GC is right.

Candelas are a direct measure of both brightness/intensity and longitudinal throw.

Candelas and lux are both measures of brightness or intensity of light.

The difference is that one Candela is one lumen per steradian angle, whereas one lux is one lumen per square meter.

Candelas can be converted into lux, and vice versa, if the distance and steradian angle is known.

Here is a Candela to lux calculator:

http://www.ledrise.com/shop_content.php?coID=18

Now, about those lumens.

Remember how mechanical power in a motor car engine = mass x acceleration x speed.

In other words, power = force x speed.

Or power = torque divided by time.

Likewise, electrical power = voltage x current (in Amps).

Similarly, the total power of light (in lumens) = the brightness/intensity (in lux) multiplied by the surface area of illumination (in square meters).

The brightness/intensity in lux is the total lumens per square meter.

The brightness/instensity in lux determines the longitudinal throw.

The brightness/intensity in Candelas also determines the longitudinal throw.

The total lumens gives you an idea of the product of the brightness in lux, and the size of the surface area illuminated in square meters.

Lumens gives you an idea of both the lux and the surface area illuminated, combined in one total.

**Now, your formula of Candelas divided by lumens factor?**

Is basically lux divided by lumens.

One lux is one lumen per square meter.

Thus, lux divided by lumens = (lumens per surface area in square meters) divided by lumens.

This equals (lumens per surface area) multiplied by (1/lumens).

Note the lumens cancel out, leaving you with (1/the surface area); that is, the inverse of the surface area of illumination.

Thus, if I'm not wrong, your Candelas per lumen, is actually roughly equal to the inverse of the surface area illuminated.

If you invert your mathematics to lumens per Candela, or in other words, lumens per lux, this will give you the surface area of illumination alone.

Lumens = lux x surface area.

Surface Area = lumens divided by lux!

In other words, surface area = lumens/lux.

Which is similar to lumens per Candela, remembering that if the distance and steradian angle is known, then Candelas can be converted to lux, and vice versa.

One Candela is just one lumen per steradian, while one lux is one lumen per square meter.

Thus, rather than your original Candelas per lumen, try lumens per Candela, or even lumens per lux - this will give you a rough idea of the lateral surface area of illumination in square meters, or at least a surface area of illumination co-efficient for comparison purposes.

You know the total output in lumens, giving both the brightness in lux and the surface area of illumination, combined in one total.

You know that Candelas and lux are a measure of the brightness and hence longitudinal throw.

The higher the Candelas per lumen ratio, or lux per lumen ratio, the greater the surface area of illumination factor.

Your E21 XP-E with 25mm bezel diameter has 154 lumens output and 4510 Candelas for 134 meters of throw.

Meanwhile your TK15 XP-G with 34mm bezel has 337 lumens with 11592 Candelas for 215 meters of throw.

The TK15 has more Candelas, and throws further.

The TK15's total lumen output has roughly more than doubled too, because the brightness/intensity in Candelas has more than doubled.

Note that the TK15's reflector is larger, so the hotspot will be smaller and more intense for greater throw.

If the TK15's reflector has a deeper relative depth, then the spill will be narrower, the hotspot possibly even brighter still, or at least the corona will be more prominent.

If we calculate the lumens per Cd ratio, the E21 comes out at 0.0034, while TK15 is 0.0029; meaning that the E21 has more surface area of illumination, or spill; consistent with a smaller reflector.

So does the E21 have a wider spill than the TK15?

However it gets confusing, because the TK15 may have a larger hotspot than the E21?

Meanwhile we know that your Maglite has a high Candela, long throw, but very dim spill, but the lumen/Candela factor of 0.0043 says very little.

**The reason the lumen/Candela, or your Candela/lumen factor says little is because the light beam is not uniform on the white wall.**

The beam is inconsistent, with a concentrated hot spot of variable size, a less concentrated corona of variable size surrounding that hot spot, and a variable size and intensity spill.

Thus, when purchasing or comparing flashlights, the best way to compare their lateral performance or "spill", is not by numbers, but by white wall shots that are controlled.

For example, Selfbuilt keeps the flashlights 0.75 meters away from the white wall, while the camera is 1.25 meters away.

The aperture and shutter speed are known.

Then the beam on the white wall are compared.

Look for six factors: the intensity and size of the hotspot, the intensity and size of the corona surrounding the hotspot, and the intensity and size of the spill.

Unfortunately, there is presently no objective and quantitative way to measure the lateral spill of light on the spec sheet - only a subjective and qualitative controlled white wall shot found in flashlight tests and reviews...