What is the ISO flash Guide Number methodology?

~5500K would certainly make sense for a typical electronic flash.

There's actually a fair amount of spectral variation in flash units; more than likely D55 is selected as the standard daylight.

The big advantage of the D series of illuminants is, I believe, that they can be calculated.

Again, take all this with a slight grain of salt for any important purposes.

Until we reach relativistic speeds, a meter is a meter is a meter but I can promise you that the meter stick that you buy at the home center isn't accurate to the tenth of a wave length. Your kilogram of coke isn't accurate to the hundredth of a gram, most clocks loose or gain time and don't even get me started on electricity. Do you have a volt meter that is accurate to the electron? Have you even looked at schematic for a Fender guitar or bass amplifier? "All values +/- 20%. Find me a 100 ohm resistor that measures exactly 100 ohms?

No, everyday instruments and measures aren't as accurate as a laboratory standard reference, nobody expects them to be, but OTOH they're generally not out by + 100/ - 50% either, which is pretty common for a Guide Number 'specification'.

Each of these values are simpler than the flash one too.

In what way?

We don't have to measure the force between two parallel wires, for example, to get within a fraction of a percent of knowing the current in a circuit. All it takes is an affordable meter. Likewise with measuring the intensity of a pulse of light. It's not rocket science - or at least it shouldn't be.

Photographic exposure is measured in stops. A doubling or halving of light intensity or time. That's not real measurement, that's estimation! And ought to be pretty difficult to actually get wrong.

All I'm trying to get at is some idea of how Guide Numbers ought to be measured. Otherwise we're all just urinating into the wind arguing about it.

We never seemed to have this confusion and level of secrecy before the ISO took possession of standards. In my view, they've done next to nothing to introduce new meaningful specifications, and everything to obfuscate the procedures and measurements behind anything they've had a hand in. Thankfully, our everyday SI units were common knowledge before that bunch of secretive, bean-counting Swiss Gnomes took over.

Hooray for the Indian Government and its standards institute! An enlightened view on standards in an increasingly privatised and fiscally fixated world.

Having perused those Indian standards two points immediately leap out.

1) There's a tolerance of +/- 0.5 EV allowed straight away. (You can bet it's not a pessimistic number that gets published.)

2) The description - 'room with non-reflective walls' - is very open to interpretation. Does 'non-reflective' mean matt-black, and baffled to prevent any reflection of light as much as is practically possible? Or does it simply mean non-glossy and not allowing specular reflections? Because a matt-white painted room of tiny dimensions could be called non-reflective at a stretch.

What a sloppy way to specify a methodology! If this is typical of the ISO in action, then it confirms my worst suspicions.

And 38 Swiss Francs for something hardly worth the paper it's printed on? What a rip off.

No, everyday instruments and measures aren't as accurate as a laboratory standard reference, nobody expects them to be, but OTOH they're generally not out by + 100/ - 50% either, which is pretty common for a Guide Number 'specification'.

 

In what way?

We don't have to measure the force between two parallel wires, for example, to get within a fraction of a percent of knowing the current in a circuit. All it takes is an affordable meter. Likewise with measuring the intensity of a pulse of light. It's not rocket science - or at least it shouldn't be.

 

Photographic exposure is measured in stops. A doubling or halving of light intensity or time. That's not real measurement, that's estimation! And ought to be pretty difficult to actually get wrong.

 

All I'm trying to get at is some idea of how Guide Numbers ought to be measured. Otherwise we're all just urinating into the wind arguing about it.

 

We never seemed to have this confusion and level of secrecy before the ISO took possession of standards. In my view, they've done next to nothing to introduce new meaningful specifications, and everything to obfuscate the procedures and measurements behind anything they've had a hand in. Thankfully, our everyday SI units were common knowledge before that bunch of secretive, bean-counting Swiss Gnomes took over.

Visible Light is a continuum of wavelengths the energy of these varying per photons.in proportion to said wavelength.

