Talk:Lumen (unit)

Needs clarification or at least referencing

A standard 100 watt incandescent light bulb emits approximately 1700 lumens in North America and around 1300 lumens in 220 V areas of the world. See luminous efficacy for the specific efficiency of various types of electric light sources.

At first glance, it makes absolutely no sense as to why a 100 watt bulb would be 1700 lumens in North America and why a 100 watts elsewhere would be any different. Voltage as the only controlled parameter has absolutely no affect on lumens per watt!

Lumens per watt is affected by filament temperature and envelope transmissiveness, but neither of those are controled by geographic location or voltage.

Thanks! -Jesse —Preceding unsigned comment added by 64.146.180.232 (talk) 18:32, 4 December 2007 (UTC)

The voltage is not the only controlled parameter. The thickness of the filament is another. Light bulbs designed for 220 V are designed differently than those made for 110 V. Be aware also that light bulbs are nonlinear devices. The resistance of the filament varies with voltage.--Srleffler (talk) 00:10, 5 December 2007 (UTC)

I realize that voltage is not the only controlled parameter -- but it's the only one that the article mentions as being the cause of less efficient lumens/watt. I also realize that incandescent light bulbs are non-linear, although it would be slightly more exacting to say that the resistance of the filament is affected by temperature, which is affected by wattage and dissipation, and wattage is affected by resistance and voltage, and so on. The only controlling parameters I can think of which control lumens/watt would be envelope transparency, filament surface temperature, and filament material, and none of those are tied only to voltage or operating in other parts of the world. There is no reason that an equal surface temperature, filament composition, and bulb transparency could not be done on both 110v and 220v lamps.

Is the article saying that 220v filaments are designed to run cooler, thus putting out more of their total energy in non-visible regions? or is it saying that they are running so hot that the same happens?

I'm not saying that the average 220v bulb is or is not the same efficiency as the average 110v bulb -- I do not know -- but I guess I'm saying that the article is dreadfully ambiguous, and it makes a statement that doesn't even make sense, and it doesn't give a cause for the phenomenon, nor does it cite any references. I'd say it is closer to "whimsical banter" and doesn't even rise to the level of "Original Research."

Does anyone mind if I remove that statement in a week or two if nobody finds some references? Thanks & Keep up the good work. -Jesse —Preceding unsigned comment added by 64.146.180.232 (talk) 03:30, 6 December 2007 (UTC)

I rephrased it to remove the impression that we are talking about the same bulb being run on two different voltages. It would be helpful to add a citation to support the two values. Removing the statement would not be appropriate unless you have strong reason to think it is untrue. Note that all of the discussion above is original research, and cannot be used to justify changing the article. Even if we obtained an explanation for why the efficiencies are different, there would be no reason to add that to the article. It's beyond the scope. The bulb outputs are given as a practical example of lumens. They don't need to be explained here. --Srleffler (talk) 06:26, 6 December 2007 (UTC)

I suspect the solution to this puzzle has to do with bulb lifetime. From some searching I did online, it appears that bulb lifetime changes very dramatically with voltage (proportional to the twelfth power).[1] They probably have to sacrifice efficiency to get decent bulb lifetime at the higher voltage.--Srleffler (talk) 06:40, 6 December 2007 (UTC)

Removing (or moving to talk pages) is quite appropriate for material for which there is no reliable source. See [2] (which at the time of my writing) says: "Any material that is challenged or likely to be challenged must be supported by a reliable source. "Original research" is a claim for which no reliable source can be found. Producing a reliable published source that advances the same claim taken in context is the only way to disprove an assertion that a claim constitutes original research. If there is a source, but the source or claim is disputed, that is not original research but rather of a question of reliable sourcing or undue weight. However, using information from references out-of-context or to forward claims not directly supported by the sources is original research."

