Make no mistake I get excited when new emitters come out. The tantalizing possibility of an even brighter and more efficient LED allows me to imagine my future perfect EDC light. But after a few days I realize that a lot of the lights I have right now work really, really well. I have to admit though, it was the insane 200 lumen high of my Peak Eiger AAA that enticed me to buy the light, so even I can be bit by the emitter bug. After writing this blog for two years though I realized that old emitters, properly implemented, can run with the big dogs long after they ceased being cutting edge.
The secret of flashlight output is that it is the reflector and the optics, more than the emitter, that determines how useful your flashlight really is. A cutting edge emitter in a substandard optics package (which is what the majority of budget lights are) can give you really great "emitter lumens" numbers but terrible and disappointing output in reality. Emitters are important, but only a part (small part) of the puzzle. They are, however, the thing that sells a light, so it is important to understand what all of the names and letters mean.
To that end, I am going to explain the nomenclature and a little about the more popular emitters and then give you a few things to consider in addition to the emitter that will have a big impact on the light's performance. I am not an engineer and I have very little technical knowledge, so if I make a mistake, please chime in in the comments below.
Before I get started, here are two excellent references on LEDs: 1) photo thread of all major LEDs used in flashlights; and 2) tech specs on emitters.
A Brief History of LEDs in Flashlights
About ten to fifteen years ago LEDs started showing up in flashlights. They were weak, produced icy cold blue light, and did not offer any price advantage over incandescent lights. But slowly all that changed and then about five years ago the LED revolution kicked into high gear. Instead of massive shower-head type arrays of weak LEDs combining together to get modest output, a single emitter could hold its own with an incan light. The colors were still awful, but things were moving along.
Three things happened that really spurred this revolution. First, the historic explosion of manufacturing in China allowed a huge number of flashlights to be produced quickly and cheaply. Major names like Fenix got their start as China's rise began. This massive surge in capacity gobbled up LEDs quickly and demanded better ones. So there was a real market for LEDs. Second, the internet allowed people that liked flashlights to gather in a single place and swap information and lights. Doug's old Flashlight Reviews was a mainstay of mine and reading about LEDs as they overtook incans was pretty exciting. I still remember his first Surefire LED review. Finally, flashlights were expensive enough to warrant the use of power regulating chips. Without these circuits LEDs would have a hard time operating. An incan is least efficient when it is on high, meaning most of the power used by the filament is lost to heat not more light. The reverse is true of an LED. It is most efficient when on high, and a low is much more taxing on the emitter. Without power regulation, less power means greater inefficiency, which consumes more power from a dying battery and so on. These power regulating circuits also allowed for multiple output modes when coupled with other technology.
About five years ago LEDs moved into a tie or actually outpaced incans, in every area but color rendition. Then about three years ago, Hi CRI emitters came out. You can read more about them here.
LEDs: Companies and Names
There are three parts to the name of an emitter, the company or brand name, the series designation, and the bin code. None of these things are the end all be all in terms of telling you how good a flashlight is, but they can be useful comparing models.
First, the company or brand. There are only really eight companies making LEDs: CREE; Lumileds; Edison Opto; Led Engin, Inc.; Nichia; Seoul Semiconductor; OSRAM; and Luminus. Of these only a few make commonly used emitters. The most famous, as of this minute in the flashlight world, is CREE. Their XPG and XML emitters are found on a ton of different lights. Nichia makes some very good Hi CRI emitters, while the others have had their moment in the sun in the past. I particularly liked the OSRAM Golden Dragon emitter I had on my Nitecore EX10. The vast majority of emitters right now come from CREE. They are the company producing the XLamp series of emitters (XPG, XML).
Bin codes deal with four characteristics: flux rating, tint, Vf (voltage forward), and color. Simply put, with most emitters, the higher the letter, the more light, given all of the other settings being identical. A "U" bin designation will produce more light than an "R" bin designation. Here is more information bin codes. This doesn't however, tell you a whole lot because emitters can be configured very differently. They are likely only helpful in situations comparing two identical lights, driven the same way. In that case, the emitter with a later occurring bin letter will produce more light. The bin code does not describe ANY other characteristics other than the four it is designed to represent and in a vacuum it means very little.
