Law of Diminishing Returns RE: Lumens

I recently purchased my first significant LED flashlight an Olight SR51. Coming from an electronics/amateur radio back ground my question is at what point is increasing lumens output insignificant. For example: In amateur radio the law of dimishing returns would say that going from 100 watts to 1000 watts is the most significant increase of power output which would be 10 dB...going from 1000 to 2000 watts output is only a 3 dB increase. Generally speaking to move the receivers S meter (dBm) one S unit requires approximately a 6 dB increase, furthermore a 2 dB increase is barely perceptible.

Having come from an EDC (every day carry) of a streamlight TL-3 to the SR51 the difference is overwhelming. Roughly speaking 200 to 900 lumens. Is going from 900 lumens (SR51) to 2200 lumens (SR90 intimidator) just as significant? At what point does increasing the lumen output become insignificant other than quote "bragging" rights. If I'm standing say 100 yards away from the light source how perceptible of a difference is 900 vs 2200 lumens? I suspect that being on the receiving end the difference is more noticable than perception/viewing from the source. In addition I suspect the type of LED and reflector play a significant role in that perception.

Oh...the Olight SR51 is just awesome!

Thanks
Brad
South of Pittsburgh
First Post

Re: Law of Dimishing Returns RE: Lumens

:welcome:Brad

Good to have a fellow Amateur here

Norm

Re: Law of Dimishing Returns RE: Lumens

I think there will be two aspects to your question, one being the physics of it and one being the biology.

With the physics of this there will be factors including losses, scatter and the distance of observation etc.

The biological one, and the eye's non-linear perception of brightness, and of course the viewing conditions at the time, will give another interpretation of the significance of this increase in output.

From my own experience I would say that the 6600Lm HID was about the same perceived jump in output from my 760Lm Fenix TK45, as the TK45 was compared to a 200lm Cree R2 light. I don't have a 2200Lm light, so can't say for definite, but it seemed a similar jump in output to me.

Re: Law of Dimishing Returns RE: Lumens

I have read many comments on the perception of light, but no scientific paper. Some people say the perception is logarithmic, others say it is proportional to the square root of the increase in brightness.
The important aspect is, the perception is also dependent on the brightness (measured in lux or candela) and not the overall amount (Lumen). The difference in brightness between the different reflectors is similar to the difference between a normal antenna and a directional (parabolic) antenna.
You should have a look at selfbuilts 100 yard beamshots and other beamshot comparisons in the reviews subforum. Goinggear has published a video comparison of more than 60 lights on Youtube. By watching this pictures and videos You will get an idea of the different reflector and beam characteristics.

Re: Law of Dimishing Returns RE: Lumens

Erzengel said:

I have read many comments on the perception of light, but no scientific paper. Some people say the perception is logarithmic, others say it is proportional to the square root of the increase in brightness.
The important aspect is, the perception is also dependent on the brightness (measured in lux or candela) and not the overall amount (Lumen). The difference in brightness between the different reflectors is similar to the difference between a normal antenna and a directional (parabolic) antenna.
You should have a look at selfbuilts 100 yard beamshots and other beamshot comparisons in the reviews subforum. Goinggear has published a video comparison of more than 60 lights on Youtube. By watching this pictures and videos You will get an idea of the different reflector and beam characteristics.

Click to expand...

I appreciate you pointing me in the right direction regarding lumens/lux/candela. Having performed some quick calculations (all things being equal and lux at 1 meter) the difference between 900 lumens and 2200 lumens at 100 yards is significant! A couple of YouTube videos comparing various throwers confirm my mathematical computations. Time for something to drink it's too early here to scratch grey matter that deep! :thinking::thinking:

Re: Law of Dimishing Returns RE: Lumens

Erzengel said:

I have read many comments on the perception of light, but no scientific paper. Some people say the perception is logarithmic, others say it is proportional to the square root of the increase in brightness.

Click to expand...

Here you go - I discussed this at some length in my Sunwayman V10A review, post #3 (complete with links to actual scientific papers).

Simply put, the logarithmic method of linearizing our perceived brightness dates back to Victorian-era science (although still forms the basis of camera f-stops and some astronomy conversions).

The square root method would be appropriate if we were talking about a point source of light. But for a flashlight beam (which corresponds to >5 degree arc), the most accurate perceived brightness power relationship is a cube root. One of the papers I cite actually reports a table of all the measured power exponents from different studies up to that time, and 0.33 is basically both the mean and the median.

That said, there are issues with how perceived brightness is estimated for Stevens' power law calculations (as discussed in the wiki links). The point would be to say that cube root is the best estimate we have for perceived brightness.

The important aspect is, the perception is also dependent on the brightness (measured in lux or candela) and not the overall amount (Lumen). The difference in brightness between the different reflectors is similar to the difference between a normal antenna and a directional (parabolic) antenna.

Click to expand...

Yes indeed! The perceived brightness calculations above all assume a uniform testing target. In practical terms, our subjective experience is heavily influenced by relative beam pattern and distance (plus effect of surrounding structures, etc).

This does show up somewhat in my 100-yard shots (although I can tell you is even more noticeable in real life). Simply put, at close range, "floody" lights seem brighter than "throwy" lights, even if they are outputing the same lumens. The reverse is true at a distance. This is why you need to take both overall output (as measured in lumens) and throw (ideally as measured in "beam distance", since this is normalized by an inverse square for distance).

In the case of the SR51 vs SR92, that can be demonstrated in the outdoor shots and summary tables from my reviews of those lights:

Re: Law of Dimishing Returns RE: Lumens

Diminishing returns regarding lumens . . . Only when a light becomes bright enough to actually cause permanent eye damage. Be sure to wear some shades.

