• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Cell phone flash

Martin Evans

New member
Can anyone give me an approximate value for the Guide Number of a typical flash as fitted to a cell phone ("mobile phone" in Britain) or smart phone with built-in camera?

I am updating my "essay" on flash photography in galleries and museums, and would like to calculate the gallery-lighting time equivalent, as it might affect conservation issues. The technical data that I have seen for cell phones do not give any values; just a simple statement that the flash is a dual-LED unit or some-such.

I don't want four-figure accuracy! Just a ball-park number. From real data or personal knowledge, please, not a guesstimate.

Thanks and best wishes to all; apologies for long absence.

Martin
 

Asher Kelman

OPF Owner/Editor-in-Chief
Martin,

Why can't you measure the output with a flash meter? If you don't have one, I'll try to do some measurements when I'm back in my studio.

I doubt that there's enough energy in the flashes to compare with having lights on in the museum gallery!

Asher
 
Most phone "flashes" are not flashes at all, but white leds. They cannot be measured with a flashmeter.

Hi Jerome,

I agree that it may be difficult to get a reliable flash-meter reading, or any reading at all. It also depends on the actual flash-meter used, because it triggers on a steep enough ramp in the lighting level. Some are more sensitive than others.

I managed to get one reading of f/2.2 at approx. 1 metre distance @ ISO 100, suggesting a metric Guide Number of 2.2, but subsequent readings failed with an underexposure warning. I was using the dual LED flash on my smartphone. Probably the ramp between the constant focus assist light, and the 'flash' for the actual exposure, is not steep enough, so the reading is probably also not accurate at all.

Maybe it is more useful to measure the focus assist light intensity, since that would last longer and have more impact on the gallery-lighting time equivalence. For that purpose one should perhaps also consider the spectral composition of the light emitted by the LED(s). Usually 'white' LEDs are used, which are a mix of UV/blue and fluorescent Yellow/Green emission. Newer 'flash' units can use a combination of R/G/B LEDs to influence the color temperature.

Cheers,
Bart
 

Jerome Marot

Well-known member
You can't get reliable metering with a flash meter, because those are designed to measure for a few thousands of a second, so as to ignore the continuous light that is often present in a typical photo studio. The typical led lights of cell phones will be turned on for, probably, about 1/50s - 1/10s for exposure, so much longer. Your flash meter will not integrate all the light.

I could fire the oscilloscope connected to a solar cell or use the high-speed video modus of a small casio camera to measure the illumination time of these leds (hint...), but quite frankly, I don't see how it would answer the question which was about conservation issues in galleries.

If one wants an idea on how much energy is emitted by these lights versus a small flash during the length of exposure, taking a picture in darkness at increasing distances until exposure is too dark is the way to go. But that still will not tell us about conservation issues, since we do not know how much UV they emit.

With iOS 7, the flash led can be switched on to be used as a flashlight, BTW. It is as powerful as a 5-10 lumens led flashlight, by direct comparison. But maybe Apple uses a brighter mode for photography.
 

Doug Kerr

Well-known member
Hi, Martin,

We discussed this general matter a while ago.

It would seem as if (assuming that the sort-term irradiance is not "extreme" - that is, so potent that it does not cause the oil paint to immediately "fry") that the deleterious effect on artwork of exposure to electromagnetic radiation (visible or otherwise) likely relates to the radiant exposure (He) (the time interval of the irradiance, Ee), except with some weighting by wavelength (dependent on the artistic media involved) applied.

There is not the slightest reason to believe that the weighting would be the same as used to determine photometric quantities from radiometric observations, as is done in photographic exposure meters, photoflash meters, and the like.​
Has such a relationship been established (or even identified) by published research?

If such is the case, then to evaluate the impact of time-limited irradiation by a device such as one burst by a photoflash unit, or one blink of a smartphone illuminator, or 10 seconds of a certain kind of flashlight, or a day's exposure to the ambient lighting in the gallery, would need to be done by an integrating radiant exposure instrument provided with a filter that implements the applicable weighting by wavelength (which of course I'm sure would vary for different kinds of artistic media).

And I would think that such an instrument would be of great value to conservators.

Perhaps if none exists, you should devise one, and could become wealthy from it.

Best regards,

Doug
 

Jerome Marot

Well-known member
If one wants an idea on how much energy is emitted by these lights versus a small flash during the length of exposure, taking a picture in darkness at increasing distances until exposure is too dark is the way to go.

