Supernova E3 Pro 2 dynamo front light

I have had this headlight for two years and have just replaced it with the SON Edelux II. Despite all of the super reviews received, which is why I bought it, I personally have found the performance disappointing. The light has no real power to it and is spread too wide, where as the Edelux II just outshines it with a really powerful beam focused perfectly where you need it and this is from a lamp which is rated at half the power of the Supernova. I have a Busch Muller IQ-X lamp on another bike which also outshines it.

Morning,

Did you ever get round to reviewing the rear light too?

Hi Gregor, and thanks for the reply! It's great to hear that you're tracking customer comments.

100-130 lumens "measured" by whom and by what method?

I believe Olaf Schultz measured it at 140; I think Wouter Scholten agreed that that seemed roughly correct (his review and discussion of these results may be found at http://swhs.home.xs4all.nl/fiets/tests/verlichting/koplampen/Supernova_e... ). My own measurements suggested slightly less light, but I was using a homemade integrating box (a small, mostly empty room, white walls) and computed brightness using a camera's light meter at a few points around the room (turned out to be pretty consistent on the scatter-lit surfaces) vs. a couple of NiteRider ("we measure our actual output") lights at various lumens from 80 to 450. I drove the light using my Schmidt SonDelux on my 20" (406) wheel, I believe at a steady 16km/h using an electric drill and verified using my bike's wheel-speed-sensing computer. Measurements were at room temperature, no wind, but the casing didn't heat up a whole lot during the brief testing periods, so wind chill probably wouldn't have made much difference. A rear light was attached; it was your own 3-LED light that's meant to work with the headlight. I believe I disconnected the taillight and repeated the experiments and wasn't able to measure any difference using my crude equipment. Incidentally, my eyeballing of light output side-by-side of the "300-lumen" lights (yours, NiteRider's, and a more modern NiteRider with a "low-power" 300-lumen mode) also suggests that the NiteRiders is vastly brighter than the e3 Pro.

Now, the discrepancy might be due to any of a few things:

• Measurement error? I'm confident that my measurements are not terribly accurate, but also sure (based on eyeballed sanity checks (e.g., that if I shine a 300-lumen NiteRider at the ceiling, I can easily read small text under my desk, and that that's impossible with my e3 pro at 16km/h)) that they're not inaccurate enough to account for the measured discrepancy.
• NiteRider habitually understates their lights' output across a few different models, by at least a factor of 2? Possible, but seems like a poor business decision...
• I got two NiteRiders whose quality control was such that they were both much brighter than they should have been? Given their atrocious quality control (I've had electronics problems with 4 NiteRiders now) this wouldn't be too surprising except for the direction and magnitude of the error required.
• Supernova overstates their lights' output at least for the e3 Pro? This seems likely, since the new model with a new LED isn't being claimed to be any brighter than an older model (or has Germany changed the laws such that they now limit lumens?).
• Peter White Cycles sent me an older light, mistaking it for a newer one? What do you think of this hypothesis?
• I got a dud? Coincidentally, so did the couple of other people whose measurements of the e3 pro I've heard about. Of course, those who got lights that put out the advertised 305 lumens might not bother to measure them.

The Terraflux2 Lens excels at throwing a lot of light onto the ground and efficiently illuminates the road ahead.

Yes, mine is good for riding in straight lines (as long as you're not worried about prairie dogs running across the path in front of you ) Not so great for winding bike paths, as I described. Again, it's possible that this has been improved.

it's not the brightest out there and we wish we could make it brighter without sacrificing the longevity of the capacitor,

Perhaps a bigger capacitor? A standlight ought to be bright enough to stand out against a sea of car headlights. Other than second-guessing your design decisions, it's encouraging to hear that you build the lights to last!

The new Tail Light2 is 4x brighter and has a much wider beam angle.

Those are both good things, but what about emitter size? As I tried to describe above, one thing that is crucial in a taillight (and is also nice in a headlight, incidentally) is having a large enough area that motorists can judge closing speed by how quickly the emitter's apparent size is changing. That's difficult with your taillights. Another consequence of the tiny-but-very-bright taillights is that cyclists may have difficulty comfortably riding behind you; this too is somewhat alleviated by having the same total amount of light coming from a larger surface area. At least yours don't blink the way most lights here in the USA do... those things are a holy terror to ride behind! (Effective vs. motorists, though

I really do appreciate the response. I'd love it if more people would chime in in order to generate a more balanced discussion of the goals of bicycle light design, and/or tell me that I got a dud, as I think that Supernova does some things very well indeed, and other things not so well.

