Lighting Measurements - Reality Check

One thing that I wanted to mention is calculated light VS light delivered to the actual surface. I see a lot of recommendations and numbers thrown around, but not a lot to back them up.

My concern is that some of the advice has not been vetted by real world experience or any type of measurement.

View media item 36871
Photo of a work surface with a light mounted directly above it at a height of 36 inches. The light was mounted 24" away from a white wall.

730 LUX works out to 67.8 lumens per sq foot.

View media item 36877
The light I tested.

View media item 36873
Working height of the light.

View media item 36882
Another set of lights with similar output levels.


Far from the magical 100 lm/ft sq. This type of light fixture has often been recommended as a light for ceiling mounting. One of the main reasons it was recommended seems to be that its a wrap around and has a diffuser.

Here is my take, if the light cant hit 70 lumens 3' directly under the very center of it, how in the world is it ever going to get to 100 lm . / ft sq mounted 10' up?

I ran this test on three versions of lights wrap lights with diffusers.

Same basic result: None of the recommended lights were past 67 lumens / sq ft. at 3'

At six feet the I got about 240 lux (22 lumens / sq ft).

Three feet from the lamp, that level of lighting was pretty good for doing assembly work.

I also measured a three bulb, T8 Troffer. Standard industrial fixture mounted at 9'. 6' from the fixture it had an output of 68 lumens /sq ft.

Major Points:

1: 70 lumens / sq ft is pretty bright ( 760 LUX).
2: A lot of the recommendations are really, really optimistic with regards to lighting efficiency.
3: The eye is not very sensitive to changes in light intensity. Going from 300 LUX to 180 LUX is not very noticeable.

More BS:

The task light on my desk puts out about 2080 lux /193 lumens / sq ft (meter is set at 10x) 18" from the work surface and is very good for detail work.

View media item 36878
Good lighting is a combination of sources and work surfaces.

It seems that the thing to infer from what you are saying, is that you need general illumination of a certain level, plus task lighting.

bczygan said:

It seems that the thing to infer from what you are saying, is that you need general illumination of a certain level, plus task lighting.

Click to expand...

Agreed. That and hitting the often quoted light levels for general lighting are pretty optimistic. Task lighting is a lot more efficient use of resources.

I will have some more measurements of my own shop lights this afternoon. First, I need to get the shop a bit warmer. 56°F is a little bit cool and will affect the output of the lights.

Four Bulb T8 from Home Depot. Light is mounted at 8'


View media item 36888
The first group of measurements were taken directly under the light at varying distances.

Test Conditions:

Night ( no outside light)
All other fixtures were off
Ambient temp: 62°F
Fixture warmed up for 30 minutes.
Fixture: Home Depot utility light Model # 1284GRD RE retrofitted with a Phillips 1.18 factor ballast (PN IOPA-4P32-HL ). http://www.homedepot.com/p/Lithonia-Lighting-4-Light-Grey-Heavy-Duty-Shoplight-1284GRD-RE/202968125#

Dead center:
- Distance from the bottom of the fixture-
---Distance-- Lux----FT candles
---- 12",----- 4100,---- 380
---- 24",----- 1600,---- 148
---- 36" ,---- 1130,---- 104
---- 48" ,---- 710,---- 65
---- 72" ,---- 360,---- 33

Next Test- Light Drop off from the end of the Fixture (taken six feet away from the bottom of the fixture).



Moving parallel to the floor, away from the end of the fixture:

2' away = 370 LUX, 34 foot candles
3' away = 270 LUX, 25 foot candles
4' away = 190 LUX, 17 foot candles

Light drops of pretty quickly at the ends of the fixture.

Sides of the fixture ( measured from the center of the fixture, same 6' distance from the bottom of the fixture):

3' away = 520 LUX, 48 foot candles
4' away = 360 LUX, 33 foot candles
5' away = 270 LUX, 25 foot candles

As predicted, the side light does not drop off as fast.