The Heavy Hitters
Emitters are all about trade offs. There are three attributes that I think are essential to evaluating emitters--max output, runtime, and color rendering. All three are in tension with each other and all three are important. Super high max outputs usually come from large lights and large batteries and usually at the cost of good runtimes. Max outputs also sacrifice something in terms of color rendering. Methods of producing accurate colors almost always siphon off a few lumens. The best emitter, in my opinion, is one that hits over 100 lumens, can do so for about an hour (total, not all at once), and has Hi CRI color rendering. This is, of course, very hard to achieve. So emitters, like blade steel, can give you some of what you need, but not all of it. No emitter does everything. Finally, much of what we actually see, that is, what the light looks like in reality, has nothing to do with the emitter. Nonetheless, it is probably helpful to explain a little about the big name emitters on the market now.
XPG (and XPG2): This is a CREE emitter and the product page for the XPG can be found here. It is no longer their highest power emitter on the market, that would be the XML, but the emitter itself is much smaller allowing for a different beam pattern and reflector shape. The updated version of the XPG, XPG2, is much more efficient, giving you 20% more lumens per unit of power and it runs longer. Here is CREE's product page. Lots and lots of people still like and actually prefer the XPG series to the XML series. The reason is pretty simple--a smaller emitter allows for more reflector surface and that, in turn, "scoops" up more of the light produces and pushes it out the front of the flashlight. The smaller emitter also allows for different shaped reflectors, as less of the reflector surface is committed to the hole the emitter has to fall into. This also has a little to do with the size of the light's head. If you want the most reflector surface possible (which you should, as it means less lost light from the emitter) the XPG, because of its smaller size, will give you more reflector surface in the same sized head as an XML, all other dimensions being constant (the loss comes from the need for a bigger hole in the reflector to accommodate the larger emitter). The XPG comes in all tints, and there are some Hi CRI XPG emitters, but they are rarely used (for more on Hi CRI see here). The issue, if I had to guess (and it would be just a guess as I am not a flashlight maker) is that the Hi CRI XPGs score a 90, while their competitors score at least a 92 are the same or similar price and have the same or similar runtimes and outputs. Malkoff lights, for example, are available with an XPG Hi CRI emitter.
XML: This is also a CREE emitter and the product page can be found here. This is the newest emitter from CREE and there are lots of variations. The big difference between the XML and the XPG is the size of the emitter and the max output. The XPG series emitters are 3.45mm x 3.45mm. The XML emitter is monstrous by comparison: 5 x 5mm. The drawbacks associated with this size are outlined above, but the advantages are pretty obvious--XMLs pack more punch. You can get higher lumens counts on XMLs than you can on XPGs. Significantly higher. The jump in lumens outputs that happened last year, where lights went from 300 lumens on a single CR123a to 500 lumens is associated solely with the emitter upgrade. A good maker can still coax amazing performance out of an XPG or XPG 2 emitter, but the easier way to get more punch, is to simply jump up to the XML series. There is one big drawback besides size--energy consumption. XMLs, to be used at their peak level of performance, need some serious juice. They can run on 1.5v, such as in a AA light, but to really open them up, you need 3.0v or even more. Ideally they'd like to be driven by something like an 18650 battery that can pass through super high voltages like 4.2v. One other interesting feature of the XML, is the ease with which the emitter can be modified or changed. CREE has a series of different XML emitters, all that can do different things and this is in part, I guess, because the larger emitter allows them the ability to alter the insides more easily. If you have the batteries to really take advantage of it, an XML is probably the state of the art in terms of output.
Nichia 219: Here is a good sales page with specs. If the XML emitter is the high horsepower muscle car, the Nichia 219 is the almost as fast, but vastly more refined European sports car. This is, in my opinion, the connoiseur's choice. The Nichia 219 is a small emitter, similar to the XPG (it is actually sold as a direct replacement to the XPG). It has similar power needs, outputs, and runtimes. But, and this is a big difference in my opinion, it produces the most accurate light of any regularly used emitter anywhere. It gives off a radiant beam of sunlight, hitting 92 on the color rendering index. The result is an emitter that never washes out colors, that makes reds look red and greens look green. The light I am currently testing, the Overready Edition of the Peak Eiger, runs a Nichia 219 emitter and the light is just perfect. If you have a choice, this is the emitter you want. There are some sacrifices though. Hi CRI emitters, all else being equal, produce less lumens than their non-Hi CRI counterparts, so there is a bit of a brightness sacrifice (like 15% of output). Second, they are marginally more expensive, compared to their XPG counterparts, around $10-$20 more on the consumer end. That said, these are sacrifices I am more than willing to make. We carry lights to see in the dark. Why not see better?
