View Full Version : Headlight Shootout... Blind Opinion test.

10-16-2008, 09:09 AM
Was poking around Mtbr.com's forums (http://forums.mtbr.com/) when I found this in the bike lights forum:

Light Shootout - blind opinion test (http://forums.mtbr.com/showthread.php?t=459490)

I looked at the pictures, and I thought that it was a pretty cool idea. I myself wasn't interested in trying to figure out which light was which, but it gave me an idea.

Wouldn't it be cool if we could come up with our own version to help others in making purchasing choices. Perhaps, do it regularly (once a year or so) before the bike light season starts so new riders can see what the latest and greatest lights are, their cost, and make informed decisions.

Pick a night, and a location, and anybody who was interested could come out and see how their light (bike or flashlight) would compare to other lights. Now, I'm not talking about a "Best Lit Bike" Contest (which might be fun too), but a best bang for the buck comparison. People just show up with their bike lights (and fresh batteries), and we take pictures. Perhaps, from three views, the rider's, oncoming traffic, and overhead (beam pattern).

We could create a standard course (say on a football/soccer field/or track), that anybody could replicate anywhere else (from Anchorage to Zimbabwe) if they have the requisite objects (bowling pins, garden gnomes, hurdles, signs, etc).

http://www.planetbike.com/imgs/3007_1view.jpg http://www.planetbike.com/imgs/3060view.jpg (http://www.planetbike.com/page/learn/lightfinder)
Something similar to these shots from PlanetBike

Major problems I see with this is... I don't know where a good (and dark) soccer/track field is in PDX (I doubt attendance would be good in Washington County). We'd need a good night for doing it (clear, no moon). And the biggest problem I see, I don't know how to "fix" my camera to take pictures with a standard aperture/shutter time.

Just an Idea...


Bent Bloke
10-16-2008, 10:01 AM
That cave looks cool, literally!

I just added a small light to my helmet, but it's only a single led and not very bright. I use a Planet Bike Beemer 3 in blinky mode for my bike, but wanted something to illuminate the road for my helmet. Something not too heavy, and less than $50 (or not much more than that), that will easily attach to a helmet. Something with 3-5 leds should be sufficient.

Someone mentioned looking for outlets that sell to the tactical market, but not having to do any "special ops" for, oh -- ever, I'm not sure where to go to buy tactical gear.

10-16-2008, 10:11 AM
An informal 'best bang for the buck' light convention like that might help focus attention on what kinds of improvements are needed in the way of bike light beam characteristics. While overall brightness is important, so also is the illumination field of a lights lamp and reflector array. Any effort made to find and/or raise awareness of lights and user techniques that adequately light the road yet avoid blinding oncoming riders in places such as MUP's would be helpful.

Some time back, Wyeast made some very useful posts featuring demonstrations he'd made of 4-5 different flashlights for use as bike lights; their comparative brightness, beam pattern and cost.

Flashlights as budget headlights (http://bikeportland.org/forum/showthread.php?t=2149)

The tests Wyeast did with flashlights are great work and could be very helpful to people trying to work out their lighting needs. :cool:

On a different thread, also supporting the idea of using flashlights as an alternative to bike specific lights, Vincent Paul had the following to say:

The Niterider X.2 actually puts out only 150 lumens on high, and the 300 lumens quoted for a dual x.2 system is deceiving. These 300 lumen flashlights will light something up at a greater distance than a dual x.2 system. One of the intended uses of these new flashlights is as tactical para-military gear - they'll temporarily blind someone from a considerable distance, even in daylight.

The new Cree and equivalent LED emitter flashlights are considerably more powerful than any comparable bike light of equivalent weight and size for a fraction of the cost. Bike lights are always going to be much more expenseve than flashlights of equivalent power. It's simple mass production principles. The tactical flashlight market is huge. Military, police agencies, security, hunters, hikers, timber industry, repair, etc. I hesitate to give a guess at the market discrepancy between high-power flashlights and high-power bike lights. A factor of ten? Twenty? One hundred? The bike market is always going to be a niche market behind the curve on lighting technologies.

This situation (bike light market/niche market) could change as the realities of relying on bike light illumination for night-time riding becomes better and more widely understood.

