Some links on our site may earn us a commission at no extra cost to you. Learn more.
A sunrise alarm clock is a convenience product, not a medical device. If you are dealing with insomnia, seasonal low mood, or another sleep or mood condition, talk to a healthcare professional about what is right for you rather than to a lamp.
Updated August 2026: the Philips SmartSleep HF3650 has come off Philips’ own US light-therapy listing page, so it finishes last on the tally. This comparison scores light output, the distance it’s measured at, and what each maker puts in writing about it; the subscription-cost math for Hatch+ now lives on our dedicated Hatch+ page. Prices and warranty terms were re-read at each maker’s own site this month.
The Lumie Bodyclock Shine 300 wins this on a weighted tally, 82.1 out of 100, and the Philips SmartSleep HF3650 loses it at 49.3. But the number worth carrying away is not either of those. It is this: the JALL Wake Up Light, at $34.99 on JALL’s own store, puts 545 lux on a meter twelve inches away. The Hatch Restore 3, at $169.99, puts 285 lux on the same meter at eighteen inches. Six inches of nightstand is worth more than $135 of clock.
Four Sunrise Alarm Clocks, Four Product Pages, One Lux Figure
A sunrise clock has exactly one job that a $12 clock radio cannot do: put light on your closed eyelids before the sound starts. So the first thing to check is what each maker is willing to say about that light. I read all four product pages and the support and warranty documents behind them.
| Maker | What it publishes about brightness | What an independent meter recorded | Verdict |
|---|---|---|---|
| Hatch (Restore 3) | No lux, lumen or nit figure on hatch.co. Its own support article on brightness says only to move a slider to your preferred setting. | 610 lux at 12 in, 285 lux at 18 in | No claim to check |
| Philips (SmartSleep HF3650) | Nothing on the US product page. Philips’ own manual for the HF3520, a different wake-up light, prints ‘Approx. 1 lux – 300 lux at 45 cm’. | 650 lux at 12 in, 325 lux at 18 in | The only conditioned figure in the field — and it is on another model |
| Lumie (Bodyclock Shine 300) | No lux or lumen figure on either the US or the UK page. Both publish the sunrise duration range instead. | 730 lux at 12 in, 360 lux at 18 in | No claim to check |
| JALL (Wake Up Light) | No lux or lumen figure on jall-us.com. The listing counts colours and nightlights. | 545 lux at 12 in, 275 lux at 18 in | No claim to check |
Every measured figure in that right-hand column comes from one place: Optimize Your Biology’s light-meter testing of 15 sunrise clocks, whose author describes the method as putting each clock in front of a spectrometer and measuring lux every six inches across brightness settings. That is a single independent tester carrying the entire evidentiary weight of a product category. More on what that limits at the bottom.
Three of four makers publish nothing. That is not an oversight and it is not shyness about a bad number. Brightness without a distance is not a specification, and a maker that prints one has to pick a distance and defend it.
Six Inches of Nightstand Is Worth Roughly Twice the Light
The same test publishes readings at two distances I could read consistently across the whole field: one foot and eighteen inches. That gives you something the marketing does not. Divide the eighteen-inch reading by the twelve-inch reading and you get how much of the light survives half a foot of extra nightstand.
If these clocks behaved like a bare bulb, the answer would be fixed. Illuminance from a point source falls with the square of distance, so going from 12 to 18 inches should leave (12 ÷ 18)² = 44.4 percent. Here is what the meter actually recorded.