I don't know who made the claim that 220v bulbs produce less lumens per watt then 110v, but I don't take it to be true. I know that it's certainly not always true (for example, the most efficient incandescent 220v bulb is better then the least efficient 110v. The reverse is also true.) WP Official Policy is that the only way to counter the claim of original research is to provide reliable sources. I don't know who made the claim, and I don't know if they were correct -- maybe they were, maybe they weren't. But they provided no references, so an average reader has no way of knowing. By the way, in my comments here, I am challenging the afore mentioned statement as being "Original Research."

I would propose that the comparison of incandescent bulb efficiencies be replaced with a simple statement like "The ability of lightbulbs to produce visible light is measured in lumens. Some types of lamps, like incandescent, tend to produce more of their output energy in non-visable regions and less in the visible, and so produce less lumens per watt, then, lets say, a flourescent or sodium vapor lamp, which tend produce more visible light per watt." (If lamp efficiency even needs to be discussed here -- since I think lamps have their own section..) Can anyone give me a single reason that removing an unreferenced statement of questionable accuracy would be somehow not appropriate? Having that unreferenced OR statement there is inappropriate, according to WP Policy.. The statement is not always true, and I'm not convinced that it's even "usually true."

By the way, I forgot to ask before, but is this article someone's pet project? If so I understand how you could feel intruded upon by my wishing to effect WP Policy compliance and meaningfulness to this article. If it's someone's pet project, just say so, and I'll go away and not bother you any more! I wouldn't want someone else trying to edit my personal webpages..! Otherwise, there is A) No reason whatsoever that a disputed unreferenced statement should exist on WP, and B) No reason whatsoever that said disputed unreferenced statement shouldn't be removed. Thanks and keep up the good work! -Jesse

PS:The fact that our conversation is original research is besides the point. The point is that the material in question is original research, and according to WP policy should not even exist to begin with (unless with valid references) -- you make it sound as if unreferenced statements can only be removed if they are proved wrong using valid references. But that of course is exactly the opposite of WP Policy which says they may be removed unless proved TRUE with valid references. -Jesse —Preceding unsigned comment added by 64.146.180.232 (talk) 05:31, 7 December 2007 (UTC)

I completely missed this reply (and your subsequent edit to the article). You are quite right that unsourced material can in general be deleted from the article. However, deleting material that is not really in doubt or that can be easily sourced might be seen as disruptive editing. The reason I brought up the fact that the conversation above involves original research was to point out that argument about why a 230 V bulb might be less efficient than a 120 V bulb is completely irrelevant to whether that "fact" should be in a Wikipedia article. Either the claim can be sourced or it can't. Argument about why it might or might not be true is interesting, but irrelevant.--Srleffler (talk) 04:08, 27 May 2008 (UTC)

Just for fun, I looked up some 230 V bulbs. Philips has a 100 W 230 V bulb that puts out 1380 lm. My favorite local industrial supply company, McMaster-Carr, stocks 3 standard "general purpose" 100 W bulbs: a 230 V model that produces 1270 lm, rated for 1000 hours, a 120 V model that produces 1600 lm, rated for 750 hours, and a long-life 120 V model that produces 1470 lm, rated for 1500 hours. A survey of 120 V bulbs' package statistics[3] gives output ranging from 1585 to 1750 for standard bulbs. So, a quick survey seems to support the statement that standard 230 V bulbs are less efficient than standard 120 V models.

Having done some reading and thinking on this since we last talked, I've learned that one can produce bulbs that are much more or much less efficient, at any given voltage. There is a tradeoff, though, between efficiency and bulb life. Standard ("general service") 100 W light bulbs are designed for bulb life in the 750–1000 hour range. Designing a bulb to run for this length of time at 230 V requires a cooler filament temperature, and therefore less efficiency. One can design a bulb to be more efficient but last for a shorter time, but that wouldn't be a "general service" (standard) bulb. Projector bulbs take that approach. Alternatively, one can design a bulb for longer life but at the expense of reduced efficiency, as in the example of the 1500-hour "long-life" bulb above. At any given rated lifespan, though, a 230 V bulb will be less efficient than a 120 V one with the same power rating.--Srleffler (talk) 05:31, 27 May 2008 (UTC)