10-16-2008, 11:14 AM
After thinking about this some more, I got the idea to have a time lapse video made of each light to see what things look like over the run time.

I'd predict that the first to fail will be the commercial battery halogen lights (using AA, AAA, or C rechargeable batteries, then alkaline). Next would be the button battery powered LED's (ie Trimet's blinkies). Then it would be the Halogen or HID lights, followed by the powerfull LED lights, and last would be the smaller commercial battery LED lights.

Damn, now I'm thinking that this belongs in the Consumer Reports magazine... Add battery brands (Eveready, Energizer, Etc) into the fray and we could really learn something.


10-16-2008, 12:29 PM
Why limit it to battery-operated lights? :)

Got the B&M IQ Fly (http://www.bumm.de/index-e.html), now lusting after the Supernova E3 (http://ncrandonneur.blogspot.com/2008/05/supernova-e3-glowing-review.html)...

10-16-2008, 12:49 PM
Why limit it to battery-operated lights? :)

Got the B&M IQ Fly (http://www.bumm.de/index-e.html), now lusting after the Supernova E3 (http://ncrandonneur.blogspot.com/2008/05/supernova-e3-glowing-review.html)...

You could certainly bring out your dynamo powered lights... Problem I see is how to do stationary pictures with them.

You'll have to forgive me, I've moved so far away from them following my problems with them while I was in the UK, dynamo's didn't even cross my mind.

I just found that Busch & Müller also produce gas powered lights... I didn't know they existed...

http://www.bumm.de/docu/grafiken/tunnelbilder.gif (http://www.bumm.de/index-e.html)
various beam patterns from the Busch & Müller lights.

Basicly, I'm saying is if you have something that produces light, bring it...

As for batteries (or gas), I'd be interested to know that Brand X runs for this amount of time vs. Brand Y, and they produce insert amount of light here lumens (lux/candlepower/whatever). If I can get away with buying a Dollar Tree brand of batteries and get similar results, I'd be interested in saving the money, if Brand X is far superior, then I would go that way. I don't think that comparisons that use mechanical devices (say toys) rather than lights, produce acurate results.

Rubberside Down!

10-16-2008, 03:27 PM
Why limit it to battery-operated lights? :)

Got the B&M IQ Fly (http://www.bumm.de/index-e.html), now lusting after the Supernova E3 (http://ncrandonneur.blogspot.com/2008/05/supernova-e3-glowing-review.html)...

I can see how the Supernova E3 might generate a bit of lust as well as excellent illumination; it's a great looking, and by the sound of it, performing light. Nice if they'd send one to Wyeast for free so he could do a test for us ;). Having virtually no personal experience with it, one thing that reflexively puts me off dynamo lights is the tire power take off.

Of course Supernova offers what looks to be excellent dynamo hub wheels; both road and mountain bike wheels. $400 is a lot of money for poor schmo's like me. The lights at $240 are expensive too. Those prices would probably make the lights out of the question for many people.

10-16-2008, 09:36 PM
I grant the entry price is expensive. I expect my requirements are perhaps a bit more stringent than most cyclists. Ride all night? (http://lynnerides.blogspot.com/2008/05/fleche-ouragan.html) You betcha!

The hub is the generator; it isn't a rim generator. It does slow me down 1-2 mph when the light is on. I've got a SON dynamo hub, which, as near as I can figure, will not need maintenance for a Very Long Time. Ditto for my light.

I bet I haven't spent more than K'tesh on lighting :)

And I surely appreciated the light heading into Newberg on 99W this past Sunday night!

I can see how the Supernova E3 might generate a bit of lust as well as excellent illumination; it's a great looking, and by the sound of it, performing light. Nice if they'd send one to Wyeast for free so he could do a test for us ;). Having virtually no personal experience with it, one thing that reflexively puts me off dynamo lights is the tire power take off.

Of course Supernova offers what looks to be excellent dynamo hub wheels; both road and mountain bike wheels. $400 is a lot of money for poor schmo's like me. The lights at $240 are expensive too. Those prices would probably make the lights out of the question for many people.