| Clock | Lux at 12 in | Lux at 18 in | Light retained | Point source predicts | Gain from moving 18 in to 12 in |
|---|---|---|---|---|---|
| Lumie Bodyclock Shine 300 | 730 | 360 | 49.3% | 44.4% | 2.03× |
| Philips Hue Twilight (fifth clock, same test) | 840 | 395 | 47.0% | 44.4% | 2.13× |
| Philips SmartSleep HF3650 | 650 | 325 | 50.0% | 44.4% | 2.00× |
| Hatch Restore 3 | 610 | 285 | 46.7% | 44.4% | 2.14× |
| JALL Wake Up Light | 545 | 275 | 50.5% | 44.4% | 1.98× |
All five hold on to more light than a point source would — between 46.7 and 50.5 percent against a predicted 44.4. That is expected physics, not a measurement error. A 2025 paper by Cecilia Stari, Marcos Abreu, Martín Monteiro and Arturo C. Martí, Light illuminance does not always decay with the inverse of the square of the distance (arXiv 2501.00622), sets out the cases: a flat source keeps illuminance nearly constant while you are closer to it than the source is wide, and only at distance does it start behaving like a point. A sunrise clock is a broad diffusing dome rather than a filament, and it is read from one to three feet away. That is the region where the simple rule stops applying.
The practical version fits in a sentence. Move any of these clocks from eighteen inches to twelve and you get about twice the light — 1.98× on the JALL, 2.14× on the Hatch. That swamps every difference between the four. The JALL at twelve inches delivers 545 lux against the Hatch’s 285 at eighteen: 1.91 times the light for 4.86 times less money.
What the test does not publish, and what I am not going to invent, is where each clock falls below a useful level. Optimize Your Biology names roughly 250 lux at the pillow as a working floor and reads out to 36 inches on per-model charts; the two distances I could pull cleanly for all five are the two above. If your clock sits across the room on a dresser, none of these numbers describes your morning.
300 Lux at 45 Centimetres, Printed on a Model Philips Sells Instead
One document in this whole category does the job properly. Philips’ own user manual for the HF3520 wake-up light, dated 2022, states the light output as approximately 1 lux to 300 lux at 45 cm. A number and the distance it was taken at. That is the format a buyer can use.
Forty-five centimetres is 17.7 inches, near enough to the eighteen inches the independent meter used that the two are comparable. Philips claims up to 300 lux on the HF3520; the meter recorded 325 lux on the HF3650 at 18 inches. Different models, so this is not a claim checked against itself — but the one maker that publishes a conditioned figure is also the one whose products land where an outside meter says they land. That is worth something.
It also makes the rest of the field look worse rather than better. Philips proved the number can be printed. Three makers chose not to print it.
$34.99 to $169.99, and What Each Price Buys
- Lumie Bodyclock Shine 300, $169.00: the brightest reading in the test, the longest sunrise ramp, and the longest warranty. On pre-order at Lumie as of this update.
- JALL Wake Up Light, $34.99: within 10 lux of the Hatch at the same distance, for a fifth of the money. The warranty is a claim you cannot read.
- Hatch Restore 3, $169.99: the app, the sound library and the routines. The shortest warranty here and the clock’s own light is third of four.
- Philips SmartSleep HF3650: the loser. Second-brightest, shortest sunrise ramp, no price on Philips’ own site, and gone from Philips’ US listing page.
Five Axes, and What Each One Is Worth
Here is the tally, with every weight stated so you can move one and see what happens.
- Light at 18 inches — 30%. The heaviest weight because it is the only thing a sunrise clock does that a cheap clock radio cannot. Scored as the measured lux divided by the highest in the field, 360.
- Longest settable sunrise — 20%. A slow waker needs a long ramp and cannot add one later. Scored as the longest ramp divided by 90 minutes, the longest available.
- Price on the maker’s own site — 15%. Light, deliberately: the spread here is real but the light difference is bigger. Scored as $34.99 divided by the price.
- Warranty as published — 20%. Months of coverage you can read before you pay, divided by 36, the longest verifiable term here. A warranty you cannot read scores zero.