Re: Law of Dimishing Returns RE: Lumens

Recently I've been thinking along the same lines as the original post. Just about any modern flashlight can do well over 100 lumens. It is becoming more important to have small size for EDC, simple user interface, decent color rendition, useful beam shape, and long run time. At the 700-1000 lumen level you are reaching white-out for nearby objects and complete loss of night vision. I confess I love a "wow" light. However, for actual use I usually prefer floody neutral-warm tint lights on low or medium output. Throwy lights are good for use with firearms, they put the light where it is needed for target ID and sighting, and reduce the amount of backscatter from the spill beam. Bicycling requires both throw and significant flood. Spotting things at a distance also calls for strong throw. So it's a case of "the right tool for the right job."

I still want a light so bright it has a recoil. I guess that means I belong here.
--Max_Power

Re: Law of Dimishing Returns RE: Lumens

cube root method . does that mean 3.5 times the lumens is about 50 % brighter ? so go from 100 lumens to 350 lumens is circa1.5 times the brightness ? seeing as how when i use my fingers and toes to count the cube root of 3.5 is about 1.5

+1 on selfbuilts lumens vs intensity peice. it can vary depending on factors like whats in the beam and your awareness level of what you are looking at.

cheers

Re: Law of Dimishing Returns RE: Lumens

For me the Law of Diminishing Returns applies more specifically to application and physical size. I want a flashlight that I can easily carry, an EDC if you like, that has sufficient output to get the job done: Having several light levels makes for a much more useful tool. Being able to put a flashlight in your pocket makes a lot of sense to me and with the introduction oh the XM-L emitter, EDC Flashlights have more than enough light for most tasks. I particularly like being able to use 18650 cells and many pocket-friendly lights offer this feature. YMMV.

It's a great time to be a Flashaholic. :welcome:

Re: Law of Dimishing Returns RE: Lumens

It's hard to tell the lumens that a light puts out, since the beam focus (lux) adds an extra dimension.

If a light with all factors being invariable except for the one variable; lumens, then it isn't too hard to tell the difference between 600 and 800, 900 or 2000, 100 or 200. It's just hard when comparing a floody 300 to a throwy 200.

Gotta make a custom light that can multi-mode from 0 to 1000 in even intervals.

--

I find that on a dark night I don't need more than 50-150 lumens for any task. Cranking up the output past that just reduces battery life and doesn't add much value to the 5 to 35 yard range that I require the light for.

Re: Law of Dimishing Returns RE: Lumens

Max_Power said:

Recently I've been thinking along the same lines as the original post. Just about any modern flashlight can do well over 100 lumens. It is becoming more important to have small size for EDC, simple user interface, decent color rendition, useful beam shape, and long run time. At the 700-1000 lumen level you are reaching white-out for nearby objects and complete loss of night vision. I confess I love a "wow" light. However, for actual use I usually prefer floody neutral-warm tint lights on low or medium output. Throwy lights are good for use with firearms, they put the light where it is needed for target ID and sighting, and reduce the amount of backscatter from the spill beam. Bicycling requires both throw and significant flood. Spotting things at a distance also calls for strong throw. So it's a case of "the right tool for the right job."

I still want a light so bright it has a recoil.

I guess that means I belong here.
--Max_Power

Click to expand...

Good points.

Re: Law of Dimishing Returns RE: Lumens

My totally unscientific experiments showed that the consensus that you need about 30% more brightness to actually see a difference (with the same beam profile) sounds about right and this relation seems to need a higher percentage difference as you go higher in the lumen range. For example, you turn on a light that has 100 lumens, turn it off and quickly turn on a light that has a 130 lumen output, you will be able to tell it's a little brighter, but probably not enough to justify spending 50$ more. If you try a light that has 200 lumens and one that has 260 lumens, the difference seems harder to perceive.

The second part of my observations is that when you double the output (always with the same beam profile) the perceived difference is more and more impressive with brighter lights. As an example, I've often played with 2 100 lumens lights, 2 240 lumen lights and doubling the output made a difference, but I would not have justified carrying both all the time. When I turn on my 2 predators (500 lumens) I am impressed, but no more. The one time I played with my TK70 and my cousin's, I was already impressed with the one TK70, but turning both on at the same time made me drool, feel weak and I suddenly wanted a light THAT bright. When I tried just one of them afterwards, it seemed a little dim even after 10 minutes of walking in the dark.

Re: Law of Dimishing Returns RE: Lumens

Cataract said:

My totally unscientific experiments showed that the consensus that you need about 30% more brightness to actually see a difference (with the same beam profile) . . . .

Click to expand...

I agree

Re: Law of Dimishing Returns RE: Lumens

Very interesting thread here !


lovecpf




Allow me to add one additional point to consider . . . .


Once the Light Beam is bright enough to cause your pupils (iris) to "stop-down",
then any additional brightness is simply, pretty much Wasted.


:candle:
_

Re: Law of Dimishing Returns RE: Lumens

Burgess said:

Very interesting thread here !


lovecpf




Allow me to add one additional point to consider . . . .


Once the Light Beam is bright enough to cause your pupils (iris) to "stop-down",
then any additional brightness is simply, pretty much Wasted.


:candle:
_

Click to expand...

I've definitely noticed that with my XM-L floodmonster; If after using it on low (about 60 lumens i think) around the house i turn on the floodmonster to max, it's much much brighter for about 20 seconds and then it's just as bright, because my irises contract so much due to the amount of light.

of course, the extra brightness is still very much noticeable when using it to light the whole room in ceiling-bounce