The drop-offs look extreme but are as noticeable to the eye.

More to be added----

Explanation of different types of reflections. To be added later.

bczygan said:

It seems that the thing to infer from what you are saying, is that you need general illumination of a certain level, plus task lighting.

Click to expand...

^^^^^ This Guy Gets It ^^^^^

And some really good scientific study goes a little further to describe types and direction of task lighting vs: ability to focus, determine depth, and comfort level

I got a Lux meter a couple of years ago and checked eighteen 8 ft fixtures to see what was "best". My goal was to see if I needed to add fixtures when changing from a T12/T8 mix to all T8, since T12s are phasing out.

I have quite a hodgepodge mix of T8 and T12, new ballasts and old magnetic ballasts, some 2x80 watt, some 2x60 watt and some 4x32 watt. All are mounted at ceiling height, 10 ft off the floor with no diffusers.

I checked them all at a level 4' off the floor, so just shy of 6ft from the light. All lights had been on for a few hours.

I don't have the numbers in front of me but here's what I remember ......

Readings varied from around 580 to 840 lux.
The light that appeared brightest visually was not the brightest with the meter. Same for the dimmest.
The color temperature of the light makes a difference to my eye but not to the meter.
The brightest was an old T12 fixture using fairly new 80 watt bulbs (840 lux).
The dimmest was an old T12 using 'energy saving' 60 watt bulbs (580 lux).
The newer T8s were around 700 lux.

What's the formula for calculating lux to lumens/sq ft?

djjsr said:

I got a Lux meter a couple of years ago and checked eighteen 8 ft fixtures to see what was "best".

What's the formula for calculating lux to lumens/sq ft?

Click to expand...

Djjsr, great info, thanks for the post.

I have been using an online calculator:


http://www.unitconversion.org/illumination/lux-to-lumens-per-square-foot-conversion.html

Here is a link to a best practices spreadsheet for motor vehicle plant lighting levels. It is pretty specific for different types of areas to get energy star ratings.

This is for general illumination...NOT task lighting.

Might be a good starting point for us.

http://www.energystar.gov/buildings/sites/default/uploads/tools/Motor_Vehicle_Plant_Lighting_Best_Practices.xls?2f61-be6e

I looked at general assembly which recommended 50FC and then compared it to paint, which suggested 120 for production and 100 for inspection.

Based on this spreadsheet, I can see how a more sensitive lighting design needs to have different general illumination levels for different areas and functions. So my prior advice to get 100FC everywhere is invalid. It wastes energy and fixtures and causes eyestrain.

In addition, these plants, which include all the major manufacturers, use daylight where they can, to save energy and produce color temperatures that are better.

Of course, task lighting is also needed in conjunction with general lighting. And that is a whole nuther subject.

Bill

That calculator makes it easy!

Bottom line for me is that my new T8s (total 128 watts) seem to be only about 75-80% of the light of my old T12s (total 160 watts) according to the meter. I understand that they use less power, but if I have to add more fixtures to maintain the same level of light, I haven't saved energy.

Maybe I need to consider the power used by the new electronic ballasts vs the old magnetic ballasts.

djjsr said:

That calculator makes it easy!

Bottom line for me is that my new T8s (total 128 watts) seem to be only about 75-80% of the light of my old T12s (total 160 watts) according to the meter. I understand that they use less power, but if I have to add more fixtures to maintain the same level of light, I haven't saved energy.

Maybe I need to consider the power used by the new electronic ballasts vs the old magnetic ballasts.

Click to expand...

One way to increase the output of a T8 is to use a higher factor ballast.

