, 2026. All specifications are the manufacturers’ own published figures, checked on that date. Standards citations verified against ANSI, PLATO, IEC and NIST. Popular EDC does not perform laboratory testing — every number below is either published by the manufacturer, published by a standards body, or arithmetic derived from those.

What is the ANSI FL1 standard, and which version is current?

FL 1 is the performance-reporting standard behind the icons on flashlight packaging. Its lineage matters, because the category is citing a version that has been superseded twice.

Status per the ANSI Webstore listings and PLATO’s publication announcement, checked August 19, 2026.

The name most often printed in the wild — “ANSI/NEMA FL1” — credits a publisher that has not maintained this standard since 2016. PLATO, the Portable Lights American Trade Organization, did not exist when the 2009 edition was written; per its own history page, members began discussing an industry body in May 2010 and PLATO incorporated that September.

Worth knowing who wrote it. The 2009 committee roster, printed in the standard’s own foreword, is the manufacturers the standard governs: Dorcy, Princeton Tec, Coast, SureFire, Golight, Petzl, Brinkman, Energizer, ASP, Streamlight, Cat Eye, Black Diamond, Coleman and Duracell.

What FL1 actually measures

Six rated parameters in the 2009 edition: beam distance in metres, peak beam intensity in candela, run time in hours, light output in lumens, impact resistance in metres, and water ingress protection as an IPX rating. FL 1-2025 adds a seventh, Short-Duration Elevated Brightness — turbo — plus dust ratings IP5X and IP6X.

The three protocol facts that change how you read a spec sheet

1. Light output is measured 30 to 120 seconds after switch-on. The 30-second mark is the defined initial output value that anchors the runtime test. That instant sits inside the turbo window for essentially every high-output light. Acebeam’s X75 holds 80,000 lumens for 30 seconds; the FL1 reading is taken at the exact moment the light is about to collapse to 12,000.

2. Run time is measured until output falls to 10 percent of the initial value. Not 50 percent. Not “until it gets noticeably dim.” Ten percent. A light that starts at 3,000 lumens is still officially running at 301 lumens, and the clock only stops at 300.

3. Beam distance is the distance at which illuminance falls to 0.25 lux — roughly the light of a clear full moon. Because illuminance follows the inverse-square law, that makes beam distance a pure function of candela:

where I is peak beam intensity in candela and d is the FL1 beam distance in metres.

Two consequences worth internalising. Beam distance scales with the square root of candela, so quadrupling candela only doubles the rated distance. And 0.25 lux is barely enough to perceive terrain — FL1 beam distance is a detection figure, not a useful-illumination figure. At half the rated distance you get 1 lux; at one-tenth, 25 lux.

FL1 compliance is self-declared, and third-party certification barely exists yet

A UL and PLATO certification partnership launched September 14, 2020. As of August 19, 2026 PLATO’s own news index lists Coast Products as the first approved manufacturer — approved December 22, 2025 — with SATCO/NUVO joining as a trial member February 1, 2026 and NEBO joining August 13, 2026. None of the 25 models in the table below come from a certified manufacturer. Every “ANSI FL1” claim in this article is a manufacturer’s own assertion about its own product.

What is the difference between lumens, candela and lux?

Three different quantities that answer three different questions.

The candela is an SI base unit. Per NIST, following the 2019 redefinition: “The candela is defined by taking the fixed numerical value of the luminous efficacy of monochromatic radiation of frequency 540 times 10 to the twelfth Hz, Kcd, to be 683 when expressed in the unit lm per W.” That frequency is about 555 nanometres, the peak of human photopic sensitivity.

The relationships you actually need:

Candela is obtained by taking a lux reading at a known distance and multiplying by the square of that distance. A lux reading taken at exactly 1 metre is the candela figure.

Why more lumens can mean less distance

Lumens are the total photon budget. Candela is what happens after the optic decides where to put them. A deep smooth reflector paired with a small, high-luminance emitter concentrates flux into a tight cone: high candela, long throw. A large multi-die array under a shallow optic spreads the same or greater flux across a wide cone: enormous lumens, low candela, short throw. Because distance follows the square root of candela, it tracks concentration and never total output.

The cleanest demonstration is a single flashlight contradicting itself, from one manufacturer’s own product page. Acebeam’s Terminator M1 carries both an LED and an LEP emitter:

One-fifth the lumens, seven times the distance. Both rows published by Acebeam on the same page, checked August 19, 2026.