Xenon flash tubes emit a wide range of wavelengths going well beyond the visible range on both sides (visible is roughly 400-700nm, xenon tubes emit over at least 200-2000nm)

Flash guns are usually filtered to block most of the energetic UV - The filter on the one I was working with today blocks ALL UV below 380nm and is up to full output by 405nm. This cut of will vary from flash to flash.

At the other end most flashes allow all the infra red to be emitted along with the visible light.

So we have a variable spectrum being transmitted which different meters will record to different extents. (Note human eyes also vary as to which wavelengths they can see. Younger people typically see further into the UV).

This is much more complicated than measuring voltage, current, weight or indeed monochromatic light.

Visible Light is a continuum of wavelengths the energy of these varying per photons.in proportion to said wavelength.

Xenon flash tubes emit a wide range of wavelengths going well beyond the visible range on both sides (visible is roughly 400-700nm, xenon tubes emit over at least 200-2000nm)

Flash guns are usually filtered to block most of the energetic UV - The filter on the one I was working with today blocks ALL UV below 380nm and is up to full output by 405nm. This cut of will vary from flash to flash.

At the other end most flashes allow all the infra red to be emitted along with the visible light.

So we have a variable spectrum being transmitted which different meters will record to different extents. (Note human eyes also vary as to which wavelengths they can see. Younger people typically see further into the UV).

Interesting, but totally irrelevant.

And telling us nothing we didn't already know, thanks.

All that we're interested in as photographers is the effect the flash has on film or a digital sensor, and with a digital sensor, the exposure can be seen immediately. It also agrees with and confirms the reading from any decent flashmeter.

So yes, determining the light output of a speedlight is as simple as pressing a button. On either a digital camera or a calibrated flashmeter.

This is much more complicated than measuring voltage, current, weight or indeed monochromatic light.

Total tosh!

As explained above, it's even easier than connecting a volt- or ammeter up to an electrical circuit.

Interesting, but totally irrelevant.

 

And telling us nothing we didn't already know, thanks.

 

All that we're interested in as photographers is the effect the flash has on film or a digital sensor, and with a digital sensor, the exposure can be seen immediately. It also agrees with and confirms the reading from any decent flashmeter.

 

So yes, determining the light output of a speedlight is as simple as pressing a button. On either a digital camera or a calibrated flashmeter.

 

 

Total tosh!

As explained above, it's even easier than connecting a volt- or ammeter up to an electrical circuit.

It's simple enough measuring it on a digital camera but the values WILL vary from one camera to another depending on the cameras internal cut filter & sensor charcteristics. The spectral output of the flash will alter how the differences are seen from camera to camera, so normalizing the cameras is not possible.

It's simple enough measuring it on a digital camera but the values WILL vary from one camera to another depending on the cameras internal cut filter & sensor charcteristics. The spectral output of the flash will alter how the differences are seen from camera to camera, so normalizing the cameras is not possible.

Look, a photographic exposure only has to be accurate to 1/3rd of a stop - and that's set down by the ISO itself. Plus or minus one third of a stop allows for a measurement tolerance of +26% or -20.63% in the linear value of illumination. That's as wide a target as hitting a barn-door with a blunderbuss at two paces.

The piddling difference between camera spectral responses - most of which use similar sensors manufactured by a very limited number of companies - is as nothing compared to that huge measurement leeway.

The same applies to the majority of films, which use similar sensitising dyes and have very similar spectral responses.

Not to mention that speedlights nearly all use a polycarbonate Fresnel and safety cover over the tube, which is very efficient at filtering UV. Considerable Infrared still gets through, but digital camera makers take great pains to fit effective IR filters over their sensors. And Sunlight itself emits well outside of the visible band, and has no UV or IR filters built in!

The fact remains that there's a sizeable (< -20%) difference between published 'specifications' and actual measured or recorded light output from most, if not all, speedlights. And this has been the case ever since the darned things first went on sale.

So, if the ISO can't be bothered to tighten the measurement methodology to preclude such discrepancy, then what use is their recommendation? What use at all?

The fact remains that there's a sizeable (< -20%) difference between published 'specifications' and actual measured or recorded light output from most, if not all, speedlights. And this has been the case ever since the darned things first went on sale.