Hello bwpearre,
let me quickly address some of your points before I get ready for our Tuesday nightride.

1) Lumens brightness
100-130 lumens "measured" by whom and by what method? These measurements are simply not correct. There are so many factors that can influence this, like the dynamo used,the type of rear light attached and many people just measure the brightness by attaching 6V DC, which produces a lot less light than the AC from a dynamo the light was designed for. The data sheet does actually support our claim and that is probably the cause for this misunderstanding. LED manufacturers used to spec the LEDs at a temperature of 25° C. In "real life" the chip actually heats up to 80°C even in our well-cooled housings and loses efficiency. The general spec-sheet rating of e.g. 130 lumens per watt (meaning LED current 350 mA = 1 W) cannot be directly scaled up to the 3.5W we use to power some LEDs. The next generation Cree XPE2 and XPG2 LEDs are binned at 85° junction temperature and therefore will have more realistic specs. The real brightness also varies with the temperature of the environment, riding speed (wind chill), and ride duration. Therefore, the lumens value can only ever be a rough indicator for the average "brightness class" of a light. The human eye has great difficulties perceiving brightness differences below 30% (which is 90 lumens if you take a 300 lumen light, for example) and it adapts to the brightest area in a beam, which is why light distribution and beam shape are far more important than raw brightness. The Road CC beam comparison does a great job comparing beam shapes and allows the user to get an idea of how the different beams compare. Of course every beam shape has advantages and disadvantages. The Terraflux2 Lens excels at throwing a lot of light onto the ground and efficiently illuminates the road ahead. A symmetrical beam is often better for off-road riding (although we also use our Airstreams to do 24hour-races).

2) Stand light
Yes, it's not the brightest out there and we wish we could make it brighter without sacrificing the longevity of the capacitor, but we give a 5-year warranty on full fuctionality and we have to stay within the recommended operating parameters of the capacitor. The stand light is bright enough to carry the bike into a dark cellar or find a keyhole and it warns drivers of your position if you stop at a traffic light, which is its original purpose. I can promise, though, that this is one of the aspects we will improve in the future.

3) Rear light
Forget the old one, it had a bright beam at a 30° angle but the LEDs used there are now 7 years old and are just outdated. The new Tail Light2 is 4x brighter and has a much wider beam angle. It gives you a great feeling of safety, even on very busy roads. I'm certain anyone who is using it will confirm this. If you want an upgrade, just send an e-mail to contact [at] supernova-lights.com (this will land on Regina's desk so you can say hello from me).

In general, we are always happy to get input from anyone using a Supernova. We are commuting and racing with the lights ourselves but it always helps to get an outside persepctive.

Thanks,
Gregor

Supernova

Hi Dave! Thanks for the response. I don't quite agree with some of your points:

we don't set out much stall by any manufacturer's lumen ratings, to be honest. if that's your basis for calling supernova a 'shady company' then it's pretty weak,

I disagree. Sure, it's difficult for a user to verify lumen measurements since most users don't have regular access to an optics lab, so manufacturers can get away with lying. But I would say that it's exactly as acceptable to sell a 120-lumen light as a "300-lumen" light as to deliver 160 grams of coffee beans to someone who paid for 400 grams. Lumens are sufficiently well-defined--they're not a marketing term but a physical quantity that all scientists agree on. I don't see why "most manufacturers lie about it because they have been able to get away with it" means "there's no point in calling out manufacturers on their lies". You're just giving them more reason to believe that they can get away with lying to their customers.

We could argue the toss about *exactly* how much light it emits

I agree with you!! But this isn't about whether it is off by 10 or 20%, which wouldn't bother me. Rather, they're off by well over a factor of two, and in a way that their engineering department must surely know about! The older "305-lumen" Supernova was a very significant step down from my 300-lumen NiteRider (I hear that the latter company actually measures light output) even given Supernova's superior light placement (on straight roads, at least...). Sure, *exact* output is not relevant, but I'd argue that *approximate* output _is_, and that their lies are too big to ignore.