Now that the sun has set here, I tried just that with an iPhone 4S.

Interestingly, the iPhone shutter speed in total darkness with "flash" is 1/16s. I would make sense that the led is lit for the same amount of time, considerably longer than a real flash.

When the distance is just about too much for the tiny iPhone led to illuminate correctly the picture, I am about 4m50 away from the subject. The iPhone increases its sensitivity to iso 800 (which I cannot change or choose) and the lens is opened at f/2.4.

2.4 x 4.50 = 10.8, so we would have a guide number of 10.8 at 800 iso. Since guide numbers are usually given for 100 iso, that makes it 3.8.

Therefore the answer is:
the iPhone 4S outputs as much energy to take pictures as would a flash of guide number 3.8 for 100 iso, in meters (in feet, for the 3 non-metric countries in the world: 12.5). It is a very tiny flash.
 

Doug Kerr

Well-known member
Hi, Jerome,

Therefore the answer is:
the iPhone 4S outputs as much energy to take pictures as would a flash of guide number 3.8 for 100 iso, in meters
That would be so if the length of its "burst" were wholly confined to the open shutter time. Is that likely so?

Best regards,

Doug
 

Jerome Marot

Well-known member
That would be so if the length of its "burst" were wholly confined to the open shutter time. Is that likely so?

Doug, please consider the following:
-a led can be switched on and off pretty fast
-conserving power on a portable device like the iPhone is a priority
-the iPhone was designed by some of the most competent engineers on this planet.

Does this answer your question?
 

Doug Kerr

Well-known member
Hi, Jerome,
Doug, please consider the following:
-a led can be switched on and off pretty fast
-conserving power on a portable device like the iPhone is a priority
-the iPhone was designed by some of the most competent engineers on this planet.

Does this answer your question?
No.

Best regards,

Doug
 

Asher Kelman

OPF Owner/Editor-in-Chief
Getting back to the issue of safer of the pictures, I believe it's unlikely that all of the flashes of visitors will approach the total energy of the gallery lights on during the show. I believe someone once did his Thesis on this.

The real damage from people, e.g. in the Vatcan Museum in Rome, is the increase in C02 levels and production of carbonic acid with water vapor.

Flashes are really of no significance in most circumstances. It's something like hyperactivity after sugar, only useful knowledge in a "Twinky Defense" in a murder case!

Asher
 

Martin Evans

New member
Getting back to the issue of safer of the pictures, I believe it's unlikely that all of the flashes of visitors will approach the total energy of the gallery lights on during the show.
Asher

Yes, that is the point that I have been trying to make, since 1994! I then calculated that normal gallery lighting, at 200 lux, is equivalent to around 30,000 "average" flashes per day. A detailed experimental study by David Saunders of the National Gallery in London, published the following year, gave comparable results.

The power of the flashes built into modern compact cameras is lower, of course, and correspondingly less dangerous. All this is in an "essay" of mine that is at:

http://people.ds.cam.ac.uk/mhe1000/musphoto/flashphoto2.htm

My query was intended to let me update this essay to cover the smart-phone cameras that are an increasing nuisance (but not a conservation threat) in galleries. I shall assume probable guide numbers in the range 2 - 4 (ISO 100, metres).

Thanks to all,

Martin
 

Doug Kerr

Well-known member
Hi, Martin,


I must admit that, despite our having collaborated on some areas of this work, I had never read your essay before. It is excellent, rigorous but accessible, and a "smooth read".

I have only the tiniest quibble. You say:

Hanlan (1970) exposed a number of pigments to a series of intense flashes from a powerful 4500 footcandle second studio unit, that seems not to have had the ultraviolet (UV) filtered out.​

We cannot describe the time-integral luminous output from a flash unit (even in a particular direction) in the unit footcandle•second.

The unit footcandle•second is a measure of luminous exposure. A given flash unit will create a certain luminous exposure on a surface at a certain distance from the unit. (And I'm sure that is what you had in mind, for typical distances that would occur in flash unit use.)

The pertinent metric of the time-integral emission from a flash unit (in a certain direction) is denominated in the non-SI unit candlepower•second.

Perhaps the passage would better be:

Hanlan (1970) exposed a number of pigments to a series of 4500 footcandle•second flash exposures from a powerful studio flash unit, that seems not to have had the ultraviolet (UV) filtered out.​

Again, a very small editorial quibble. Congratulations on a very useful essay.