You seem to have said this over and over through this thread . . . Yet, you have not provided a single example OR the test procedure that you used . . . I may just order the damned standard since the time that I have put into reading and responded to this thread are worth at least $38.52.

You seem to have said this over and over through this thread . . . Yet, you have not provided a single example OR the test procedure that you used . . . I may just order the damned standard since the time that I have put into reading and responded to this thread are worth at least $38.52.

I use a flash meter with a flat diffuser at 10ft from the flash head. The aperture value is then multiplied by 10 to get the GN in feet. I found that all of the flashes tested registered a lower GN than specs. While there should be tolerance but they are always less and not more so that makes me believe that the discrepancy is intentional.

I use a flash meter with a flat diffuser at 10ft from the flash head. The aperture value is then multiplied by 10 to get the GN in feet. I found that all of the flashes tested registered a lower GN than specs. While there should be tolerance but they are always less and not more.

At what height from the floor? In what size room? With what color walls? How did you calibrate your meter? If the flash has a zoom head, did you test at multiple settings? Did you test at multiple flash levels? How old is the flash? How fresh the batteries? What type of batteries? How long did you allow the flash to cycle/power up? How far off where the readings? All of this and likely more goes into a scientific test. I'm not surprised that "all of the flashes tested registered a lower GN than specs." Manufacturers aren't going to round down and don't want their flash systems known for blowing out highlights.

At what height from the floor? In what size room? With what color walls? How did you calibrate your meter? If the flash has a zoom head, did you test at multiple settings? Did you test at multiple flash levels? How old is the flash? How fresh the batteries? What type of batteries? How long did you allow the flash to cycle/power up? How far off where the readings? All of this and likely more goes into a scientific test. I'm not surprised that "all of the flashes tested registered a lower GN than specs." Manufacturers aren't going to round down and don't want their flash systems known for blowing out highlights.

You missed out how well aligned was the meter, this makes a huge difference with a flat diffuser...

At what height from the floor? In what size room? With what color walls? How did you calibrate your meter? If the flash has a zoom head, did you test at multiple settings? Did you test at multiple flash levels? How old is the flash? How fresh the batteries? What type of batteries? How long did you allow the flash to cycle/power up? How far off where the readings

Jeez! All these stupid little objections are like being given half the weight of potatoes you paid for, and then the greengrocer making excuses about a local variation in gravity being to blame.

"Oh gov'ner, you see we're at a few hundred metres above sea level and close to the equator, and besides, the underlying rock strata are less dense than average. Tell you what; I'll throw you an extra spud on the scales for good measure."

Doesn't anybody actually check their optimistically specified speedlights against maker's cheating figures?

BTW. The room colour and dimensions should actually increase the GN reading if the instructions to use a non-reflective environment for standard measurement are strictly followed. Which makers obviously ignore.

Stop making excuses for corporate cheats and liars.

I'm not making excuses for anybody . . . I'm pointing the problems with comparing tests outside of a laboratory with tests inside the lab. It's the stance that I have taken from the very beginning. What happens in the lab isn't always the same thing that happens outside of the lab.

If you can't, or don't want to handle that, I'm OK. Try to duplicate the 0-60 or 60-0 figures published for your car.

I'm not making excuses for anybody . . . I'm pointing the problems with comparing tests outside of a laboratory with tests inside the lab. It's the stance that I have taken from the very beginning. What happens in the lab isn't always the same thing that happens outside of the lab.

 

If you can't, or don't want to handle that, I'm OK. Try to duplicate the 0-60 or 60-0 figures published for your car.

My point is that I accept variation but for GN it's always on the low side which makes me believe they are intentionally do it that way.

. . . and all that I have said, all along for however long we've been doing this, is that it is not a surprise because the laboratory standard is likely designed to get the maximum guide number possible. Why? I can think of two reasons but the answer isn't really relevant. First, it makes the flash look for powerful. This sounds nefarious but really isn't. Second, the manufactures always want to over estimate the power of their flash units because underestimating the power would result in more over exposed images with washed out highlights resulting in poor reviews from users relying on the guide number to make their exposures.

So . . . Maybe they DO intentionally do it this way, with good reason.