How reliable are smartphone light meter apps?

[ALToronto]

I have two different apps on my phone, and they correspond closely. But that's not surprising - both apps use the same sensor. What's worse, it's the sensor for the front facing camera, which is always inferior to the rear facing camera.

Has anyone with a serious light meter tested the meter against an app? Is the error random, or do smartphones consistently measure higher or lower? I'm interested in anything but Apple devices - Apple is the worst for colour accuracy, so I wouldn't trust it.

What's prompting this question is a recent presentation I attended by Glenn Decker of Piping Rock Orchids, where he said that most paphs require about 1200 foot-candles of light, or 'low' light level. Well, according to my phone, 1200 fc is pretty bright - I can't get close to it without some direct sunlight. And my living wall with 70 W total of LED's, can barely reach 450 fc in the areas closest to the lights, where my cattleyas are getting hyperpigmentation.

So is it 1200 fc or 1200 lux? And if it's fc, just how far off is my phone?

 

[gonewild]

Most cell phones cant measure over 1000 fc. really the apps are made by people that measure light sitting at their computer.
I tried to find one that worked on my HTC but none gave the same reading on the same spot twice. Maybe at fc under 500 they work?
I just ordered a cheap light meter from Amazon.

 

[gonewild]

1200 lux is only about 120 fc.

1200 fc is about 88% shade

 

[PaphMadMan]

I hope we get some good answers because I have the same questions.

At least one app I have looked at can be set to read in both fc and lux, and the relative readings of fc and lux seem correct on it. (~10.76 lux per 1 fc) It also has pre-set calibrations for various devices, or an option to let you calibrate it to match another meter or a known lux level. I haven't tried that.

The 1200 fc recommendation should be about right for Paphs. I have no question that he did mean fc. But I have wondered about the applicability af any light meter to LEDs or any other lighting that isn't a full spectrum similar to sunlight. You could have all the wavelengths of light your plants need at the right intensities, but the fc from a meter calibrated for full spectrum ought to read lower than the recomemnded fc for full spectrum since it is receiving much less total light energy, but much of it from wavelengths not used by plants.

 

[ALToronto]

Most cell phones cant measure over 1000 fc. really the apps are made by people that measure light sitting at their computer.
I tried to find one that worked on my HTC but none gave the same reading on the same spot twice. Maybe at fc under 500 they work?
I just ordered a cheap light meter from Amazon.

My phone gives very consistent readings in the same spots. One app samples and holds, so I don't like using it. The other app changes continuously, so I can aim the phone until I get the highest reading, and that's the value I use. I can go from the living wall to the windowsill and back, and the readings will be within a few %. So if there is an error in the sensor, it's non-random.

My phone can certainly measure over 1000 fc. Full sunlight is 65535 lux, which is 2^16-1 (2 to the 16th power, minus 1). Must be an arbitrary limit. This corresponds to 6088 fc. Dappled sunlight is around 3000 fc, but full shade in an otherwise bright window is maybe 500 fc. Does this correspond to readings from real light meters?

 

[gonewild]

I have an app called "Light Meter" sitting on my desk pointing up at 2 t8 tubes it reads 18fc. Right beside it an old photo light meter reads 46fc.
Switch apps to "BeeCam" it reads 160 lux which is about 16fc.
so both apps read about the same but only measure about have as much light as the photo meter.

maybe someone has an answer which is correct if any.

 

[RNCollins]

I don't know how accurate it is, but I use the Mobility Quotient Green Thumb light meter on my iPad. It measures in foot candles or lux.

http://mobilityquotient.com/press/plant-lovers-love-mqgreenthumb

 

[lepetitmartien]

If you have a camera (digital or film) you can use its lightmeter to calculate the incoming light and therefore verify if the apps are close to it.

Note that it's not important to have a perfect measure but an order of magnitude of what's going on.

 

[troy]

I use lux meter level pro free app it reads from the face of my phone which kind of sucks but it reads up to over 30,000 lux

 

[ALToronto]

If you have a camera (digital or film) you can use its lightmeter to calculate the incoming light and therefore verify if the apps are close to it.

Note that it's not important to have a perfect measure but an order of magnitude of what's going on.

I haven't used a real camera since 2008, when I bought a phone with a 5 MP camera. But how accurate is a camera's light meter? Most professional photographers I know use separate devices, and some are worth hundreds of $$.

And this is another question: how do analog light meters compare to each other? Is a cheap one less accurate because it has a higher random error, or is it more like a smartphone sensor that simply measures less light, or ignores some wavelengths altogether?