10-16-2008, 09:38 PM
Oh, and my light has a standlight. If you get the picture quick enough after I stop, you'll get most of the lighting effect. It lasts about 5-7 minutes, but decreases the longer I'm stopped.

10-16-2008, 10:58 PM
lynnef, thanks for the additional info about your light system. Sounds very good.

A simple search brought up a bunch of sites offering info about SON dynamo hubs. The following bike shop owner's site has what seems to be an especially thorough rundown on the subject:

Bicycle lights and generators Peter White Cycles Hillsborough NH (http://rds.yahoo.com/_ylt=A0oGkjLrK_hIqHkBBSBXNyoA;_ylu=X3oDMTEyZWRpbHA wBHNlYwNzcgRwb3MDMQRjb2xvA3NrMQR2dGlkA0gxODJfNzk-/SIG=11vdna1sh/EXP=1224310123/**http%3a//www.peterwhitecycles.com/schmidt.asp)

A couple sentences from that page that offered something very interesting I wasn't aware of: "The Schmidt dynohub meets the German government's StVZO regulations for bicycle lighting. Among other things, these regulations set the minimum speed at which a dynamo lighting system must provide adequate lighting." Peter White

I think people in Oregon would benefit from some changes to the state's required lighting specs for bikes.

StVZO would be German Road Vehicle Admission Regulation (StVZO). That just seems to make so much more sense that what we have: Department of Motor Vehicles (DMV)

Not that I've spent much time looking yet, but haven't run across the phrase 'standlight', but it's pretty clear what you mean by that. I thought of this when writing the earlier post but forgot to include a comment to that affect. Definitely sounds important to have something like that, so I'd count this feature as a plus.

10-17-2008, 12:20 PM
I went right to the top in features for my light. It has that standlight, which is basically a capacitor that charges as I ride and keeps the light illuminated for a few minutes after I stop. It also has a sensor mode - it detects when it thinks it is dark enough and turns the light on. The fog this morning did not trigger it though, so I turned it on.

The lighting pattern is approved by StVZO. The Supernova's light pattern isn't, but I'm going for, well, bright. I can always aim it down.

The light kicks on when I start walking with the bike.

Shimano also makes generator hubs, rather less expensive than the SON.

Fortunately, I bought and had my lighting setup built before the dollar tanked.

I bought the hub and rim, and had my LBS build the wheel.

10-17-2008, 07:45 PM
K'Tesh, you should've called me to tell me about this thread. *crosses arms* ;)

This'll feel a little bit of shotgunning information, but it's all pouring out at once and I'm not sure how to organize it.

Shootout Test - Is an excellent idea. However, you're faced with either a) The logistics of getting all the lights together at one place at one time to shoot them all under identical conditions, or b) Finding an (indoor) control environment where you can repeat the same test conditions. Ideally, this is probably something like a warehouse or gymnasium with no windows.

In the interim, I think the best compromise is to try to provide a "control" light for comparison with each group of shots. In that way, people can form a relationship in their mind with something familiar. What that light would be, I have no idea. I figured a 5-blinky like a Beamer is probably the most ubiquitous among cyclists, but it's on the bottom end of the illumination scale. That's the same reason that I try to make comparisons with rough equivalence among bike lights (like Newts, etc) to give people an idea of what to compare against.

Runtimes You're close in your estimate of runtimes. The only one I would change the order on is the TriMet Blinkies. Yes, the button cells have very little capacity compared to larger lights, but the little 5mm LED's are also very miserly on power. Blinkies tend to last very long on the same batteries vs a high flux emitter like a Cree - the difference of 0.5watts of power consumed (for a 5-LED) vs 3-4 watts for a Cree at full power. Then again, you're also getting several times more light out of the Cree.

It's only the last few years that LED's have become competitive (and now even beating) incandescent lights for brightness vs consumption. Well, let me rephrase that. They've always been more efficient, but it's only in the past year or two that have we gotten LED emitters that compete with 10-20-30w halogen lights in terms of power.