I used a Phillips IOPA-4P32-HL ballast to increase the output of the 4 lamp light I bought. Basically, it drives the lamps about at a factor of 1.18 compared to a ANSI standard ballast. A low factor ballast is from .7 to .8 and normal is from .85 to 1. The ballast that came out had a factor to .82 (check what you have first). My understanding is that this does not affect lamp life. Start cycles and running a lamp too dim are more significant factors.

http://assets.sylvania.com/assets/documents/FAQ0056-0605.8d13d344-4cd2-42f2-af91-100b2a1a8a4d.pdf


Three of four bulb ballast: PN IOPA-4P32-HL
Two Bulb version: IOPA-2P32-HL

These are industrial ballasts (Non consumer EMI specs).

bczygan said:

Based on this spreadsheet, I can see how a more sensitive lighting design needs to have different general illumination levels for different areas and functions. So my prior advice to get 100FC everywhere is invalid. It wastes energy and fixtures and causes eyestrain.


Bill

Click to expand...

Agreed! And its also the reason I picked up a meter.

EDIT:

This link from posted by Joe Fin in another thread is a good read. Nice to see more info on light temperature vs perceived brightness.

http://www.iar.unicamp.br/lab/luz/ld...iderations.pdf

The thing I have trouble grasping is how far from the fixture the light will go.

Lets say we are making a target of 50 foot candles at a 4 foot height from a 12 foot ceiling.

Is there a rule of thumb for spacing of a 2 bulb 4' T8's at a 12 foot height?

jethrob said:

The thing I have trouble grasping is how far from the fixture the light will go.

Lets say we are making a target of 50 foot candles at a 4 foot height from a 12 foot ceiling.

Is there a rule of thumb for spacing of a 2 bulb 4' T8's at a 12 foot height?

Click to expand...

Look up the inverse square law for light fall off.

If I calculate it right, you get 44% of the light (Measured at floor level) from a fixture mounted at 12' AFF, that you would get from one mounted at 8'. This presupposes that both are mounted the same distance from the ceiling for reflectance.

Something else to remember is that calculations are not real world. There are lots of variables from lamp age to reflector shape to ceiling reflectance.

djjsr said:

The color temperature of the light makes a difference to my eye but not to the meter.

What's the formula for calculating lux to lumens/sq ft?

Click to expand...

We are all leaving out 1 very important aspect of Lighting Quality here even thou many of us are experiencing it in our shops.

We can flood the work space with light that emits a "Spike" in the infrared spectrum forcing the eye pupil closed were as the Lumens meter does not suffer this effect. Additionally the ways we perceive objects, depth, and acuity will be effected by this.

The bar at the bottom is the human visible wavelength range.




and here is the "Cool White Fluorescent"





remember human vision is in the 400 - 700 range

Excellent Thread

[Pardon the tardy reply to this thread. I've been sick; and then the Forum software got sick, and...]

Kevin C said:

One thing that I wanted to mention is calculated light VS light delivered to the actual surface. I see a lot of recommendations and numbers thrown around, but not a lot to back them up.

My concern is that some of the advice has not been vetted by real world experience or any type of measurement.

Click to expand...

First, thank you for making the effort to dive into some of these issues. And indeed, much of what you've posted to this thread is both interesting and potentially quite useful -- as long as it doesn't get "buried in the noise". The part a bit further down about the relative light fall-off when moving perpendicular to the tubes vs. when moving parallel to them, is especially good.

That said, I have to question some of the conclusions you're reaching in terms of absolute light levels. Whether that is because your LUX meter is not calibrated correctly, or you are using an incorrect conversion, or some other reason, I do not know.

View media item 36871
Photo of a work surface with a light mounted directly above it at a height of 36 inches. The light was mounted 24" away from a white wall.

730 LUX works out to 67.8 lumens per sq foot.

Click to expand...

Based on how many square feet? I did not see that crucial data point mentioned anywhere in your post.

Taking the example above, you presumably have, at minimum, two F32T8 tubes in that fixture. The actual lumen output will depend somewhat on both the exact tubes in question and the ballast factor of the fixture; but for broad strokes let's say 2,500-2,800 lumens per tube is a reasonable guess. So that's AT LEAST 5,000 lumens for the two tubes together.