This issue about bulb construction is pointless in an article about lumens and in my opinion it should be removed. The original sentence was inserted as reference for a luminous flux from a typical standard bulb. Rather than a US-EU standard bulb competition, it would be more useful adding also halogen and fluorescent lamps data. BTW these standard bulbs burns all at 2700°K and as you can see the us version has higher luminosity than eu, but shorter life. There are no free-lunches or 120v miracles. -- Basilicofresco (msg) 14:05, 2 January 2009 (UTC)

The article doesn't discuss bulb construction, and doesn't seek to compare US and EU bulbs. Rather, the article illustrates how much light a lumen is by telling the readers how many lumens are produced by a common light source with which they will be familiar. Because it is so common in the home, the incandescent lamp is the obvious example to provide. Since European bulbs don't produce the same light output as North American ones, the article needs to give both values so it meets the needs of readers both in Europe and in North America. The article does not explain why European bulbs are dimmer than North American ones, nor should it. If we were going to make a change, I would recommend removing the less-familiar sodium lamp, and replacing it with compact fluorescents, since these are rapidly replacing incandescent bulbs. I would still keep the incandescent bulb values, though, since this is a very familiar light source.

I'm not sure what you thought those two datasheets prove. The numbers agree well with those in the article and cited by me above. The 130 V bulb is brighter and more efficient than the 240 V model. Note also the long-life 120 V bulb I mention above, which is rated for 1500 hours with an output of 1470 lm—still brighter than the 1000-hour 230 V bulb you mention. --Srleffler (talk) 17:53, 2 January 2009 (UTC)

I moved the 230 V case out of parentheses. The article should not presume that the reader is in a place with a 120 V grid. The 120 V and 230 V cases are of equal importance. I removed the halogen lamp because I don't think we need three examples and because consumer halogen lamps vary a lot more in output than the others. Where I happen to live, the commonly available medium-base halogens are not significantly more efficient than a typical incandescent bulb, but they have long bulb life. The purpose here is to provide common examples, such that a reader can look at a light bulb in their own home, not think too much about the technical details, and get a feel for what a lumen is (or what 1500 lm is). I rephrased the CFL example to make the emitted lumens clearer, and added references for 120 V CFLs.--Srleffler (talk) 19:14, 4 January 2009 (UTC)

Just to clarify the original point. A 120 volt 100 watt incandescent lamp does give more light than its 220/240 volt counterpart. This comes about because the temperature at which the filament is run determines the efficiency of the lamp - the higher the temperature the better the efficiency (more Lumens per watt). The temperature is determined by a number of compromises but generally, the temperature is aimed to give a 1000 hour life on average. The life is determined largely by the surface area of the filament, as it is from here that the tungsten evaporates. A 120 volt bulb requires a filament with one quarter of the resistance of a 240 volt bulb (half the voltage time twice the current for the same wattage). As such the filament is half as long and twice the cross sectional area giving one quarter of the resistance. But the resultant 120 volt filament has a smaller surface area than the 240 volt filament and can thus be run hotter for the same life giving a 30% gain in efficiency.

This principle is exploited in 12 volt low voltage halogen lighting where the very thick filaments coupled with the halogen cycle allows the filament to be run even hotter giving nearly a 70% gain in efficiency compared with a 240 volt lamp and the lamp still has a 4000 hour life. 20.133.0.13 (talk) 13:52, 9 September 2009 (UTC)

Differences between lumens and lux: This section has some poor grammar and is quite confusing to the reader. —Preceding unsigned comment added by 211.30.168.79 (talk) 12:43, 19 July 2009 (UTC)

More clarification

What is the purpose behind this strange equivalency string?

1 lm = 1 cd·sr = 1 lx·m2

How is it useful or informative? What is the meaning of "cd x sr" or "lux x m2"? Wouldn't it be more useful to say that the lumen is defined by reference to a source with intensity of 1 cd in a given direction, where 1 cd = 1 lm/sr? An isotropic 1 cd source then emits 4pi lumens... Cluginbuhl (talk) 02:12, 21 August 2009 (UTC)

It's usual for articles on SI units to provide the equation for the unit in terms of SI base units, in the introduction. The expression is there for those who need it. If it doesn't help you then you aren't the target audience for that equation. :) The article does go on to say that this means a light source that uniformly radiates one candela in all directions radiates a total of 4π lumens.