Candela to beam distance, worked out

Distances calculated as 2 times the square root of the manufacturer’s published candela. Note the teaching point: 1,200 cd gives 69 m, and four times that (4,800 cd) gives 139 m — only double.

How far below the headline number does a flashlight actually settle?

This is the table the category does not publish. Every figure is the manufacturer’s own specification. The column that matters most is not any number — it is whether the brand publishes a sustained figure at all.

All values are manufacturers’ published specifications, retrieved from their own product pages and manuals on August 19, 2026. Download this table as CSV, including impact and ingress ratings and full runtime strings.

The gap, where it can be measured

Who discloses and who does not

Acebeam is alone in full disclosure. Its spec tables use a notation that answers the question in three seconds: the L35 2.0 reads 5000~1800lm with runtime 60s+1h15min. The tilde gives the step-down levels, the plus signs partition the runtime across them. The X75 extends it to three stages: 80000~12000~900lm, 30s+42min+8min. This notation should be the industry benchmark.

Partial disclosure: Streamlight’s Wedge publishes a genuinely useful “1,000 lumens, 35 second burst” — and is notably the only Streamlight in this set with any turbo disclosure at all. Olight publishes turbo durations but not the levels it steps down to.

Prose only: Wuben’s 3,000-lumen sustained figure appears in marketing copy as “1.7 hours of constant current at 3000 lumens,” never in a spec table beside the 12,000-lumen headline.

Acknowledges step-down without quantifying it: Fenix. The E35R and PD36R Pro manuals both state that at 60 degrees C or above “the light will automatically step down a few lumens to reduce the temperature.” A few lumens, on a light rated at 3,100.

Silence: SureFire, Nitecore, Modlite, and Streamlight’s ProTac, Stinger and TLR lines.

Why is a long runtime figure not what it sounds like?

Because FL1 runtime runs from the initial output value until output falls to 10 percent of it, and the standard says nothing about what happens in between. A light can be advertised with a long runtime while spending the overwhelming majority of it at a small fraction of the advertised brightness, and the claim is fully compliant. The number is not false. It measures something other than what a buyer assumes it measures.

It compounds, because the initial value is the turbo value. On the Acebeam X75, 10 percent of 80,000 lumens is 8,000 — below its own 12,000-lumen sustained level. The runtime clock would never even reach the threshold during the sustained phase.

Fenix’s own footnote is the clearest statement of the problem in the industry

From the Fenix E35R manual’s technical parameters table, printed under the heading “ANSI/PLATO FL1”:

And the asterisk, verbatim from Fenix:

That is the manufacturer stating, in its own manual, that the 2 hours 30 minutes is not 2 hours 30 minutes at 3,100 lumens. It sits in a footnote, in a PDF, on page two.

Two ways to prove the footnote is doing real work, with a calculator

Route one, internal inconsistency, no battery data needed. Fenix’s own table claims Turbo delivers 3,100 lumens for 2.5 hours, which is 7,750 lumen-hours. High delivers 1,000 lumens for 3.667 hours, which is 3,667 lumen-hours. From the same cell. The Turbo claim implies 2.1 times more total light energy out of the same battery. That is impossible, and the only resolution is that most of those 2.5 hours are spent far below 3,100 lumens.

Route two, energy budget. Fenix publishes the E35R’s cell as a 21700 at 5,000 mAh, which at a nominal 3.6 volts is 18 watt-hours. At a generous 150 lumens per watt system efficacy, 3,100 lumens draws about 20.7 watts, giving 0.87 hours. At a realistic 120 lumens per watt, 0.70 hours. Fenix claims 2 hours 30 minutes — roughly three times the physical ceiling.

The PD36R Pro shows the identical pattern: 2,800 lumens for 3 h 30 min, only 25 minutes short of the 1,000-lumen mode’s runtime from the same cell. The product page then advertises “Max Lumens: 2800” beside “Max Runtime: 42 hours,” the second figure belonging to the 30-lumen Eco mode.

Can you audit a spec sheet without any equipment?

Yes, and this is the most useful thing on this page. Because FL1 beam distance is a pure function of candela, every manufacturer’s spec sheet can be checked against itself with a calculator. We ran the formula against all 30 published mode-rows in this survey where a brand gave both values.

28 of 30 match d = 2 times the square root of I to within 2 percent. That independently confirms both the formula and that these brands are applying FL1 arithmetic honestly.