If we assume that the "square feet" you're talking about is a two-foot by four-foot workbench top, that's about 625 lumens/ft.^2 (!), less any intervening losses. Yet, you're claiming to be measuring less than 11% of that.

Yes, the lens/diffuser will "cost" you a little, but not enough to make anywhere near THAT big a difference. And at a source-to-workplane distance of merely three feet, the inverse square rule is not going to take all that big a bite either.

Now obviously, there are MANY other factors which are also getting into the act here, some of them near-impossible to quantify -- such as, for example, the degree to which that fixture is ALSO distributing some significant portion of its output outside the bounds of your workbench top.

So my point here is not to take issue with (at least most of) the raw data itself; but rather, to caution against attempting to apply it too literally or too broadly.

2ManyProjects said:

Based on how many square feet? I did not see that crucial data point mentioned anywhere in your post.

Click to expand...

I believe using the measurement unit of LUX means X number of lumens on a surface of one square meter.

2ManyProjects said:

[Pardon the tardy reply to this thread. I've been sick; and then the Forum software got sick, and...]


That said, I have to question some of the conclusions you're reaching in terms of absolute light levels. Whether that is because your LUX meter is not calibrated correctly, or you are using an incorrect conversion, or some other reason, I do not know......

Click to expand...

My suggestion: Grab a light meter and repeat a similar series of tests to get a feel for the units and results. This is why I ran this series of tests.

I am only concerned about light delivered to the work surface. In this case, at a distance of 3 ft, directly under the light source.

he lux (symbol: lx) is the SI unit of illuminance and luminous emittance, measuring luminous flux per unit area.[1] It is equal to one lumen per square metre.

Click to expand...

http://www.unitconversion.org/illumination/lux-to-lumens-per-square-foot-conversion.html

...

A lumen per square foot (lm·ft⁻²) is a US Customary and British Imperial unit of illuminance and luminous emittance, measuring luminous flux per unit area and equal to 10.76 lux.

Click to expand...

1 lumens per (sq foot) = 10.7639104 lux


I can crunch the numbers a bit later... But a good start is to repeat the test I ran; any similar light will get you close enough.

djjsr said:

I believe using the measurement unit of LUX means X number of lumens on a surface of one square meter.

Click to expand...

Right. I should have put that together. My head is still full of... well, let's just call it "icky green stuff".

Kevin C said:

My suggestion: Grab a light meter and repeat a similar series of tests to get a feel for the units and results. This is why I ran this series of tests.

I am only concerned about light delivered to the work surface. In this case, at a distance of 3 ft, directly under the light source.

Click to expand...

And the tests you ran to produce RELATIVE results were indeed both illustrative and useful. It's the absolute light levels I'm questioning. I'm having a very hard time buying the idea of only 68 lumens/ft.^2 directly below and only three feet away from a light source that will putatively produce 5,000+ lumens. That's just too far off-scale to be plausible.

2ManyProjects said:

And the tests you ran to produce RELATIVE results were indeed both illustrative and useful. It's the absolute light levels I'm questioning. I'm having a very hard time buying the idea of only 68 lumens/ft.^2 directly below and only three feet away from a light source that will punitively produce 5,000+ lumens. That's just too far off-scale to be plausible.

Click to expand...

I ran the numbers last night, my calculations are withing 30% of measured.

How this could work out: We could debate the merits of my calculations and measurements and possibility that I'm wrong on both counts.

Or I could explain why your numbers are so far off and it will probably end up a never ending debate... Or you could try the same test yourself.

The title of the this post is Reality Check, in keeping with the subject of the thread, I would suggest you do the same.

Get a meter and verify your calculations.

Post a reply once you have run the same test.

2ManyProjects said:

It's the absolute light levels I'm questioning. I'm having a very hard time buying the idea of only 68 lumens/ft.^2 directly below and only three feet away from a light source that will putatively produce 5,000+ lumens.

Click to expand...