One can approach defining the lumen the way you suggest, but it isn't very clear unless the reader has a good understanding of solid angles already. A source that emits with an intensity of 1 cd into a solid angle of 1 sr emits a total of 1 lm. You have to specify the solid angle that is illuminated by the source for the definition to be correct. Isotropic sources are convenient because it's easier to visualize an isotropic source than a source that emits only into a solid angle of 1 sr.--Srleffler (talk) 02:49, 21 August 2009 (UTC)

The whole detailed discussion about lumen output of 120V vs 230V lamps is off-topic for this page defining and discussing lumen. Take it to a luminous efficacy page, if there is one. Here just give an example like a "typical" candle at about 10 lm, a 100W bulb at about 1500 lm, a streetlight at 20,000-50,000 lm. That's it. Don't get distracted. Cluginbuhl (talk) 02:20, 21 August 2009 (UTC)

That's pretty much how it started out. The article initially gave the lumen output of a typical 100 W bulb, because that is a light source that would be familiar to most readers. This had to be adjusted because European 100 W bulbs are dimmer than American ones. I'll see if it can be simplified a bit.--Srleffler (talk) 02:53, 21 August 2009 (UTC)

Not correct methinks.

That is, a light source that uniformly radiates one candela in all directions radiates a total of 4π lumens. If the source were partially covered by an ideal absorbing hemisphere, that system would radiate half as much luminous flux—only 2π lumens.

I don't believe that this is correct. If the hemisphere is absorbing the light over half the sphere over which it radiates, then the light flux over the remaining half must be the same at 4π lumens. If the hemisphere redirected the light into the system forcing it to combine with the unshielded half, then it would become 8π lumens, but the source would remain a 1 candela source. 20.133.0.13 (talk) 13:57, 9 September 2009 (UTC)

It is correct. Candelas are a measure of intensity, but lumens are a measure of the total flux emitted. If you block half the emitting area with an absorber, the amount of lumens emitted are cut in half, but the intensity in candelas remains the same over the unblocked portion of the source. (The intensity is of course zero where the source is blocked.)

If you used a reflecting hemisphere instead of an absorbing one, the intensity would double over the unblocked half of the source, but the total lumens emitted would remain the same. (This assumes, of course, that the source does not absorb any of the reflected light, and that the reflected light is uniformly distributed. In practice, the source would have to be much smaller than the diameter of the hemisphere for this to work.)--Srleffler (talk) 16:27, 9 September 2009 (UTC)

Thanks for the responses. I had a think about it and it did make sense after all. I wonder if the article requires some clarrification? 20.133.0.13 (talk) 12:41, 10 September 2009 (UTC)

Plural

As the plural in English language is usually adapted from the language the word comes from e.g. radius -> radii, I changed that. Hope you're fine with that, otherwise please discuss here. example of a lighting firm's homepageThanks, Saippuakauppias ⇄ 23:48, 28 October 2009 (UTC)

"Lumens" is correct, at least in American English.^* If you want to assert that "lumina" is the correct plural in English somewhere else, you are going to have to provide a reference to support your claim. I believe that "lumens" is the correct English plural everywhere.

