H31w and H51w rated lumens vs real life

Can anyone answer this for me, please. Many people have said that when comparing two lights, you need to have significantly (up to twice as much) more lumens to notice a big difference. Why is it that when I compare the H31w and H51w on the fonarik website (http://fonarik.com/test/indexen.php), this is what I see:

H31w on high (189lum) is significantly brighter that H51w (172lum)
On the 103lum setting, the H31w is still brighter or very similar to the H51w on high.

I want to buy just one of the two, and use my light for night time running, hiking and rock climbing. The ability to see that little bit further (which the H31w most definitely does) is appealing to me. Why does everyone say that a few lumens here and there makes little difference yet many beamshots I see show an obvious difference in these lights?
Also, am I right in saying that if I ran a H51w on L91 lithiums, I can expect extended runtimes but not really an increase in brightness?

Thanks very much

Want to see something even more confusing? Go to the fonarik site and compare the H31w and SC51w side by side. The outputs appear to be close and more like what you might expect between the H31w and H51w based on specs. Then compare the SC51w and H51w side by side. The SC51w and H51w are supposed to be internally identical, but the pictures do not seem to agree and the SC51w appears to be much brighter. Maybe the particular H51w they tested was a dud. Maybe they used a bad battery. Maybe the camera exposure was different.

There's also the photo issue. Photos do a poor job of duplicating what the eye sees in a beam of light. Photos can magnify OR diminish differences of brightness (depending on the exposure chosen), and generally magnify differences of tint.

If a mfgr's published lumens are not agreeing with photos uploaded by a user, I tend to go with the published lumens as a more reliable guide. It isn't always, but more times than not, because of the errors that can creep into beam photography. The best comparative beamshots show two beams side-by-side in the same photo...if the exposure's correct, such a photo can do a decent (but not great) job of displaying relative brightness.

I came up with a way to use a DSLR as an ambient lumen meter and it has proven to be quite accurate (given a 1/3 stop meter granularity) with my various light's listed specs. My ZLH51w max measured the furtherest off spec, @ 125 lumens, or about equal to my D25A clicky and ThruNite T10 on Eneloops, which are spec'd at 122 and 115 lms, respectively. A few others have noticed the same on this thread:

http://www.candlepowerforums.com/vb...white-better-for-trail-running&highlight=h51w

Is it sample variation, or is the light simply just over-stated?

BTW - the "Needing twice the lumens to see a difference" is an old flashaholics tale....total bunk.

A ~ 5-10 % difference can be noticed by most people if they see lights side by side for example...and a 10-20% difference is noticeable for most people if they have to see the light at separate times, etc.

If you want to see what YOUR lumen resolution is, take a light with multiple modes, and go between them....you'll see a difference, and you can then compare the lumen changes that you saw to what you think is a reasonable resolution.

Of course, even if you can't tell if one light is brighter or not, you CAN note what's visible when a light is on, and if there are multiple objects in your view...some of them will be furthest to the sides, furthest away, and so forth. If light A sees things further away it has more throw whether you perceived it that way w/o a reference point or not...and an object may come into view with even a few more lumens, especially when the light has more throw to its pattern, etc. (A floody light takes more total lumens to "make a difference" as it spreads them out more, etc...)


Now, if the beam PATTERNS are different, its very hard for people to make an apples and apples comparison....as a higher lumen light with a floodier beam can produce an equal lux on target to a throwier beam pattern, but over a larger area...but, we tend to focus on "the hot spot", which fools us into evaluating the throwier pattern as "brighter".

That means that if there are two different lights, NO particular ratio of lumens is going to cause, or not cause, a perceived change in "brightness".

People who look at the beam from my 131 lumen LED maglight typically say it looks brighter than the beam from my 750 lumen ZL SC600. Their eye goes to the searing yet very small hot spot...and "takes a brightness reading" from that little spot...and says :"its very very bright".


So any comparison of lumen output and perceived brightness is only valid for a particular light...and ITS output.

IE: If I ramp my SC600 up from ~ 500 L to 750 L (~ 33% increase in lumens), its VERY obvious that the light got brighter...even though its only 33% brighter rather than twice as bright.

Add to all that the above explanations of why a camera is not as good as an eye at showing what your eye would see...and you have a real dilemma trying to use lumens alone, or a picture alone, or lux at 1 meter alone, as a comparison of "How bright" a light is...relative to what YOU want to see.

TEEJ said:

BTW - the "Needing twice the lumens to see a difference" is an old flashaholics tale....total bunk.

Click to expand...

Well...as written I agree, but usually the statement is contextualized like this: "100% increase before the beam looks significantly brighter." And here "significant' doesn't mean "detectable," it means "enough brighter that it's unambiguously brighter," often thought to appear about one-quarter brighter to the eye. Or for me that means, "enough brighter that I should maybe consider an upgrade."

TEEJ said:

A ~ 5-10 % difference can be noticed by most people if they see lights side by side for example...and a 10-20% difference is noticeable for most people if they have to see the light at separate times, etc.

Click to expand...

Our forum's perceptual physiologist (Bowzer) has mentioned several times on this forum that 10% is a common threshold for many perceptual detectable differences (weight, light) in a controlled laboratory situation, with the stimuli presented simultaneously or in quick succession. I find 20% difference to be slightly noticeable under controlled sequential situations.

So this argument generally devolves into (1) what's the smallest perceptual difference can be seen under ideal conditions (10-20%) and (2) what's a usably brighter light in a real-world situation, an upgrade-worthy number (often given around 100%, but very subjective).

In a nutshell: No way would I upgrade my 200 lumen light to 240 (20%), just not worth it as I'd barely be able to detect a difference under ideal conditions, let alone real-world conditions.

Upgrade from 200lm to 300lm, a 50% increase? Marginal, if I have money burning a hole in my pocket and I want bragging rights. From 200 lm to 400 lm, a 100% increase, which will probably appear a quarter brighter to my eye? Yes please, where do I pay, here's my money. But that's just me.

(Remember "twice as bright" is generally given as a 300-400% increase, or in the case of a 200lm light, 800 to 1000lm)

I agree with TEEJ's methodology of how to decide for yourself...that's good advice. And also his caution not to compare lumens between spot and flood, good advice also.

Well ALL of that has been very useful, thank you everyone . Referring to beam spread and throw I must admit that I thought the H51w and H31w comparison was a fair one, given their same emitters, reflectors, tint etc. Definitely seems a far fairer comparison to see two lights in the same image, and better still in real life. It is interesting the mention of the H51w being quite a way off spec for several users. Many thanks again!