You'll find the Spectral Sensitivity on page 11 here

http://www.histest.com/blog/files/2012/08/LX1330B-Manual-new-version.pdf

JoeFin said:

You'll find the Spectral Sensitivity on page 11 here

http://www.histest.com/blog/files/2012/08/LX1330B-Manual-new-version.pdf

Click to expand...

Now THAT could (and probably does) explain a LOT.

It shows that the spectral response of Kevin's meter is VERY peaked at around 550 nm, and down by about 80% (relative to that peak) at both 500 nm and 650 nm, which of course are both well within the visible spectrum.

Fluorescent tubes are also notoriously "peaky" in terms of their spectral output. Different tubes, with different color temps and different CRIs will have peaks in different places; but NONE of them have anywhere near "smooth" spectral curves. Compare, for example:

http://en.wikipedia.org/wiki/File:Spectrum_of_halophosphate_type_fluorescent_bulb_(f30t12_ww_rs).png


http://en.wikipedia.org/wiki/File:Fluorescent_lighting_spectrum_peaks_labelled.svg


We have no idea what tubes Kevin has in his workbench task light, or what their spectral curves actually look like; but putting two and two together, it seems likely that much of their output is in (spectral) places the meter just isn't all that sensitive.

Which brings me back to my earlier statement: The tests of RELATIVE changes (based on such things as distance, direction, etc.) using the same light sources and the same meter seem reasonable, and are very illustrative. But the absolute "brightness" values are essentially meaningless, as they are simply too dependent on "coincidental" correlations of the source's spectral peaks vs. the meter's spectral sensitivity curve.

2ManyProjects said:

Now THAT could (and probably does) explain a LOT.

It shows that the spectral response of Kevin's meter is VERY peaked at around 550 nm, and down by about 80% (relative to that peak) at both 500 nm and 650 nm, which of course are both well within the visible spectrum.

Fluorescent tubes are also notoriously "peaky" in terms of their spectral output. Different tubes, with different color temps and different CRIs will have peaks in different places; but NONE of them have anywhere near "smooth" spectral curves. Compare, for example:

http://en.wikipedia.org/wiki/File:Spectrum_of_halophosphate_type_fluorescent_bulb_(f30t12_ww_rs).png

http://en.wikipedia.org/wiki/File:Fluorescent_lighting_spectrum_peaks_labelled.svg

We have no idea what tubes Kevin has in his workbench task light, or what their spectral curves actually look like; but putting two and two together, it seems likely that much of their output is in (spectral) places the meter just isn't all that sensitive.

Which brings me back to my earlier statement: The tests of RELATIVE changes (based on such things as distance, direction, etc.) using the same light sources and the same meter seem reasonable, and are very illustrative. But the absolute "brightness" values are essentially meaningless, as they are simply too dependent on "coincidental" correlations of the source's spectral peaks vs. the meter's spectral sensitivity curve.

Click to expand...

The lumen is based on the response of the eye. Certain wave lengths of light at the same radiometric level will appear brighter than others.



This is how your eye perceives light

Basically, as you head out to the far ends of the visible spectrum, we don't see as well. A long time ago, some smart guys developed a curve to chart color (wavelength) verses perceived brightness. This was done with hundreds of test subjects so it would be representative of the general population.

This curve is really useful, why? Because a source that has a lot of radiometric output in a range where the eye is not very sensitive would not be very / as useful as a light. For an extreme example, think of using tanning lights to illuminate your work bench. If you put an uncompensated photo sensor under that light, it would produce a good reading, but not be very good for reading (unintended pun). A heat lamp is another example.

To account for this, a compensated measurement called the Lumen was developed. The Lumen, uses a curve to simulate the response of the human eye.

The human eye has a response peak of around 550 nm, if you look at the chart for my meter, it has that same peak.

Other species will have very different curves. Even among humans it varies ( for example if a person is color blind).

But for the average population this is a really good way to judge how effective a light source will be. If your color blind, it shifts a little.