*Ambler Thompson & Barry N. Taylor (2008). "NIST Special Publication 811: Guide for the Use of the International System of Units (SI)" (PDF). National Institute of Standards and Technology: 31. {{cite journal}}: Cite journal requires |journal= (help)

--Srleffler (talk) 03:10, 29 October 2009 (UTC)

NB: Dictionaries are not much help, because the unrelated biological term lumen does have plural "lumina". As a result, many dictionaries list both "lumens" and "lumina" as plurals for "lumen"; the distinct meanings of "lumen" have different plurals.--Srleffler (talk) 03:35, 29 October 2009 (UTC)

I reverted you again today, because you still have not provided a reliable source to support your claim that "lumina" is a valid English plural for "lumen", in the context of the unit of luminous flux. Wikipedia's verifiability policy allows uncited and disputed material to be removed by any editor. Such material must not be re-inserted unless a citation to a reliable source is provided.--Srleffler (talk) 00:11, 30 October 2009 (UTC)

The word "lumen" is from Latin, meaning "light". Therefore latin plural applies. Now your NIST dpcument doesn't provide any evidence for anything as far as I can see, as there are no plurals mentioned. Provide a valid source for "lumens" first please. Otherwise: "disputed material to be removed by any editor. Such material must not be re-inserted unless a citation to a reliable source is provided." --Saippuakauppias ⇄ 02:38, 31 October 2009 (UTC)

Since the plural "lumens" is regular in English, it never gets mention. If you're in doubt, compare searches for "lumens per watt" versus "lumina per watt" or something of that sort. There is nothing in the literature to suggest that lumina is a plural of a unit of luminous flux, so you probably ought to drop it. Dicklyon (talk) 05:25, 31 October 2009 (UTC)

It doesn't matter what the plural is in Latin, only what the plural is in English. The NIST document states on the page cited that the plurals are formed regularly, except for lux, hertz, and siemens which don't change in the plural. --Srleffler (talk) 05:54, 31 October 2009 (UTC)

Photons per second?

Shouldn't Lumen be expressible in terms of photons per second (of a given wavelength)? Obviously the eye's spectral response and the energy per photon will make this depend on wavelength, but still, an order-of-magnitude number would be useful. According to Luminous efficacy, 100% efficiency would be 683 lm/W, so 683 lm = 1 J/s. Taking a value near the peak spectral response, λ=550 , gives 3.61e−19 J/photon by the Planck relation. Dividing by that gives

$683\,\mathrm {lm} =1\,\mathrm {J} /\mathrm {s} =2.77\times 10^{18}\mathrm {photons} /\mathrm {s}$

$1\,\mathrm {lm} =4.05\times 10^{15}\mathrm {photons} /\mathrm {s}$

or

$2.47\times 10^{-16}\,\mathrm {lm} =1\mathrm {photon} /\mathrm {s}$

Again, this is just for 550 nm green light. Does that sound right? —Ben FrantzDale (talk) 13:01, 13 September 2010 (UTC)

It's misleading to talk about expressing lumens "in terms of photons per second". It implies a general relation where there is none. Obviously if you know the spectrum of a source you can calculate one from the other. When you do this you are not "expressing" one in terms of the other. Rather, you are describing the source in two different ways.--Srleffler (talk) 04:22, 14 September 2010 (UTC)

Sure. My motivation is to try to get a ballpark sense of photon-limited shot noise, at least to get some bounds on that. Since I rarely work with lumens, I'd like a little confirmation that the above calculation is about right :-). —Ben FrantzDale (talk) 11:35, 14 September 2010 (UTC)

Example Table?

As a reader of this page, I didn't find the thing I was looking for: a table relating incandescent bulb wattage to lumens. —Preceding unsigned comment added by 192.132.94.130 (talk) 17:10, 11 January 2011 (UTC)

You want Incandescent light bulb#Efficiency comparisons.--Srleffler (talk) 03:30, 12 January 2011 (UTC)

Kitchen example

I removed the mostly-redundant kitchen light example, for two reasons: first, the flux given seemed a bit high for an actual fixture. Four 4' T8 fluorescent tubes output about 10 klx. More importantly, putting the same fixture in a larger room does not uniformly reduce the illuminance. Assuming the same ceiling height, the illuminance right under the fixture might be almost the same but the illuminance near the walls of the larger room would be lower. One could think about average illuminance, but that would be pointless. You don't measure the average illuminance across a whole room, nor is that what is important for use. --Srleffler (talk) 17:26, 23 June 2011 (UTC)

NITS

This is an article about lumen. There should be a related link, or a small paragraph about the comparison to nits too!