- Listed on the maker’s own catalogue page — 15%. Not the product page, which can outlive the product. 100 if listed and shipping, 60 if listed on pre-order, 0 if the maker has dropped it.
| Clock | Light 30% | Ramp 20% | Price 15% | Warranty 20% | Still listed 15% | Total |
|---|---|---|---|---|---|---|
| Lumie Bodyclock Shine 300 | 100.0 | 100.0 | 20.7 | 100.0 | 60 | 82.1 |
| JALL Wake Up Light | 76.4 | 66.7 | 100.0 | 0 | 100 | 66.3 |
| Hatch Restore 3 | 79.2 | 66.7 | 20.6 | 33.3 | 100 | 61.9 |
| Philips SmartSleep HF3650 | 90.3 | 44.4 | 0 | 66.7 | 0 | 49.3 |
The Lumie’s 82.1 breaks down as (100.0 × 0.30) + (100.0 × 0.20) + (20.7 × 0.15) + (100.0 × 0.20) + (60 × 0.15) = 30.00 + 20.00 + 3.11 + 20.00 + 9.00. Run the other three the same way and the order does not change.
Two cells are zeros with reasons rather than measurements. The JALL scores zero on warranty because jall-us.com advertises a lifetime warranty behind an activation link and publishes no terms document I could retrieve; that is a mark against the disclosure, not a claim that JALL refuses service. Philips scores zero on price because usa.philips.com renders no price numeral for the HF3650 and I will not substitute a retailer’s. An empty cell with a printed reason beats a filled one resting on a guess.
The gap from first to second is 15.8 points, which is wider than any weight I would defend moving. No tie-breaker is needed here.
360 Lux and Three Years: the Lumie Bodyclock Shine 300
The Lumie takes four of five axes. It measured brightest of the field at both distances. Its sunrise runs 15 to 90 minutes, a range Lumie prints on its own page and the outside test confirms, and the 90-minute end is genuinely rare. Its warranty is three years, worded on the US page as a 3-year limited warranty and on the UK page as a 3-year warranty, with the same repair-or-replace remedy either way.
Two honest marks against it. Lumie’s US page listed the Shine 300 as pre-order with an expected shipping date of 25 September when I read it, which is why it lost 40 points on the availability axis rather than none. And the same independent test rates its flicker as risky at low brightness, which matters if you use it as a bedside reading lamp as well as an alarm.
Price is $169.00 on Lumie’s US store. The UK store lists the same model at £149.00. I am not converting one into the other — the two stores ship to different places and price independently — but it is worth knowing both figures exist before a reseller quotes you a third.
Skip this if you need the clock in the house this month, or you want it doing double duty as a reading light.
$34.99 Buys 275 Lux: the JALL Wake Up Light
Second place, and the reason is arithmetic rather than affection. At 275 lux at eighteen inches the JALL sits 10 lux behind the Hatch Restore 3 and 85 behind the Lumie, on a meter that cannot tell what anything cost. Its sunrise runs 10 to 60 minutes. It is listed and shipping on jall-us.com at $34.99.
The warranty is the problem, and it is a disclosure problem rather than a coverage one. The site carries a banner inviting you to activate a lifetime warranty. I could not retrieve the terms behind it, so I could not score a duration, and a term you cannot read before buying is worth nothing at the point you are deciding. Adjust that cell upward if you are willing to take the banner at face value. Score the JALL 100 on warranty and it goes to 86.3 and takes first place from the Lumie by 4.2 points. That is the single weight in this tally most worth arguing about.
The build is what $34.99 buys. Plastic, an interface that takes a week, and a brand whose US presence is a Shopify store and an Amazon shelf. Nothing here is precious.
Skip this if you want a maker who will still be answering the phone in five years, or if the clock is a gift and the box matters.
One Year From Registration: the Hatch Restore 3
Third at 61.9, and it loses on paperwork more than on light. Hatch’s own warranty page covers the Restore for one year from the date of registration, not the date of purchase — so the clock a buyer forgets to register is a clock with no term running at all. That is the shortest coverage in this field, attached to the joint-highest price at $169.99.
What it does well is not in dispute. The sound library is the best here, the app builds routines that people actually keep, and the free tier is real: Hatch publishes it as 40 sleep sounds, 20 alarm sounds, 20 lights and five unwind options, with Hatch+ at $4.99 a month or $49.99 a year adding roughly 100 of each. If the subscription is the thing you are weighing, we run that math separately in our comparison of the Hatch+ membership against a $50 sound machine, which puts the Restore 3 against the Yogasleep Dohm and the LectroFan EVO on running cost. This page is about the light.