The Take Away

The light sources in the fixture have their output measured in Lumens.

Light meters for Photometry use the same curve; that is the curve my meter uses.

My measurements would be wrong if weren't taken using that "peaked" curve.

My low cost meter is not as smooth as a lab instrument, but its rated to within 5% for accuracy.

Peaks are weighted for spectral response of the eye. The luminosity is determined by the area under the curve (kind of sounds like calculus, cause it is).

Photo meters do have limitations, and spectral peaks do cause some inaccuracy, however test data typically shows this to be less than 15%, not the order of magnitudes we have been discussing.

Photometry involves the physical measurement of visible light
energy and attempts to compensate for the psychophysical
attributes of the human response and physical units of power.
Photometry is just like radiometry except that everything is weighted
by the spectral response ( luminosity function) of the human eye as
defined by the CIE.

Click to expand...

http://www.ies.org/pdf/education/IES-Color-3-Webcast-Handout.pdf This is a great article on light and measurements.

P.S. Do you think you will make the leap and purchase a meter to take a few readings? The feedback and understanding gained cant be understated. Its not like I'm suggesting getting an integrating sphere to put a T8 light fixture in.

Photo Optic Response Curves: Humans VS Chickens and Color Blindness.



If I wanted to measure lighting for chickens, would I need a light meter with using the curve on the right?



Shifts for red and green color blindness.

http://www.answers.com/topic/luminosity-function

Kevin C said:

The lumen is based on the response of the eye. Certain wave lengths of light at the same radiometric level will appear brighter than others.

Click to expand...

Yes, I understand that; and I didn't mean to imply otherwise in my earlier (and possibly too "concise" or over-simplified) comments. So for sake of simplicity, let's stipulate that the response curve in your meter is intended to mimic that of the human eye. How well it accomplishes that goal is an open question. There are a bunch of reasons why the "standard" curve may be more or less accurate in and of itself, as is always the case with something which is fundamentally based on statistical sampling; but I'd really rather not get bogged down on that right now. The BIG problem is attempting to apply that "one size fits all" result to different specific situations.

Perhaps I can illustrate this better by referring to a somewhat related area which has more directly captured my attention from time to time over the years: Namely, the age-old question about what color lighting is "best" for your night vision. The answer is ANYTHING but simple; and at the end of the day, depends almost completely on EXACTLY how you ask the question. For example, how EXACTLY do you define "night vision" in the first place? The lowest absolute level of illumination at which you can make out something, anything? And is that something/anything static, or moving (and at what rate in which direction reative to the observer, the light source, or both)? Or perhaps the best resolution of detail at some arbitrary constant level of varying-color light? And are we talking about discerning the light sources themselves, or objects illuminated by reflected light? Are you primarily concerned with being able to maintain visual acuity while operating under such restricted-spectrum lighting (say, for example, accurately reading navigation charts on an otherwise blacked-out ship's bridge)? Or rather, with being able to maintain sensitivity to extremely low-intensity full-spectrum light (i.e., seeing objects in the distance in near-full darkness) immediately after being exposed to said restricted-spectrum lighting? Or, for yet another incarnation, which color of low-intensity light on that channel marker will remain most discernable from two or three miles away? In clear weather or in pea-soup fog? As I said, FAR from a simple question, despite usually being initially posed as such. And so, we wind up getting a bunch of folks parroting such "wisdom" as "The Red Light Myth" (cf. http://www.google.com/search?q="Red+Light+Myth") -- which isn't a "myth" at all; it's just the inevitable result of applying the right answer to the wrong question!

Getting back more to the topic at hand, my point was (and is) that in the SPECIFIC case of measuring the output of artificial light sources (such as fluorescent tubes) which have EXTREMELY non-linear spectral outputs, relatively small differences in the details of that spectral output (which would "normally" have little or no effect on perceived "brightness") can get blown all out of proportion to effectively make HUGE differences in the measured results. Then too, regardless of the accuracy or applicability of the curve itself, there of course remains the question of how precisely it is implemented in a ~$30 meter, as opposed to, for example, http://sensing.konicaminolta.asia/products/t-10a-illuminance-meter/, at over $1,000 http://www.tequipment.net/MinoltaT-10A.html).