Best regards,

Doug
 

Martin Evans

New member
Hello Doug, and many thanks for your knowledgeable feedback. I don't have your engineering background, so I didn't spot the error.

I was citing Hanlan's work second-hand. I never got to see his original paper, but David Saunders quoted his work in his own excellent 1995 paper based on experimental work at the National Gallery. I'm surprised that Saunders would slip up over photometric measurements, but the relevant paragraph is:

"Hanlan's flash produced 4500 foot candle secs, which is equivalent to approximately 48,500 lux•s, 75 times more light output than the 650 lux•s from the Mecablitz flash unit."

Is Saunders' figure of 48,500 lux•s a meaningful one for a studio flash? I could use that value in place of the erroneous one if it is less misleading.

David Saunders moved from the National Gallery to a post at the British Museum a few years ago. Perhaps under his benign influence, the BM has a very tolerant attitude to photography in the galleries, even letting visitors use flash.

Best wishes from a chilly night in East Anglia! The moon is out and they forecast fog later tonight. No doubt the sun is still shining in New Mexico.

Martin
 

Doug Kerr

Well-known member
Hi, Martin,

I was citing Hanlan's work second-hand. I never got to see his original paper, but David Saunders quoted his work in his own excellent 1995 paper based on experimental work at the National Gallery. I'm surprised that Saunders would slip up over photometric measurements, but the relevant paragraph is:

"Hanlan's flash produced 4500 foot candle secs, which is equivalent to approximately 48,500 lux•s, 75 times more light output than the 650 lux•s from the Mecablitz flash unit."

Is Saunders' figure of 48,500 lux•s a meaningful one for a studio flash? I could use that value in place of the erroneous one if it is less misleading.

In fact 4500 foot candle•sec is equivalent to 48,500 lx•s. The problem is that both are a measure of the photometric exposure on the target surface, not of the output of the flash itself.

So the question is, "at some specified distance, is a photometric exposure of 48,500 lx•s credible from a studio flash."

And of course "studio flash units" come in a gigantic range of outputs. And most can be equipped with a wide range of reflectors, and the luminous intensity-time product (which, with the distance, is needed to determine the photometric exposure) varies greatly with the nature of the reflector.

And often the manufacturer is reticent about actually specifying the output, with any particular reflector, in the proper system of units.

But I know some do, as I recall a study I did a few years ago in which I found quite a bit of such data from some manufacturers. So let me try and dig up some of that scoop and we'll see what we can make out of it.

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Martin,

This has been revised based on further manufacturers data I received since the original post.

*******

As you know, studio flash units are generally "rated" in term of energy storage, in Joules. This does not bear any direct relationship to total luminous output, although for typical units the relationship can be estimated.

A rating in terms of Guide number (GN) can be fairly well related to luminous output (in cd•s).

About the most potent studio flash units encountered in "normal" photography are those rated at about 5000 J.

Based on some manufacturer's GN data for a studio flash unit rated at 4800 J with a reflector giving a beamwidth of about 60°, if I go through a whole trail of stipulations, assumptions, and extrapolations, I get a SWAG (scientific wild-assed guess) that the luminous intensity-time product of the flash beam would be on the order of 60,000 cd•s.

If we contemplated this being used at a distance of 10 ft from the object, then the photometric exposure on the object would be about 6000 lx•s.

I hope. This was all done on the back of an envelope.

Best regards,

Doug
 

Martin Evans

New member
Good morning Doug!

Very many thanks for clarifying those points. I had not properly understood the differences and it seems that Dr Saunders was a bit careless in the way he quoted Hanlon's work. I have still not seen Hanlon's original paper so it is impossible for me to know exactly what quantity of light his specimens were exposed to.

For that reason I have re-worded that part of my essay, and simply left Saunders' assertion that Hanlon's trials used a flash unit 75 times more powerful than his own. That might not pass a referee's criticism for publication in a learned journal, but it will do for the purposes of my essay.

Thank you for your interest and expertise, for the time you have spent (I hope that it didn't make you late for dinner) and for the sacrifice of an old envelope!

Best regards,

Martin
 

Doug Kerr

Well-known member
Hi, Martin,

Good morning Doug!

Very many thanks for clarifying those points. I had not properly understood the differences and it seems that Dr Saunders was a bit careless in the way he quoted Hanlon's work. I have still not seen Hanlon's original paper so it is impossible for me to know exactly what quantity of light his specimens were exposed to.
I understand.