There are a lot of things that go into making a light seem bright or correctly tinted. Surefire lights have been able to "punch above their weight" in terms of lumens compared to perceived brightness for years because Surefire has the best reflectors you can find on a production light. They gather all of the lumens and pump them out into a narrow beam. Similarly, the lens on the light itself is a big deal. If it is not as clear as it can be and treated with anti-reflective coating, it can rob you of lumens. Lens are the most straightforward so I will address them first.
Lens are pretty simple. They are usually made of one of three different kinds of material: 1) plastic (polycarbonate); 2) glass (mineral glass); or 3) sapphire (artificial corundum). Plastic is the least clear and least scratch resistant. It is relatively tough. It is made for cheaper lights. Glass is better in terms of clarity and scratch resistance. It is more expensive. Sapphire is the best in terms of scratch resistance, as it is harder than steel by a substantial amount, and it can rival the clarity of some UCL glass (ultra clear lens glass). It is the most expensive and because of that high hardness a little more brittle than glass. The best lens, like sapphire and UCL, will ward off scratches and coupled with anti reflective coating will allow as much as 10% more light through than cheap or uncoated lens. That's not enough for your eye to notice a difference, but if we are at the point of debating bin numbers on emitters, then we are already at the point where 10% more light matters. Simply put--avoid plastic if you can.
Reflectors are much, much more complicated. Designing reflectors is something that requires a ton of math and optics knowledge, far more than I have or have the ability to acquire. Reflectors, more than emitters and lenses, can make the biggest difference in how a light performs. A well designed reflector is, in my opinion, vastly more important than the latest emitter. Two general points about emitters are important.
First its surface. There are two types of surfaces, stochastic and smooth. Stochastic surfaces are referred to as orange peel reflectors. Stochastic refers to the method by which the surface is patterned, and a truly stochastic surface is entirely random. Lots and lots of math and optics are needed to explain this, but simply put a stochastic surface produces a smooth beam. The slightly bumpy surface scatters the light and smooths out the beam, giving you a clean spill and an edgeless but intense hotspot. Novatac's reflectors were wonderful for a production light, as were the aforementioned Surefire reflectors, but nothing I have ever seen comes close to the perfect beam profile put out by the McGizmo Haiku reflector. This is the zenith of reflector design in my opinion. If you need a more focused beam and the rings, holes and other artifacts that are present in most lights don't matter, you can opt for a smooth surfaced reflector. Personally, I like the floody and clean look of a textured (orange peel or very light orange peel) reflector over the swiss cheese like look of the smooth reflector.
Second there is a new trend towards more sophisticated reflecting technology called a total internal reflector. The TIR is not a silvery surface of domed metal at all, but an actual optic inserted into the head of the flashlight that sits around and above the emitter. It gathers up light and sends it out the head, just like a reflector does, but it uses optics and angles to bend light from the emitter and send it forward. Surefire has a few TIR lights. They are known for producing better throw than normal reflectors with smoother beams.
Finally, there is the combination of a lens and reflector called an aspheric lens. Lots of focusable lights, like the previously reviewed LED Lenser M7R (which has a TIR inside and an slightly aspheric lens outside) and the LensLight have convex lens that send light out the front in a tightly focused pattern. These optics create a magnifying glass effect that can be used to spread out or shrink down the beam.
In the end, the performance of a flashlight is probably something like 25% emitter, 65% reflector, and 10% lens. TIR and aspheric lens throw those ratios off a bit. Emitter designations are important to know and you can tell a little about the light from this information, but real world performance is more closely tied to a well designed reflector. Nonetheless, if you want the most eye scorching light ever, know which emitter is which can help. If you can, get the Nichia 219. The XPG2 is a nice choice as well. If you want nothing but lumens, then most lights on the market right now that can deliver use XMLs.