In short, there are SO many other variables in play that it is completely UNsurprising that the "absolute" measurements you observed seem at least somewhat at odds with common sense. In contrast, your tests which used the same light sources (whatever they may be) AND the same meter (whatever its "absolute" accuracy may be) to illustrate RELATIVE illumination levels based on geometry, etc., take at least most of these "unknown" and uncontrollable variables out of the equation, and thus remain both reliable and repeatable -- hence, useful.

Please understand, I am NOT saying your measurements are "wrong", per se. It's just that attempting to make them mean more than they really do is a fool's errand.

P.S. Do you think you will make the leap and purchase a meter to take a few readings?

Click to expand...

Frankly, at this point, with everything else going on (two broken houses, some health issues, etc.), it would be a rather low priority. And I would not expect it to produce all that much in the way of "new information" anyway. But hey, you never know.

In the meantime, if the spirit moves you, it might be very interesting to see the results of you repeating some of your tests using a variety of different-type tubes as the (sole, of course) light sources.

Update... I double checked my results, calculated and measured. The meter checks out to a calibrated unit to within 10%, the calculated numbers back up the readings.

FWIW, I used to test light sources for a living... This included spending time in the "Black Lab" testing light sources, AR coatings and other optical materials. This little exercise is not rocket science and crazy OCD long posts don't change the facts.

None of the recommended lights were anywhere near 625 lumens / sq ft. at 3'



Using the calibrated meter I got about 73 lumens / sq ft. at 3' VS my cheepo meters 67.

Calculated was about 75 lumens / sq ft. at 3'. Photometric data from the manufacture of the light fixture correlates to within 20% my measurements and calculations.

2ManyProjects said:

[Pardon the tardy reply to this thread. I've been sick; and then the Forum software got sick, and...]



Taking the example above, you presumably have, at minimum, two F32T8 tubes in that fixture. The actual lumen output will depend somewhat on both the exact tubes in question and the ballast factor of the fixture; but for broad strokes let's say 2,500-2,800 lumens per tube is a reasonable guess. So that's AT LEAST 5,000 lumens for the two tubes together.

If we assume that the "square feet" you're talking about is a two-foot by four-foot workbench top, that's about 625 lumens/ft.^2 (!), less any intervening losses. Yet, you're claiming to be measuring less than 11% of that.

Yes, the lens/diffuser will "cost" you a little, but not enough to make anywhere near THAT big a difference. And at a source-to-workplane distance of merely three feet, the inverse square rule is not going to take all that big a bite either.

Click to expand...

1: The the inverse square rule only applies to a point source. Once you use some type of optics to shape the light it no longer applies.

2: Rule of thumb of optical surfaces: Un-coated plastic will cost you about 8%. AR / Index matching coatings can cut that to 1% per side. Unfortunately they are not used on diffusers. Rule of thumb, is a 8-10% loss from the diffuser.

3: Surface reflectivity: Again, figure at least a 8 to 10% loss for each reflection. Multiple bounces mean a 8-10% loss for each bounce.

4: The tubes shadow the output of half the lamp. Light coming off the top of the light reflects off the top of the fixture and out or back down to the tube. Light reflected off of the top of the tube is bad. Transmission through the tube is very poor (90% attenuation). Typically Light off of the top of the lamp takes at two to three passes to get out and loses 8 to 10 % for each reflection.

This is where a T5 has an advantage, its smaller diameter reduces how much it shadows its own light. Same thing on a T8 VS a T12.

What is wrong with this calculation?

Using your numbers: 5000 Lumens. 625 Lumens per sq ft = 8 sq feet. For Broad strokes, why in the world would start with 100% coupling between the source and the work surface? Wrap fixtures are designed to have a very, very broad light pattern. Just look at how much light is thrown horizontally onto the walls.