There are a handful of quality runtime tests of some bike-specific lights, but not many of them. For handheld flashlights-used-as-headlights there's a pretty good stock of information at a variety of flashlight-dedicated sites. ;) If anyone is interested, I can give you those URL's. Otherwise, I won't detract too much from what we have here. As a very rough rule of thumb, you're looking at around 45-min to 3-4 hours for a high flux LED (LuxeonIII, Seoul, or Cree) in a compact package (1AA to 4AA) In general, those three emitters are usually consuming the same amount of power - it's the efficiencies (light vs current) that is why Crees are brighter than LuxIII lights.

From the flashlight-heads, a good rule of thumb is to consider battery life is what takes you to 50% of full-charge brightness. Some lights will be advertised as having hundreds of hours of life, but most of that may be in a very dim "tail" at the bottom end of the curve, which is really not useful to us in this application.

Batteries: How to get more life? - If you are using a high flux light (or incandescent) with a common off-the-shelf battery size (AA, C, D, etc) there are a few things you can do to maximize your brightness over time. In general, Lithium batteries will hold their voltage better over the life of the battery. So for most lights, it will maintain better brightness until the batteries finally give out. To a lesser extent, rechargable NiMH batteries will behave the same way. Alkalines tend to have a more linear discharge, they gradually lose voltage the whole way down - so when you hit that 50%, you're literally at half life on the battery, compared to Lithiums and NiMH's which won't drop to 50% voltage until they're closer to actual 75-90% of capacity used up.

Of course price plays a factor into that. If you don't need to squeeze every minute out of the battery, alkalines are far cheaper than lithiums.

As for K'Tesh's question about brand x vs brand y battery, that's also heavily researched among the flashlight heads. :D I can't remember the numbers off the top of my head, but I do recall that the Kirkland batteries at Costco fared quite competitively with the big names like Duracell and Energizer (slightly less performance at much less cost) so that's probably your best bang for the buck. More specific than that, and it kinda depends on the light as some batteries do better than others based on how much power that specific light tries to draw.

On that note, I've said this before, and I'll say it again. Unless you go on very short rides, avoid 3-AAA type lights if you can. They're very popular because they don't cost as much as other types of lights - this is because they don't require a voltage boost circuit. The downside is this very lack of regulation plus the naturally low capacity of AAA cells to begin with, and you have a light that will typically get noticably dimmer in a very short amount of time, 15 minutes even. You don't notice this in most cases where you're turning on the light for a few minutes then turning it off (like camping) - but on a bike where the light is constantly on, you find the batteries give you very poor bang for the buck.

Helmet Light For Bent Bloke - one popular method of helmet light is to use a bikeblock or just velcro straps to mount a small 2AA light - this will be the most bang for the buck, but a little tricky to mount on certain helmets. Another popular method is to modify a camping-style headlamp (like a Myo XP or Princeton Tech EOS) to fix to your helmet. Depending on how fast you ride, 3-5 LED's (imho) is not sufficient to "light the road" for you at any decent amount of speed.

What makes a good light? This is a very subjective question, and cannot be answered without looking at what you want out of your lights. Brightness and runtime alone are not enough to define what makes a good light. Beam pattern plays a big role. A demonstration of this is the Task Force 2C vs Coleman MAX in the thread wsbob linked. Those two lights have nearly the same lumen count (brightness), but the size and shape of beams are very different.

There's also color. Many LED's tend to have harsher blue tones or even some green tones. Some people find this hard on the eyes and/or makes it difficult to make out details on the road. Often people say this tends to "flatten" their perspective.

Size and shape of the light itself can come into play. Some lights are heavy, bulky, and may not mount well onto your bike.

Reliability is an underestimated factor (again, imho). Often people will steer towards some lumen monster for $10. Then again, I've seen $30-50 lights that were built pretty cheaply as well.

I think that's enough of my jibba jabba for now. I'm sure more will follow later. :D

10-17-2008, 11:25 PM
I studied the Bush and Muller website again tonight Bush and Muller bike lights (http://www.bumm.de/index-e.html). Amazing. Germany's market for bike lighting seems to be light years ahead of that here in the U.S. B & M has just an amazing array of precision designed headlights.

