It varies depending on who you ask, but the difference probably works out to something around 20%. The SC52 is pretty floody, so the difference would be even harder to notice.
I measure the SC52 at a genuine 210 OTF on eneloop and 480 on 14500. It’s all relative though, Zebralight do use OTF ratings but could be described as more generous in those than some other companies. It’s hard to say for sure who’s readings are most accurate, but they are on slightly different scales. If you take a look at selfbuilt’s reviews and his lumen readings then they are generally in accord with Zebralight’s scale and you find Foursevens and Fenix lights often show output higher than their claimed ANSI specs.
The jump from 200 to 280 is an absolute difference of 40% but brightness is perceived logarithmically, so that difference will not be perceived as a 40% increase. This basic idea is explained quite well here: richardbrice.net. The next link goes into it a bit further, and contains a nice graph showing the actual exponents for different senses according to Steven’s power law: Psychophysics | in Chapter 04: Senses. A more complete list can be seen here: Stevens's power law - Wikipedia; the exponent differs whether it relates to a point source, a lit target, or a brief flash, but for what we are considering the exponent would be 0.33.
The way I understand it this means that in practice you have to multiply a light output by 3 (well actually 3.3 but I have rounded to 3) in order to give an incremental increase in perception. For instance if 1 lumen was viewed as 1 brightness (arbitrary unit), then you would need 3 lumens for brightness 2 and 9 lumens for brightness 3, i.e.
1 - 1
2 - 3
3 - 9
4 - 27
5 - 81
6 - 243
7 - 729
8 - 2187
9 - 6561
This idea is explained quite well here: http://www.indiana.edu/~p1013447/dictionary/log_r.htm, which concludes quite nicely: ‘The reason the senses are built this way is the enormous range of stimulus intensities to which our senses, especially vision and hearing, can respond. Full sunlight produces an intensity of reflected light that is 10 billion (10,000,000,000) times as intense as the faintest light that a human can detect. Because brightness is related to the logarithm of light intensity, the range of brightness is much smaller than the range of stimulus intensity. For the same reason many instruments that measure physical quantities that vary over a very large scale are also equipped with logarithmic scales.’
Further to this, when you take into account the massive variances in beam profile between different lights as well, and the fact that we are poor at estimating light output in terms other than lux at a point, the brightness differences that you will actually perceive become even more difficult to predict.
Left to Right… SK 68, SC52, & Mrs D’s lipstick
Now that I have the luxury of quite a few lights to choose from, I have found that for serious night hiking or skiing, especially on sloping and rough terrain, a light with a flood lens clipped to my hip belt that augments a good strong headlamp throwy beam on my head is much better than just the headlamp alone. Aside from getting a better view of what you’re about to step on (or off), I notice a psychological effect of feeling more comfortable, probably due to have my previously dark peripheral area illuminated.
At the moment I use a Spark SD6 with the interchangeable flood lens clipped to my hip belt (and an ST6 on my head), because that’s what I have available, I assume that the Zebralight or any other one with similar characteristics and a NW tint would work just as well for this purpose.
Yes
You know, I think the OP (or anyone) might like one of the 2100s that Eric did - moonlight to 2.8A blaze. Hmm, I wonder if the driver would swap with one of the Convoys?
It’s the same driver (105c) but with custom programming so it should fit.
Swap the driver into a Convoy host and you’ll probably end up with a better light. Eric did a good job on his mods, but mods are always limited by the quality of the original host, and those UF2100 were not that great as hosts.
I knkw that there has been much made here about the difference beteen 200 and 280 lumens. But what light actually produces 200 lumens on 1xAA? That’s REALLY the poiint of this thread: is there a decent light out there that has output between the SK68 and SC52 that is perhaps cheaper. And it doesn’t seem like there are any. Although some lights state high outputs, the fine print always reveals that you need a 14500 to get it. But with NiMH, you are basically at SK68 levels of brightbess.
As RedForest UK mentions, that 40% sounds like a lot to a meter, but to the human eye it’s nothing. You’d notice it if you were rapidly switching from one to the other, but your eyes would soon adjust and the difference would seem lost.
The EagleTac D25A will give you 200 lm, as will the Nitecore EA1 with the XP-G2 if you’re looking for an outdoor light with more throw than the Eagtac or Zebra, but less than the Sipik.
Apart from the battery format they are about as different as it gets.
Apart from price and output they also aren’t that different. 
Features …. SK68 Clone ……. SC52 (if only there was a clone)
Size ………… Pocket …………… Pocket small
Output ……. Good ……………. Very good, best in 1x AA
Price ………. Dirt cheap $5 …. Not Bad about $60
Quality …… Not bad, varies … Very good, occasional issues.
User Interf. Good ……………… Very good
Mods ……. Potential good …. Already there.
Resale …… Not worth post … Very Good, maybe 80 to 90% less shipping.
Bragging … LOL ………………… Distinct envy from lesser light owners.
I’m not certain what ‘already there’ means with regard to the modding capabilities of the SC52, but according to most who try, modding the SC52 is really a no-go. If you can even pry the bezel ring out without cracking the lens, the LED is mounted to the main electronics PCB. The SC52 is really not a good modding platform at all.
Yup, I don’t see what is left to mod on the SC52 without repurposing it.
I think what he means is that the performance of this light is at a point where you are not going to make it any better by modding it - at least with NiMH. Let’s face it. You don’t exactly see people modding 1xAA lights like the SK68 and getting, say, 400 lumens from a single Eneloop. That’s because the SC52 REALLY pushes the envelope when it comes to what is possible from a single AA NiMH battery. I think alot of it is the driver. Simply put, you’re not going to find an off the shelf 1xAA driver at Deal Extreme or Kaidomain (or anywhere else) that comes close to what you get in the SC52. So unless there is someone out there who is an electrical engineering wizard who happens to design their own driver that can, say, kick out 1.5A to an XM-L2 from a single Eneloop, you’re not going to do better than an ‘off the shelf’ SC52.
Hi,
I scanned but didn’t read through this entire thread, but are you limiting to 14500 only?
If not, take a look at take a look at the UF-K21:
http://www.fasttech.com/search?uf-k21
They’re 123/16340 lights, but VERY bright and cheap.
I do have an SC52, and the 5-mode UF-K21 I have is, to me, quite close to the SC52 in brightness.
For a step up (in price at least), there’s the Xtar WK21. md-lightsource has it.
http://md-lightsource.com/searchquick-submit.sc?keywords=wk21
I don’t have that yet, but it’s specs are not bad for a small 16340 light.
Well, its a CR123 light?
sc52 is clearly the winner!!
FT’s spec says “CR123A”, but I’ve only used that once (a TF CR123A, it’s a primary/non-rechargeable), and then have been using it with 16340s since then.
Their specs also say:
EDIT: One thing to note is that both the UF-K21 and the Xtar WK-21 are twisties.
Armytek has 300 lumens but with no low modes is horrible. No low mode is a no go!
Probably not much in between that puts out 250+ lumens..
just wait a few months, and more lights will reach that point I think.
options are: Lumintop ED15, Olight S15, Thrunite Neutron 1A
Olight should be in the $40 range..