Advice Needed On Lux Meters

[CelticCross74]

Its about high time I finally got a lux meter. What is everyone using, where did you get them and for how much?

 

[HKJ]

Get a Gossen, it is very good. It will show correct values!
But it is a bit on the expensive side

 

[hiuintahs]

I use an Extech HD450 data logging light meter. I use it to make run time graphs and compare the efficiency of flashlights to another. Also I have found it useful for comparing beam intensity. I picked one up for around $175 used on Ebay a few years ago.

 

[CelticCross74]

Gossen and Extech well alright that is a great starting point.

 

[NoNotAgain]

CC74, I've got the Gossen Mavolux M502B USB model. I've also got the software for charting out the readings.
The M502C has finer measuring resolution than the B model.
You definitely want the software for the plotting function.

I've used Gossen meters for 30 years starting with the Luna Pro then the best photographic hand held meter IMO the Ultra Pro.

The Mavolux can be found on eBay at times for a good prices. Expect anywhere between $150-200.

 

[Viking]

An important feature is maximum peak hold. Make sure it has it.
I have an Extech EA30.[h=1][/h]

 

[ShineOnYouCrazyDiamond]

If you're looking for a lower cost meter just to do some comparison testing with the Dr.Meter is a $40 alternative. It has good reviews on Amazon https://www.amazon.com/gp/product/B005A0ETXY/?tag=cpf0b6-20 and I've been happy with it for the little bit I use it for.

 

[CuriousOne]

Is luxmeter useful without the integrating sphere?

 

[ShineOnYouCrazyDiamond]

A lux meter tells you the brightness of a point of light.

A lux meter is used in conjunction with an Integrating Sphere to determine lumens. Calibration is required for this type of setup.

 

[CelticCross74]

Extech EA30 is looking good

 

[Grizzman]

I've had an EA30 for a bit over 3 1/2 years. I've had no complaints so far, and no desire to upgrade....and doubt I will.

 

[CelticCross74]

wow well thats a pretty solid endorsement Grizzman thanks!

 

[hiuintahs]

Is luxmeter useful without the integrating sphere?

If you want to do run time graphs, you need to have an integrating sphere ($$) or some type of home made light box that will provide a constant medium. Then all your measurements will be relative to each other. After doing dozens of tests and comparing to the manufacturers stated lumens, I have an equation that converts lux to lumens that works pretty good for my setup.

I patterned the light box after Selfbuilt and Quickbeams milk carton design.

 

[HKJ]

If you want to do run time graphs, you need to have an integrating sphere ($$) or some type of home made light box that will provide a constant medium.

Not really, measuring lux works fine.

 

[hiuintahs]

Not really, measuring lux works fine.

I guess I could just shine the light at the meter at constant distance. However, the light box does keep other ambient light sources out such as the room lights.........and when I got interested in this sort of stuff, Quickbeam and Selfbuilt were doing a lot of run time graphs and so I thought those guys were pretty good examples to follow.

Also another good reason to have the light box is that I have found when doing intensity measurements at 2 meters, that the reading is all over the place depending on just a slight movement angle of the flashlight. Maybe that is why Viking suggested peak measurement. In the light box, the light is averaged due to the reflection on the walls.

Edit:
Thinking about this further, I think what HKJ might be referring to is bouncing the light off of the ceiling. The room would create a constant medium. Basically the measurements are compared against each other is the main purpose when you can't actually measure lumens.

 

[billcushman]

The Extech LT45 will give good results on most conventional and LED light sources. Download and read the User Manual. It is the most accurate cheap (less than $300) meter and it compares pretty well to a $10K+ Spectroradiometer. Accurate Lux measurements are very difficult to achieve and highly dependent on the spectrum of the light source. Only a spectroradiometer can replicate the photopic curve with high precision. If the meter specs don't include the f1' data and a graph of deviation from the photopic curve the results are very likely to have large error.

 

[HKJ]

Only a spectroradiometer can replicate the photopic curve with high precision.

That is not correct, the precision depends on the filter, it has to match the eye. Any meter can have a good filter, but due to the cost of a good filter very few meters have a good one.
The first category to avoid is meters where you have to select light source (Incan., sunlight, fluorescent, etc.), that means they do not really have a filter, but the calibration is only valid to a few specific spectrum's.
One test is to shine a UV lamp on the sensor, it must not show brighter that it looks, i.e. like a flashlight in moonlight mode.

 

[reppans]

I'll +1 for the Dr. Meter...... and the BLF lightbox idea:

With this sized tubing, my lumen conversion is a straight factor of 10... only need to move the decimal point. For calibration, I tie-up closely to a couple of ti-force's Quark reviews - the only CPF reviewer to claim ANSI accuracy with laboratory tested calibration lights. I can also tie-up closely with my HDS's 0.02/325 min/max modes, the only light said to be individually calibrated (by Henry). Interestingly, half (3 of 6) of the calibration sources that Selfbuilt uses also happens to be ti-force, 4Sevens, and Henry (Novatac days), although SB doesn't seem to match-up to them anymore.... while the three sources seem to match each other well.

 

[ShineOnYouCrazyDiamond]

^^ That's pretty cool. I may have to give that a try.

 

[billcushman]

My statement is absolutely correct. Even the most expensive filters cannot approach the accuracy of a spectroradiometer. NIST has published extensive data on this subject. A typical spectroradiometer has 256 channels and measures the stimulus in 5 nanometer bands. The photopic curve is calculated with almost no error. Even in 5K photometers using error correction for a very costly filter stack cannot approach the accuracy of the calculated photopic curve of the best spectroradiometers.

That is not correct, the precision depends on the filter, it has to match the eye. Any meter can have a good filter, but due to the cost of a good filter very few meters have a good one.
The first category to avoid is meters where you have to select light source (Incan., sunlight, fluorescent, etc.), that means they do not really have a filter, but the calibration is only valid to a few specific spectrum's.
One test is to shine a UV lamp on the sensor, it must not show brighter that it looks, i.e. like a flashlight in moonlight mode.