On the light it is third of four, and 285 lux at eighteen inches is 79 percent of what the Lumie delivers at the same distance for a dollar less.
Skip this if the light is the reason you are buying, or if you will not register a product within a year of unboxing it.
Zero Listings on Philips’ Own Catalogue Page: the SmartSleep HF3650
This is the named loser despite testing second-brightest of the four at 325 lux, which beats the Hatch and the JALL comfortably on light alone.
Then I read the catalogue page rather than the product page. Philips’ US light-therapy listing carries eight products — the EnergyLight HF3318 at $299.99, two goLITE BLU models at $199.99, the briteLITE 6, and four Wake-up Lights including the HF3471 at $129.99 and the Wake-up Light Plus HF3485 at $169.99. The SmartSleep HF3650 is not among them. Its own product page still resolves, still renders the specification list, and still carries a 2-year warranty line, but it renders no price and Philips no longer puts it in the menu.
A product URL that still loads is not evidence a maker still sells the thing. Philips has not announced anything about the HF3650 that I could find; it simply stopped listing it, which is the quieter of the two ways a product leaves. If you want a current Philips, the HF3485 at $169.99 is the one the company is actually selling — and there is no independent lux figure for it, so you would be buying it on faith.
Skip this if you want a warranty you can claim on, which is most people. Buying a delisted model means betting the parts and the support outlast the listing.
One Recall on Record, and It Is Not This Clock
Hatch has a recall in the CPSC record and it deserves stating plainly, because searching the brand will surface it. On 3 July 2024 the agency published recall 24-291: about 919,400 US power adapters sold with Hatch Rest 1st-generation sound machines, plus about 44,352 in Canada, because the plastic housing could come off in the outlet and leave the prongs exposed. The notice records 19 reports of the housing detaching and two reports of a minor shock.
The adapter is model CYAP05 050100U and it shipped with the Rest 1st generation, a baby sound machine. The Restore 3 is not in the recall, and neither is the Rest 2nd generation. I found no CPSC recall touching Philips, Lumie or JALL sunrise clocks.
One Tester, One Unnamed Meter, and No Published Protocol
Every lux figure on this page traces to a single source. Optimize Your Biology bought 15 clocks and metered them; that is more than any of the makers did in public, and it is still one person. The instrument is described as a lab-grade spectrometer and is not named. No protocol is published for how the clocks were aimed, what brightness setting each reading used, or how ambient light was excluded. I could not find a second independent light-meter test of this category to check it against.
So the ranking on light is as good as that one test and no better. It does not compare products this operation has handled — we buy nothing and meter nothing, and every number above is somebody else’s reading or somebody’s own published document, named in the sentence where it appears. Our sourcing rules are on the editorial standards page.
What does not depend on that test is the shape of the finding. The distance arithmetic holds whatever the absolute numbers are, because it is a ratio of two readings from the same meter on the same clock. And the warranty terms, the prices and the catalogue listings came straight from the makers.
Prices are approximate, current as of publication, and change frequently — confirm the latest price with the retailer before buying.
Two Decisions, in Order
Decide where the clock sits before you decide which clock it is. Twelve inches from your face rather than eighteen is worth about double the light on every unit measured, which is more than the entire spread between the cheapest and the brightest.
Then buy on the tally. The Lumie Bodyclock Shine 300 if you can wait for stock. The JALL if you want to find out whether light wakes you before spending $169 on the question. The Hatch Restore 3 if the sounds and the app are the point and you will register it. Not the Philips SmartSleep HF3650, which Philips has stopped listing.
If the clock is one part of a bedroom you are rebuilding, our white noise machine comparison covers the sound side and our blackout curtains, shades and blinds comparison covers the dark half. The rest of the category sits in Home & Living.