Their light called the 'Big Bang' : "As bright as a car headlamp For cross-country tracks and road traffic" . The demo pics on that light are something to see. The one showing '(with light-dark-line)' was interesting to me. The headlights in my dad's car have that effect. It seems like this design could correct the problem of riders on MUP's being temporarily blinded when riders with bright lights approach them from the opposite direction.

B&M has tail lights too (or as they call them rearlights) Even have one with a brake light feature. Wonder how well that works. I think they have a headlight with a hi beam-low beam feature. The text for that light was in German. Since I don't speak/read it, I couldn't be sure.

10-18-2008, 06:32 AM
European standards for headlights in general are pretty strict compared to the US with regards to stray light. So it doesn't surprise me that this translates to their bike lights using optics (or in some cases, a variation on "reflex" type reflectors) to make the most of the lights, putting it on the road instead of up into the trees and without a glaring hotspot in the center.

We've had HID lights available for bikes here stateside for a few years, I think. Niterider Rage, Topeak Moonshine, for starters.

The latest "triple" Cree light setups like the TriNewt or Light & Motion's Seca are the first LED-based bike lights that are competitive with HID's for overall brightness. I don't have any experience with either to be able to discuss what their pattern looks like or if there's any sharp cutoff (K'Tesh, I'm looking at you, man :D). Being reflector-based, my guess is the cutoff won't be very sharp - just the nature of the emitter.

10-18-2008, 09:03 AM
For those of us finding themselves lacking in familiarity with some of the lighting technology terms and concepts being thrown out in some of these comments, good ol' wikipedia has a nice article to help us out:

discussion of light source types on Wikipedia: Headlights (http://en.wikipedia.org/wiki/Headlights#HID_.28xenon.29_light_sources)

Simple Nature
10-18-2008, 10:50 AM
To add to what Wyeast so elequently disclosed, you will also see the use of the higher density batteries. Most Cree flashlights use CR123A or 18650 type Litium batteries. These are anywhere from 3 to 3.7v each. In the case of the CR123a, they have 3 times the energy of the standard NiMH AAA cell in about the same volume.

Also, one of the earlier discussions mentioned *strobing* the LED to conserve energy. while watching my chain with the Cree helmet light last night, it is quite obvious that Cree modules do this strobing. I suspect is it more than energy concervation, though... it is also a great way to manage heat.

Indeed, when you pull that much juice out of a battery, useful battery life is greatly diminished. I decided to see just how long 2x cr123a's [3.7v @ 1000mah] would last at 300 and 200 lumens. On the 3 hour ride, I used the light about half the time and the light was pretty much down to 33% or full output by the time I got home. A bit disappointing and pretty much tells me to carry the extra set of batteries.

Having said that, the 18650 batteries are essentially double the cr123a's They are mostly rated at 3.7v at 2500mah. With an extended body tube, a pair of 18650's in the flashlight would last better than twice as long.

BTW, my main headlight is boom mounted [mounted to the frame]. It is a Nashbar special 5 led. Daylite use is in flash mode... quite visible... and nighttime use: it will light your way and have side lenses for visibility. It is useless when a car is coming towards you, though.

10-18-2008, 02:28 PM
Re: those German bike lights. They are ahead of us because all bicycles sold there are REQUIRED to have lights, and, I believe, they must be generator powered. I think most bikes come with a rim generator (the kind you all remember not very fondly; I'm told they've improved...).

I think the Big Bang retails for $900. Or, it did before the dollar tanked. Not planning on getting one :)

10-19-2008, 12:41 AM
If I were going to go about measuring lights, here's what I'd do. First, get permission of the local school to use their gym some evening or weekend. Get yourself a macbeth colorchecker card or two, and a photography light meter. You'll also need a camera (preferably digital) with a manual mode. Temperature will have a significant impact on test results, so arrange to do all your testing when the gym is the same temperature (which to verify, you'll need a thermometer too).

Set up your lights one by one on a stand at one corner of the gym. Mark points on a polar coordinate grid, at say 0, 7.5, 15, 30, 45, 60, 75, 90, and 95 degrees at distances from 0m to as far out as you can in the gym. At each point, take a picture of the color card from a fixed distance as illuminated by the light. Also at each point, take the median of a few readings on the light meter. You can compile these numbers into a plot showing quantitatively the illumination and lighting pattern of each light. The amount of shift to the picture that's necessary to make the gray panel gray will define the chromatic aberration of the light. Having quantitative data is important because you can't trust the manufacturer ratings.