And, yes I measured it.

My Estimate:

Two Tubes, 2800 Lumens per tube = 5600 Lumens of raw output.

Since its a wrap fixture that's putting out light in a 180° pattern the light is spread out over a much larger area than you assumed.

Light from the top 180° of the lamp is attenuated by about 25% (on average three reflections plus shadowing).

Light from the top of the lamp= 5600 / 2 * .75 = 2100 lumens.

Light from the bottom 180° of the tube= 5600 /2 = 2800

Total light hitting the diffuser = 4900 lumens.
The diffuser cuts that 4900 lumens by 8%, to give us about 4500 lumens.

Adjusting for a .9 Ballast factor and we have 4050 lumens.

The wrap fixture has a better than 180° pattern. I measures it at about 200°. Basically, its shooting light up at an angle above the light. However, the intensity was dropping off, so I will go with 180°.

Also, the light pattern extends about 1' to each end of the light.

The 4500 lumens is radiating out in a cylindrical shape that is about 6' long. Since I measure the intensity at a distance of three feet, its pretty simple to calculate the area of the cylinder and the lumens per sq foot.

Formula for the area of the outside of a cylinder: 3.14 * D * L

3.14*6"*6"=113 sq feet. Adjusting for the degrees of light emission (180°)

(113 / 360)*180= 56.5 Sq feet (not 8 sq feet)

4050 Lumens / 56.5 sq ft = 71 lumens per sq foot.

Conclusions? My meter readings are very close to calculated, the labs calibrated meters readings are even closer. This is also backed up by the photometric data from the fixture manf (within 15%).

Four data sources have very similar numbers (my calculations, meter #1, meter #2, and the data from the manufacturer of the fixture).

I'm going with my this; My $35 meter is pretty darned accurate.
The only figure that's way off... The 625 Lumens per sq ft calculated By 2MP is off by by at least a factor 8+..

Kevin C said:

1: The the inverse square rule only applies to a point source. Once you use some type of optics to shape the light it no longer applies.

Click to expand...

That is such a gross oversimplification as to become highly misleading.

The use of lenses and reflectors can effectively change the applicable exponent value; but it doesn't obviate the underlying principle. And besides, "point source" implies spherical radiation; "inverse square" would apply to half-spheres.

Bottom Line: There is no such thing as "lossless" transmission of light over distance, even in a perfect vacuum (i.e., outer space).

The only figure that's way off... The 625 Lumens per sq ft calculated By 2MP is off by by at least a factor 8+..

Click to expand...

As you will hopefully recall, that "figure" was preceded by (and premised upon) a big honkin' "IF", and was intended to be a silly example. Besides, it later proved moot anyway; so why bring it up now?

2ManyProjects said:

That is such a gross oversimplification as to become highly misleading.

The use of lenses and reflectors can effectively change the applicable exponent value; but it doesn't obviate the underlying principle. And besides, "point source" implies spherical radiation; "inverse square" would apply to half-spheres.

Bottom Line: There is no such thing as "lossless" transmission of light over distance, even in a perfect vacuum (i.e., outer space).



As you will hopefully recall, that "figure" was preceded by (and premised upon) a big honkin' "IF", and was intended to be a silly example. Besides, it later proved moot anyway; so why bring it up now?

Click to expand...

I like it... When called out on BS it your response is more BS; Pretty much what I expected.

The point of my starting this thread was a reality check on lighting calculations that were very far off from reality.

Ironic that you tried to discredit that reality check with some really , really bad lighting calculations that you now call silly. What was the point of that?

Why bring it up? Why not... You linked to this thread as an example of why field measurements are not accurate. I had let it drop, you brought it back up, pretty simple.

Where is that dead horse??????????

Thanks Kevin C for sharing your expertise !!!

Reality check is really what GJer's need to know, and filter out the noise !