The Fenix case is the notable one. Every other mode on that same E35R table — High, Med, Low, Eco — matches the formula to within 1.4 percent. Only Turbo is out, and only in the flattering direction. To justify 260 m the light would need 16,900 cd, not the 15,103 cd Fenix publishes. Nitecore errs the other way: its own candela figure would support a longer claim than it makes.

We report both directions deliberately, because reporting only the flattering error would be cherry-picking. Beyond these two internal inconsistencies we make no claim that any brand’s specification is wrong — that would require measurement, which we do not do.

What do IPX7, IPX8 and impact ratings actually mean?

Water ingress ratings come from IEC 60529, whose current designation is Edition 2.2, published August 29, 2013 as IEC 60529:1989 plus Amendment 1 (1999) plus Amendment 2 (2013). Almost every flashlight page writes “IEC 60529” with no edition, or worse, “the IP68 standard.”

Impact resistance

The FL1 impact rating is the height in metres from which the light, batteries installed and switched off, can be dropped onto concrete without cracking or breaking and still function. For ratings above 1 metre, each sample is dropped six times with a different face toward the ground.

These are manufacturer-declared ratings under a self-declared standard, not independently certified results, so they are not a clean durability ranking. Streamlight rating most of its line at 2 m while Fenix rates both examined models at 1 m is a difference in what each company chose to claim and test, not a measured comparison.

What are the rules for flying with a flashlight?

The regulation is 49 CFR 175.10(a)(18), with FAA guidance at PackSafe, last updated August 11, 2026.

• Spare batteries must be in carry-on baggage only. The regulation is explicit: “Spare lithium batteries must be carried in carry-on baggage only.” A light with the cell installed may travel either way.
• Every spare must be individually protected against short circuit — retail packaging, taped terminals, a battery case or a sleeve.
• The watt-hour limit almost never binds on an EDC light. The limit is 100 Wh for lithium ion. A 21700 at 5,000 mAh and 3.6 V is 18 Wh; an 18650 at 3,500 mAh is 12.6 Wh. Even the Acebeam X75’s built-in 14.4 V, 4,250 mAh pack is 61.2 Wh — still under, though this is the class of light where it becomes worth checking.
• If your carry-on gets gate-checked, pull the batteries. FAA is explicit that all spare lithium batteries and power banks must be removed from a bag that is planeside- or gate-checked.

Loose cells: what the regulator actually said

The U.S. Consumer Product Safety Commission issued a consumer safety warning on January 8, 2021. Its wording is broader than the version usually repeated:

Two things to read precisely. CPSC names vaping devices, personal fans, headlamps and some toys — it does not name flashlights. And its recommendation is against loose 18650 cells as a category, on the grounds that bare cells lack protection circuits, not merely against counterfeits. That sits in real tension with mainstream flashlight practice, where Fenix, Acebeam, Nitecore and Olight all sell protected cells for exactly this use. We report the tension rather than softening the quote into “avoid fakes.” Fenix’s own PD36R Pro manual takes the manufacturer side of it: “Use quality protected batteries to minimize combustion or explosion risks,” naming three approved cells and warning that “other Li-ion batteries may not function normally.”

On cell safety standards: IEC 62133-2:2017, Edition 1.0, published February 7, 2017 and still active, is the lithium-systems half of the standard that was split from the combined IEC 62133:2012. Worth knowing that it governs cells and battery packs, not flashlights. A flashlight is not “IEC 62133 certified”; its cell may be. A citation to bare “IEC 62133” with no part number has been ambiguous since 2017.

How this guide was made

Popular EDC does not run laboratory tests, and this page does not pretend otherwise. Every figure here is one of three things: a manufacturer’s own published specification with the source page and the date we checked it, a standards-body or regulator document with its designation and publication date, or arithmetic applied to those published figures. The candela-to-distance audit, the turbo-gap percentages, the lumen-hour inconsistency and the energy-budget check are all pure arithmetic on numbers the manufacturers published themselves — which is precisely why they are checkable by you, with a calculator, without owning any of these lights.

What we could not verify, stated openly. The full FL 1-2025 clause text is paywalled at $575, so the exact Short-Duration Elevated Brightness test protocol — its time thresholds and measurement intervals — is not stated here, and we will not infer it. Whether the 2025 revision makes sustained-output reporting mandatory or optional rests on a single secondary source and is therefore left open. The detailed IEC 60529 test apparatus conditions are behind a CHF 515 paywall and are not reproduced from third-party lab blogs. The FL1 drop protocol for ratings at or below 1 metre, the number of samples tested, and the concrete specification are in sections not present in the free NEMA PDF. Prices change; each is stamped with the date we checked it.