I'm not surprised by Lance's result - a factor of almost 3. So this means that my living wall catts are probably not living under a rock, but are actually getting enough light to eventually bloom. Maybe I should just trust my eyes a little more.

 

[lepetitmartien]

Metrology even on the scientific level is no pleasure. That's why I gave you the advice to search for the general idea and not be fixed over the numbers. Camera meters are reliable enough to use accurately films of defined sensibilities. Pro light meters are not only accurate but have other functions like measuring the different types of light in the area (incident, spot level on the subject etc. and flash lights), you only need to know whats going on.

Btw you can find real cheap luxmeters made in China that are reliable enough for our job…

Now if you have results in the same league measured from different means (phone, camera, cheap luxmeter) you can bet the numbers may not be accurate but the general idea is. It's like weighting one thing with a bad balance, you have to make multiple measures to get things right.

I've got to compare my luxmeter and my EOS 7D, I'll let you know.

 

[ALToronto]

Thanks lpm.

I'm glad you brought up metrology. The big challenge of metrology is separating and calculating random and non-random errors. From my own observations, the random error of my phone sensor is very small. However, the non-random error is probably significant enough that it's worth finding out by how much and possibly plotting a calibration curve.

I'll try to borrow a professional light meter.

 

[lepetitmartien]

Someone had to bring up the bad words! LOL I know how it can s…k to figure out errors and compensate them now I think it's more important to know if we have about 3000 lm or 30000 lm than 2997 and 34485.

For those who want to calculate using a camera light meter:

- switch to 100 ISO
- if you can, select Av priority (aperture), to f/8
- if totally automatic, just go on
- place a white paper on the zone you want to test, it's your target.
- measure, note the aperture (F/8 or the one decided by the camera) and the exposure time.
- Calculate!

the formula is:

Light intensity in lux = (70 x f^2)/(exposure time x ISO)

For 100 ISO, f=8 and 1/1000th of second (0,001") it'll give 44800 Lux

Else, a french calculator: http://orchideenord.free.fr/index.php?section=Lux.html (in lux of fc)

 

[Ray]

I don't know how accurate it is, but I use the Mobility Quotient Green Thumb light meter on my iPad. It measures in foot candles or lux.

http://mobilityquotient.com/press/plant-lovers-love-mqgreenthumb

As someone pretty well known to test, assess, and publicly feed back on new stuff, they gave me a free copy to test.

It seems pretty good, but needs a paper diffuser over the phone lens in order to work, and the "weight" (thickness) of the paper, and it's "brightness" (degree of coating) both affect paper's light transmission, so it seems a bit "iffy" to me.

Of course there is a way to "calibrate" it using full noontime sun, so maybe the paper quality isn't that big of a difference....

 

[Clark]

For those who want to calculate using a camera light meter:

- switch to 100 ISO
- if you can, select Av priority (aperture), to f/8
- if totally automatic, just go on
- place a white paper on the zone you want to test, it's your target.
- measure, note the aperture (F/8 or the one decided by the camera) and the exposure time.
- Calculate!

the formula is:

Light intensity in lux = (70 x f^2)/(exposure time x ISO)


For 100 ISO, f=8 and 1/1000th of second (0,001") it'll give 44800 Lux

Thanks!

 

[lepetitmartien]

I was sure it'd be of help for some :rollhappy:

 

[lepetitmartien]

My ridiculously cheap luxmeter from evil China gives me about 1580-1600 lux

And at the same time, my eos 7D :

(70 x 8^2)/(0,01 x 200) = 2240 lux

Another measure, respectively: 5800 (China) and 5600 (7D by hand) 5480 (7D using the calculator)

Note there's big variations of light due to clouds today, but it's better than yesterday, and I did that near windows at home. I'll repeat outside with full sun when I can. Seemingly, my luxmeter is not totally out in space though I would never work for a Nobel with it. :evil:

 

[gonewild]

i just got an inexpensive digital light meter.
Here is a comparison of readings...

Dr. Meter = 86fc
Old GE photo meter = 65fc
Light Meter app on HTC phone = 18fc

The old meter and new digital meter are at least in the same general range but the phone app is no where close.

 

[PaphMadMan]

i just got an inexpensive digital light meter.
Here is a comparison of readings...

Dr. Meter = 86fc
Old GE photo meter = 65fc
Light Meter app on HTC phone = 18fc

The old meter and new digital meter are at least in the same general range but the phone app is no where close.

Does the phone app allow for any kind of calibration setting?

 

[gonewild]

Does the phone app allow for any kind of calibration setting?

Yes it has several calibration options. I have not tried to adjust. Calibration requires use of a normal light meter and I just got the new one. But if you have a meter already who needs the phone app?