At the same time, I agree that you are, as you should be, testing the light's usability, and I love that you focus on what matters to riders in the real world. I'm just arguing that a manufacturer's willingness to sell something that they can't even come close to delivering may also matter

the beam pattern means it performs better than them in road situations

Actually, now that you mention it, there's a rare but potentially serious problem with their beam pattern that shows up on winding roads such as the bike paths here in Boulder (not the most common situation, but perhaps worth noting). Because the vertical cutoff slopes down towards the sides, when your bike banks into a turn, you suddenly go from having light to having no light at all right where you're trying to go (e.g. 30 degrees left of where your handlebars are aiming), which is more than a bit dangerous on a narrow bike road. The vertical cutoff should be flat over a wider range left-to-right rather than dropping down to the sides...

I do agree with you about the build quality, and it has the best mount of any light I've used. So that's a definite plus...

standlight isn't super bright, but it's plenty bright enough to get you seen when stationary

Thanks! That's part of what a standlight is for, but (a) how does it look from ahead and behind next to a row of cars' lights in a line of traffic during rush hour, or if you're, say, waiting in the middle of an intersection for your opening to turn left through a gap in the oncoming traffic? and (b) how well does it work for putting your bike in the shed after your ride? I've briefly encountered a few standlights that seem to work pretty well for this, and even for changing tires, but the Supernova of 2011 seems worse than average on these counts--especially (a). I'm glad to hear that it's a little better now; I hope they keep improving.

are you suggesting that drivers are able to triangulate distance based on the width of a wider tail light?

Absolutely, from several dozen meters away, depending on the light. I found that it's easiest to experiment with this effect with a reflector (try a 10cm wide one to start with), as those provide even illumination over a large area similarly to an ideal taillight (but, as you know, reflectors rely on optimistic assumptions about the car driver, etc, so should not be used exclusively). Maybe try a Philips Lumiring (no relation to me) or some such for comparison. You don't think so?

I'm excited if this light is better than mine and if the manufacturer is learning not to lie about it. They make a product that seems to show some promise, and frankly the state of dynamo on-road lighting was dismal enough in 2011 that even the e3 Pro was one of the best available. Here's to more better lights, and to fewer lies from manufacturers. Keep up your good work!

I'd be really curious to hear your or your readers' light output measurements. I have the older e3 pro, rated for 305 lumens as well, and I and several other people have measured it at 100-130 lumens at speed. To say that Supernova is being optimistic is to give them too much credit: they outright lie like scoundrels--even the datasheet on that LED doesn't support their claim, let alone actual measurements. So if the new one still claims 305 lumens, maybe they're realising that they're getting caught in their lie and trying to backpedal...? If this really is a 300-lumen light, I might consider it (aside from having to do business with such a shady company), but I'd want some independent verification first. Do you have any idea whether we should trust them this time around?

Also, the "e3 pro" also doesn't have a standlight to speak of--brightness goes _way_ down, almost to nothing. Has this been improved as well? Do you have an estimate of standlight output?

Finally, the matched taillight is a moderately bright but tiny point source, making it impossible to judge closing speed, and is thus about half as useful as a well-designed taillight (it tells motorists that something is out there in a given direction, but gives no information as to whether they need to take action). I'd highly recommend against using Supernova's taillight, even though it's convenient and cute.

Have the older version on the 'bent, and it's brilliant. Beam shape is the killer - it's got a really obvious horizontal top edge. Set it up so this is just below the windscreen of oncoming cars, and all the light goes somewhere useful. I've only really found it struggles when trying to spot potholes in driving rain with lots of oncoming headlights and ambient light, but then nothing really works in these circumstances.

Downside of the shaped beam is that it doesn't illuminate raised roadsigns, which is a bit of a pain when navigating in the dark. Pair with headtorch or similar.

Cleverly priced to be more or less the same as the upcoming B&M Luxos, though that will have the USB charger on it to. Looks noticeably brighter and (more importantly) wider than my B&M Cyo IQ which is more than enough light for dark lanes all night long at up to 25mph.

Dyno lights are great, speaking as a commuter and practising 'granled Audaxer'. I would take it touring but obviously you'd need a head torch as well as you can't light your camp site like you would with a battery light.

Beam shape is about as important as bang on the road, something only B&M and Supernova seem to have mastered yet.

It is a big investment all round but next time you get commuter or training wheels built I really would recommend the extra outlay, it's just great not having to phaff with recharging when you riding to work everyday of the year.