For that reason I have re-worded that part of my essay, and simply left Saunders' assertion that Hanlon's trials used a flash unit 75 times more powerful than his own. That might not pass a referee's criticism for publication in a learned journal, but it will do for the purposes of my essay.
I think that is an excellent plan!

Thank you for your interest and expertise, for the time you have spent (I hope that it didn't make you late for dinner) and for the sacrifice of an old envelope!
It was my pleasure. It gave me an excuse to refresh some of my own thoughts in that matter.

I plan to to construct from basic principles a relationship between the luminous intensity-time product of a flash unit and the guide number metric. I have been using one expressed years ago in some Kodak document, and I would rather have one whose derivation I knew! I'll let you know my result in that.

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Martin,

In my earlier work, I had used a relationship I found in some ancient Kodak document to determine the guide number for a given flash luminous intensity-time product (a quantity that sadly does not have a good nickname - it is sometimes called "LITP"). In SI units, that relationship is:

Gf = sqrt((I•t)/20)

where Gf is the guide number(I believed it to be in feet) and I•t is the luminous intensity-time product (in cd•s).

I wanted to derive the relationship from first principles for comfort.

Based on the "incident light exposure formula", with a widely-accepted value for its "calibration constant", I then derive:

Gm = sqrt((I•t)/25)

where Gm is the guide number in meters.

That would correspond to:

Gf = sqrt((I•t)/2.3)

So I have to do some sorting out of the substantial discrepancy (about 3:1 in G).

I will do some credibility checks with other data.

But now breakfast (breakfast "A" today: turkey sausage patty, banana, blueberries, half grapefruit, half apple, half orange, steel-cut oatmeal with cinnamon, selected essays from the New York Times and Washington Post). Am I spoiled or what!

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Martin,

International standard BS ISO 1230:2007 gives this for guide number (using my notation);

Gm = sqrt(0.51 I•t).

This again comes from a certain "exposure equation" implied by the standards for photographic exposure meters. I have not yet followed their derivation.

This comports reasonably well with my recent derivation (starting with a different assumed exposure equation), which can be stated:

Gm = sqrt(0.43 I•t).

I would endorse the use of the ISO equation.

More later.

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Martin,

Well, of course my comparisons were somewhat screwed up. I will get that straightened out shortly (now that I am well-fueled with Breakfast "A").

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Martin,

Well, here is a more careful story.

The Kodak equation (as I have it) is actually:

Gf = sqrt((I•t) S)/20)

where Gf is the guide number in feet, I•t is the luminous intensity time product in beam candlepower seconds (essentially identical to candela•seconds), and S is the ISO speed upon which the guide number definition is predicated.

We normally use the "ISO 100" basis, so this becomes:

Gf = sqrt(5 I•t)

Then, for G in meters, we have:

Gm = sqrt(0.46 I•t)

The ISO 1230 equation for ISO 100 is:

Gm = sqrt(5 I•t)

This is excellent agreement, especially given that the underlying "exposure equation" is rather arbitrary and there are various roundings involved, for convenience, in each equation.

My quickly-derived result of this morning is obviously cockeyed. I will reconstruct it.

Sorry for the confusion.

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Martin,

I haven't so far been able to find where I went wrong in my derivation. I'm sure when I find it, it will be a real "oh shіt".

And I've had to work on some other stuff, like trying to get my online banking at a new bank unscrewed-up.

But I'll get it.

In any case, I am confident in the ISO equation.

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Martin,

Well, it was when I divided 250 by 100 and got 25!

Using the value for C (the incident light metering constant) that best matches the exposure equation I use, and dividing correctly, I derive:

Gm = sqrt(0.45 I•t)

where Gm is the guide number in meters (ISO 100 basis, per the international standard) and I•t is the luminous intensity-time product, in cd•s.

I•t actually represents the time integral of I; its dimensionality is that of I•t.​

Best regards,

Doug
 

Martin Evans

New member
Hello Doug:

My word, you have been thorough in your analysis of photo-flash parameters! Your "Breakfast A" evidently possesses powerful mental factors - I was under the impression that your normal breakfast was simply oatmeal porridge.

As I have no details about the various factors behind Hanlon's flash unit, I am going to leave the comments in my essay to a simple repetition of Saunders' estimate that the flash was "75 times" more powerful than the ones used by Saunders.

Many thanks for taking such a keen interest. And thanks to Asher, Jerome and Bart for your information about the power of smartphone flash, which I have also added to my essay.

Martin
 
Top