This whole argument about lumens brings me back to the suggestion I made a few years ago. Take a sample of all these lights into a cave with a lux box and meters to get a practical measure of light DELIVERED TO A TARGET rather than what is generated at the emitter or what comes out the front and is wasted as spill. I still don't see why lux measurement should be so difficult or so rarely reported.
The biggest problem is that Lux is MUCH more finnicky. Some mfg's are claiming 20,000Lux and others are measuring those same lights (under optimal conditions) at like 4,000Lux. I mean, it's a HUGE swing. Lux measurements need to be taken on the same machine under the same conditions for it to really matter and light manufacturers have not been willing to do that. I believe Bobby (of UWLD) requested manufacturers send in lights for him to measure on his machine so he could get across-the-board results from manufacturer-approved units. Not a single one took him up on it. With him being a competitor, I don't blame them.....but those results would've been VERY nice to see.
The problem with the lumens rating, though, is the same as the lux. Manufacturers will continue to lie and fudge numbers to achieve better results, or just flat-out lie. With the lights that have been mentioned so far (Big Blue, LX20, UTD 35, etc).....they're all using the exact same emitter so we know what performance limits can be and what they'll actually look like. I agree that test-diving them all is the only way to know for sure what that exact combo of emitter/driver/glass/reflector/battery/wiring/programming will do for you in your conditions.....but not everybody is able to test dive all of the lights they'd like.
Manufacturers lying brings me right back around to: Use some logic along with prior knowledge to weed out garbage results. I bought some cheap $25 chinese lights that claim 1800lm on an XML emitter. I know that those put out 1000lm optimistically. I got what I expected....I knew better than to expect 1800 TRUE lumens. The UTD light is just hilariously impossible at 3500lm, considering it uses the same emitter as my chinese light. The Dive Rite LX20 uses the same, as well. The Big Blue TL4000 (or whatever it's called) uses 4 of them, but the run time numbers sound like they're not fully driving those so no way they're hitting 4000lm.....but it should be more light than the UTD and Dive Rite combined.
Prior knowledge needed when evaluating dive lights: Most guys use Cree XML emitters. XML emitters have what's called "binning"...and it's usually designated by a dash-number: XML-T6, XML-U2, etc. You can look up those specs, but they optimistically put out 1000lm at the emitter. The lens reduces light output, so 1000lm at the emitter doesn't mean 1000lm into the water. Diving in water will drastically change the beam characteristics of your light. UWLD doesn't use Cree emitters, so his stuff is harder to judge...but he also uses conservative numbers. Lux measurements on different machines is almost guaranteed to be a smokescreen for poor performance because the numbers are impossible to reproduce (and therefore refute) and the numbers sound bigger: eg, 20,000lux sounds bigger than 700lm.
OP: If you're not completely and totally fed up with all of us by now....give the industry a month or so. I dove with some prototype lights that are releasing and they're fantastic. For the short burn times being discussed, a few of those and a few spare batteries seems like the best way to go, if you ask me.
(and they are on sale)
