Usually to find out how hard is driven a led using some guesstimate logic and throwing some nuber for efficency loss in conversion. Not an exact science but you get the idea on what runtime yo ucan expect roughly and the output.
Sure a NiMH cells can handle it. At least a good one can. That NimH cell you have probably drops to 1v or so under that load, so we're only talking about 2 watts of power compared to something with a single li-on and an XM-L @ 3A which is 10+ watts. The alkaline would have VERY little capacity at that draw and could not sustain it.
EDIT: Just to be clear, if the driver was 100% efficient (which its not), with 2.6A from the cell The LED would be seeing maybe 750mA. You can't forget the voltage. You have to think about the power (voltage x current).
curiosity, safety, circuit efficiency, run time calculations, comparison to other samples (quality control indicator)
why not measure it?
thanks budge.
why dont we measure runtimes and simply clock the light until it dims? -- that's a realer measurement and no more guessing, right? My reviews (Xeno, Quark, Romisen, Tank, Worm, ..) all contain real life measurements of runtimes. no guesstimates.
output? -- the manufacturer's ANSI FL-1 Standard rating should suffice :)
yeah, smart a**ck i am ggg
I was just curious really. Wanted to see if mine was driven as hard as the ones other's reported. At this point I am wondering if the "driver" is bad, although it must be working since it's boosting the voltage to the proper level to drive the LED. I'll also check tailcap currents to see if the light is being driven as advertised, get an idea of runtime, and make sure the driver is installed properly (I posted a case where my LED's neg. lead had shorted to the ground ring of the driver and therefore was "direct driving" the LED at too high of a level.
-Garry
actually, that's enough of a good and valid point. thanks for being honest 
Out of curiosity i will punch in my DMM's leads into anything i get served on my plate. gouda, goulash, bigos, bratwurst, cozido, cidre, torta, tripas, hamburguesa.
Will feel like Bart Simpson with his new electronic toy 
I always measure pull at the tailcap, just to get an idea how hard the driver is pushing the emitter. There is a huge difference in light output on an XM-L when it measures 1A at the switch compared to 2.2A, and it is the easiest way to tell.
only my Zebra is capable of drawing such currents from nimh. All my numerous Sipiks have drawn far less…
Driver efficiency plays a big role, tailcap resistance, quality of the reflector coating, glass lens transmittance. All these can cut from the light quite good, in the end a 800 lumen approximation based on current draw can end up being with 25-30% less. I've noticed that some low cost flashlight companies coat the reflectors poorly compared to reputable manufacturers or even mid-cost manufactures, always looking considerably less shiny than other.
I posted my reply and now there are 6 new ones between Kreisler's and my reply. Now I go to post again and there are 5 more! I can't keep up! (Too many distractions.)
Anyway, so my measurements of 2.5 to 2.6 amps at the tailcap with a AA NiMh or Alkaline are feasible? I'm also comparing to measurements other's have taken (like Cheap Thrills mentions above). I'm assuming these other's measurements are straight tailcap measurements and not adjustments after the actual readings.
If my measurements are real than I don't like this at all! Runtime would be horrible! Moto - if your read this, I want to know what your lights from "campingsurvivals" measure.
I understand the tailcap reading is not the same as the current to the emitter, but I still never expected to see currents this high from single AA's!
-Garry
yesterday i did 4 runtime measurements of the Tank E09 and the results are published yet in the review of the review thread. the torch, with fresh Eneloops, got burningly hot (amazing!! no exaggeration) so i wrapped it in a wet dishcloth. guys you dont need to be afraid to measure 14500 High-mode's (Sipik, Xeno, Romisen, P1A, ..): simply wrap the torch in a wet dishcloth.
great le heatsink!!
Glass is the most ignored part of a flashlight. Some glass will cut 0.8% and some 15% (extremes, measured by 2100). Reflectors matter too.
On-topic:
Others have measured 2Amps on the Sipik with NiHMs. Depending on the leads resistance and DMM accuracy, others might have measured lower currents.
Yeah, trying to compare other's tailcap currents is tough because it seems tough to know what a "true reading" is due to DMM accuracy, resistance in leads, etc. . . Many times you don't know what other's have measured with. (I don't mean to launch into that whole discussion of meters, leads, etc. . .!) Seems you're best to compare with what you know about your own tests and your own equipment and lights.
Seems to me you're more prone to measure too low than too high, so this is why I can't write off my measurements as equipment inaccuracy.
-Garry
Yeah, remember to divide that by around 3 and a bit to get current to the LED. I often get currents of over 2A from single AA budget lights/circuits.
It doesn't mean the LED is getting that much, we only assume that as li-ion voltage is roughly the same as LED vf so if we see 3A draw from a li-ion we just estimate 3A to the LED.
some known torches are direct driven with 14500's: Jetbeam PA10, Klarus P1A, Eagtac D-series.
i guess that this means that their tailcap amperage is equal to the LED amperage.
So 2.6A / 3 = 0.87A, sounds reasonable. But, The battery is still being drained by 2.6 amps, right? So 2.6 amp draw on a 2000mAh cell isn't going to last very long, right? Can I guesstimate 2000mAh (battery capacity) / 2600mAh (current draw) at 0.77hrs ~ 45minutes? I thought others were getting over 1hr 15mins. Perhaps on models not being driven as hard?
And the "divide that by around 3" - is that because the boost circuit is boosting by about a factor of 3?
-Garry
That is partly correct. I do not measure tailcap current, especially not on single 1.5V cell lights, that would give a huge error.
Instead I prefer to do this kind of graphs:
The problem with just measuring tailcap current is that you do not know what voltage the light gets, this depends on the battery and on the meter and probes and even 0.1 ohm is very significant at 1 to 1.5 volt and 2 ampere.
The red curve clearly indicates that amperage (measured at some point in the circuit) is neither a linear curve (a straight line with constant slope) nor constant. If it was constant, we could multiply it with a factor and then guesstimate either the brightness or the runtime.
In all of HKJ's red graphs we see that a regulated light doesnt have a constant current at any time, since with time the voltage decreases and amperage has to go up to maintain constant power ( P = U * I ) which is related to brightness (e.g. proportional relationship, P ~ brightness ).
My new DMM has a data cable for PC connection (serial/USB-port) and comes with PC software. I could collect continuous data of amperage over time and get the red graph. But that's about it.
Game over kreisler!! 
could you link the meter you got?