When I first started thinking about how to evaluate a folding knife, I also began thinking about how to evaluate a flashlight. There are tons of reviews out there about runtimes and lumens. But I wanted something a little less engineer-y and more "how does this work". Those engineer-y reviews are very helpful, and I will always try to link to one, and they are a great place to start, but just like figuring out how many horsepower a given car has can't tell you what it is like to drive, this scoring system is an attempt to fill in the gap between technical specs and everyday use. Again, this is not an attempt to be objective, it is merely a quantification of my opinion. Instead it is a way to compare flashlights using one standardized method from one consistent perspective (mine). All of the scores are application-specific, again. Finally, as before, it works like this: 10 criteria, each with a possible score of 0-2. 0 means the light failed to work or perform as intended in a given criteria. 1 means it worked, and 2 means it worked well. The same exact feature could score a 0 in one application and 2 in another. For example, the McGizmo Sundrop, has no throw and tons of flood, but in its application--a portable high quality lighting source--the lack of throw is fine. In a high powered tactical light, like a shooting light, that same lack of throw is a fatal design flaw.
No evaluation of flashlights would be complete without linking to carrot's guide to high end flashlights. It is to flashlights what Joe T's posts are to blade steels...it is a great reference source. In fact, I am going to put it in the links on the right. All terminology is defined in carrot's guide.
With all that said, here is the scoring system:
A combination of brightness (lumens) and tint (color temperature). Simply because it is the unit of measure most are familiar with, I am going to use lumens. For a good explanation of lumens, see here. For more information on color temperature as designated by the Color Rendering Index, see here.
This is not about simply hitting a high number, but having a context correct output. A keychain light doesn’t need, nor should it have, a 250 lumen output. That much light in that small of a package is unnecessary and causes severe overheating and limits runtime. Its nice, of course, to have a few minutes of blinding light and also have a long useful low, so those will be considerations.
I am also going to take into account the tint of the output. If a light has a specific use that requires a high CRI tint, that will be considered. Generally, a closer to sunlight tint is better, but I am not so fanatical about it as to sacrifice a great deal of runtime or brightness.
Expressed in minutes or hours, this factor looks at how long you have useful light output. Usually measured to 50% power, most lights are now regulated. Regulated lights use special circuitry in the head of the light to make sure that an LED is functioning as close to the chosen output level as long as possible. Instead of a slow depletion of power resulting in an increasingly dimming output, regulated lights give you close to the selected output longer. There are two issues with this: 1) when the power is gone the light just goes dark; 2) there is really no warning that the light is about to go dark. Some lights, like my Muyshondt Aeon and my McGizmo Haiku have warnings to let you know this is going to happen, but very few lights do this. Again, the ideal runtime is use dependent.
Does the type of beam ("floody" or "throw") suit the application well? Throw usually requires a bigger head with more complex reflectors or advanced (aspheric) lenses or optics. The better a light is at throwing the worse it is at wide illumination, or flood. Many lights force you to choose between the two: flood or throw, though some lights, like the McGizmo Lunasol or the Surefire Aviators series don’t. There are two strategies for getting both throw and flood. The first is using multiple light sources, like the Lunasols or Aviator. Here is the multiple light source head from the legendary Lunasol light:
The second is using a focusing mechanism like the LensLights or, ugh, the Mag AA lights. Again, ideal beam type is application dependent.
This factor examines how useful and uniform the light output is. There are two parts to any beam—the hotspot (the bright center) and the spill (the dimmer outer halo of light). A good beam will softly transition between the two allowing you both highly concentrated light for maximum throw or illumination and softer, broader swathes of light for regular tasks and more area lighting. The beam should be perfectly round and free of any dark spots or rings as both can distort the image you see. Novatac’s lights have exceptionally nice beams. See here (Novatac on the right, Liteflux on the left):
How easy the flashlight is to use, how easy it is to switch from one mode to the next. Almost all high performance flashlights now have multiple outputs—a low, medium, high, strobe configuration being the most common. This also includes the mechanism of activation—a twisty, a clicky, piston drive, or some other means of turning the flashlight on and changing modes. I prefer a UI that: 1) minimizes the steps and actions necessary to change modes; and 2) is simple enough for a non-flashlight person to use. If the UI has you distinguishing between “presses” and “clicks” or requires you to do multiple things to switch modes (tailswitch activation AND twist the head), it is not as elegant as it could be. I like the Preon’s UI, but even better than that—the simple, perfect UI of the McGizmo Haiku. The Haiku uses one button and three clicks to get to everything you need—a good low, a good medium, and a nice high.
Hands Free Use (Tailstand & Antiroll)
Much of the time you use your flashlight in your hand, but in certain applications or when the power is out everywhere, you need your light to function on its own for a while. I find the ceiling bounce (aiming the light at the ceiling which is usually painted white, creating a diffuse illumination for the entire room) to be very helpful in all sorts of applications. In the shop I also like to be able to lay my light on a surface and have it stay put, illuminating something across the way. A light that has a “task attachment” in the rear is even more helpful.
How the light feels in your hand. A lot of single cell 123a lights are just too small to be held in the hand absent a pinch grip or cigar grip. I have found that I really like a single cell light to be about 3-4 inches. Smaller than that and you have to use a funky grip that is both unstable and uncomfortable. Again, ideal grip is application-dependent.
How the flashlight stows away. Usually this means that it is clipped to your pocket, but some lights, especially super small lights, live well in a coin pocket, around your neck, or even on a keychain. Some lights are too big to carry in a pocket, and for those this criteria examines how well they stow away in packs or the like.
Same as the design criteria for a knife: how does this light look on paper? Does it have ample heat sinking, is the grip designed well, does it tailstand well, is the clip integrated into the grip? Is the bezel crenelated (this allows for you to know if the light is on when placed face down)?
Fit and Finish
Again, same as the design criteria for a knife: how does the light look and feel in reality. Are the materials high quality (plastic v. glass v. sapphire lens, for example)? Are the threads cut well? Do parts thread together easily? Are the parts smooth when operating? Is the texture on the grip adequate? Are there gaps when assembled?
I am going to use the scoring system to evaluate my current EDC light--the McGizmo Haiku soon. I am not spoiling the surprise by saying that it is one hell of a light.