If you were really dedicated, you'd record continuous running time before the light at 0 degrees fell to 90% of fresh-battery-initial value, 80%, 70%, ... down to 25% or so and repeat the above measurements at each battery-life milestone. For the HID lights that take a little time to warm up, I'd measure how long that takes probably as time from power-on until intensity reached 90% of maximum value. To measure generator hubs, you'd need a small motor to turn the generator at a fixed representative speed for the duration of the measurements. Maybe you want to repeat at a few different speeds.

Just for completeness, I'd also record the upward angle (vertical angle of inclination) at which the light (fresh batteries, 0 degrees) hit 90% brightness. There's no point in wasting light illuminating trees and clouds, but SUV drivers' eyes are probably a few degrees above straight horizontal from your bike handlebars or helmet.

I don't think I"d even try measuring reliability. No matter what kind of test or accelerating factors you design, someone will complain about how it doesn't align with real life. And the available sample size would be too small to be statistically significant anyway.

People can use that information to choose whatever light they prefer. For street riding, I'd want a bright light with a narrow beam, but for off-road riding, a wider beam is necessary. I don't care too much about chromatic aberration, but I do find the overly blue or overly green lights to be annoying. That's where the color chart comes in.

To see an example of how you might present this data, check out the lens reviews over at dpreview.com. They have some experience measuring light and optics. :)

10-17-2009, 11:59 PM
Just thought I'd bring to everyone's attention, the headlight that some people over at bikeforums have been excited about....

Has anyone tried the Magicshine 900 lumen?/bikeforums-electronics lighting and gadgets (http://www.bikeforums.net/showthread.php?t=576697)

Seems to be quite a light...cheap...about $90 or less with batteries and charger....and very bright. Available from several sources, but a bikeforum member has a small business selling bike lights, including this light, to which it seems his company has taken considerable effort to inspect and test before shipping, to ensure reliable performance...also talks with manufacturer reps and apparently has been able to get them to make improvements to the light's design.

Judging from the comments over at bikeforums, the light isn't perfect, but is regarded as very, very good...especially for the money. bikeforums has a thread going strong yet for the Taskforce flashlight, which sometimes Portland Bike Forums member tested last year and reported on in another post. It's cheaper...$30...if I remember correctly...but the Magicshine is much brighter.

Guess I'm wondering if any people in the Portland Metro area are trying this light out, and how you're finding it to be.

10-21-2009, 03:15 PM
Buyer, be wary of lumen ratings on these LED bike headlights, e.g., "... produces 700 retina-searing lumens," and other such crotch-grabbing nonsense. The headlight manufacturers are typically quoting the light output of the LED chip(s) itself, which is an instantaneous measure at 25 deg C (77 deg F) provided by the LED manufacturer.

Put the LEDs in a fixture (a bike headlight, or a screw-in replacement bulb for your home), and their actual performance is diminished by thermal effects. Photometric tests (http://www1.eere.energy.gov/buildings/ssl/newtest.html) are expensive, which is why you won't see many actual lumen measurements for bike headlights. Try to check out the product and gauge it against your needs (my eyes get more feeble by the day, it seems).

As for headlights, I've had my Exposure Enduro Maxx (http://www.exposurelights.com/products/enduro_maxx_2_-_2009/index.php) for a year (got it here (http://estore.websitepros.com/1939518/-strse-Exposure-Lights/Categories.bok)), and I love it. It replaced a NiteRider MiNewt X2 (battery pack failed on a cold rainy December evening).

10-22-2009, 11:28 PM
bubba, the guys discussing the magicshine light are aware of some of the things you mention.

So without reading 9 pages of thread. Is this actually 900 lumens or just blowing smoke? commenter: operator, on bikeforums thread (http://www.bikeforums.net/showpost.php?p=9755236&postcount=212)

10-23-2009, 08:34 AM
It might also be interesting to compare water resistance and cold-weather battery performance, although it would be tougher to devise those experiments.