580EX II 14mm zoom more powerful w/ modifiers?

Hey folks,

The King Measurebator is back, with new and exciting data!

Actually this has me scratching my head a bit. When testing various flash modifiers with my 580EX II, I noticed that using the 14mm zoom setting (wide panel flipped down) gave me more light than any other setting. Furthermore, there is no linear relationship between zoom setting and exposure.

My testing was done indoors (diagram below) with the flash meter 10 feet away. Flash readings are converted to EV for simplicity.

Check out the data and let's hear your theories.

How was the meter positioned? It was in incident mode, correct?

Curtis N wrote in post #12675183
When testing various flash modifiers with my 580EX II, I noticed that using the 14mm zoom setting (wide panel flipped down) gave me more light than any other setting.

Except for the Lumiquest Softbox which beats 14mm starting at 35mm. This setup is just like direct flash, all forward light. The flash head is horizontally positioned putting the light almost inline with the meter axis resulting in higher readings.

My theory in this experiment is all about the forward component that the meter is picking up. The 14mm setting completely fills the bounce card surface which produces greater forward component. The vertical no-card setup at 14mm can easily hit walls or other surrounding objects bouncing light (forward component) close to the meter axis resulting in a higher reading than the other zoom settings.

Incident metering, dome out, pointed directly at the flash.

SFzip wrote in post #12675788
My theory in this experiment is all about the forward component that the meter is picking up. The 14mm setting completely fills the bounce card surface which produces greater forward component.

Your theory doesn't explain why the readings drop at 24mm but then rise again from 24mm to 35mm.

I'll have to try that next time I use my shoot through.

I never use any of these small accessories but with real modifiers like softbox, umbrellas etc. 14mm setting only gains very small uniform coverage at cost of lot more light loss. So I just use 24mm setting.

Interesting tests But your results are not surprising.

Of course, changing the zoom setting doesn't change the amount of light that is put out, but it changes the angle of distribution and that is changed again by the reflection and diffusion characteristics of the attachment, and yet again by the surroundings which play a huge part in the effectiveness of these things, including the walls of course which you haven't mentioned. In particular, the amount of light that goes directly forward is varied a lot, and the ratio of ceiling to direct light, and the quality of that component.

For example, I use a Lumiquest Quik Bounce (IMHO the best flash attachment by miles, similar to the 80/20 but not quite the same) and if you zoom the head back to 24mm it fills the reflector right to the edges and the direct fill-in component is slightly softer. More significantly, if you zoom out to 105mm, roughly a stop more light goes out of the top (and therefore a stop less out of the front) and this allows a useful degree of control with higher ceilings.

Tests of flash diffusers are almost always flawed because they don't consider the effect of different surrounding which is almost always the major player, and in your case you have not assessed the type and quality of the light falling on the subject, which is what matters, but only the amount of it.

And as you've discovered, there is not a linear relationship between exposure level and zoom head settings. They do not follow the inverse square law directly, or even closely, as you might perhaps assume, due to the inefficiency of the fairly crude fresnel lenses in the zoom head.

Hoppy, thank you for your input.

Hoppy1 wrote in post #12676483
you have not assessed the type and quality of the light falling on the subject, which is what matters, but only the amount of it.

The type and quality of light falling on the subject only matters if the quantity is sufficient to compete with the ambient and affect the exposure. Of course if your flash hasn't recycled and doesn't fire when you take the shot then it doesn't matter what attachment you have. So we need to have realistic expectations of the capabilities of a given setup. The data shown is a small portion of an ongoing experiment which will be discussed later on.

The inverse square law is irrelevant here. I'm just trying to make sense of the results. When you zoom from 14mm to 24mm to 35mm I would expect the exposure to trend one direction or another (not necessarily linear). Instead it starts high, goes down and then goes back up. That's the part that has me baffled.

Curtis N wrote in post #12676603
Hoppy, thank you for your input.The type and quality of light falling on the subject only matters if the quantity is sufficient to compete with the ambient and affect the exposure. Of course if your flash hasn't recycled and doesn't fire when you take the shot then it doesn't matter what attachment you have. So we need to have realistic expectations of the capabilities of a given setup. The data shown is a small portion of an ongoing experiment which will be discussed later on.

The inverse square law is irrelevant here. I'm just trying to make sense of the results. When you zoom from 14mm to 24mm to 35mm I would expect the exposure to trend one direction or another (not necessarily linear). Instead it starts high, goes down and then goes back up. That's the part that has me baffled.

As a starting point, take shots of a plain wall with the gun firing direct, without any attachments, at different zoom settings. Note how the size and shape and brightness of the light pool changes, and how it changes dramatically when you flip down the wide-panel. Also note the hotspots.

Then fit the diffusers and take wide-angle pictures from the back of the room, so you can see the shape and size and brightness of the light pool on the ceiling as well as the combined bounce/direct light pool on the wall. That will give you a good idea of where the light is going.

Hoppy1 wrote in post #12676483
In particular, the amount of light that goes directly forward is varied a lot, and the ratio of ceiling to direct light, and the quality of that component.

This. My guess is that the flash filled up the diffuser, which pushes some of the light forward (for the meter to pick up).

Fantastic test!

I knew it all along, intuitively... I've been testing these diffusers unscientifically for about 4 years and tried them all.. Surprisingly you didn't include any of Mr Fong's Tupperware in these tests, maybe for a good reason...

I think the Lumiquest Softbox would probably come third, everything being equal.. and the Lumiquest White, fired upwards, would probably top it.. Indoors anyway..

The one i have found to be the best and been using all this time is a DIY one which you probably have seen before... http://super.nova.org/​DPR/DIY01/.
I think the Lumiquest White is copied from that and mass-manufactured,,, It's, to my mind, the only one that has the best combination of Softness, No Wastage, Cheapness , Durability and Practicality.

Again, i can't scientifically prove it, but the scoop shape really helps to throw the light forward and around objects... Does light 'Speed Up' when it hits the sides of the scoop? Maybe, but your tests are showing something of that nature...

I am following it with great interest...

The Shaheen wrote in post #12678546
you didn't include any of Mr Fong's Tupperware in these tests,

I would love to test some tupperware. I don't own any but I have a friend who does. If I can get him to participate in exchange for a few beers, I'll add that data.

Curtis N wrote in post #12676122
Your theory doesn't explain why the readings drop at 24mm but then rise again from 24mm to 35mm.

The built-in wide panel is essentially a diffuser lens that spreads the light to 14mm. The mechanical "zoom" action is only between 24mm and 105mm.

The 14mm setting is really a diffuser lens placed on top of the flash, auto zoomed at 28mm.

14mm = 28mm + diffuser lens.

SFzip wrote in post #12679030

The built-in wide panel is essentially a diffuser lens that spreads the light to 14mm. The mechanical "zoom" action is only between 24mm and 105mm.

The 14mm setting is really a diffuser lens placed on top of the flash, auto zoomed at 28mm.

14mm = 28mm + diffuser lens.

I understand all that. But it doesn't explain the change in exposure from 14 to 28 to 35.

Curtis N wrote in post #12676122
Incident metering, dome out, pointed directly at the flash.Your theory doesn't explain why the readings drop at 24mm but then rise again from 24mm to 35mm.

I am not very technical, but..

Is that because the flash has gained its full power (High Exposure Value) by 24 mm, then light speeds up between 24 and 35mm, then it evens out between 35 and 70 mm, and then starts dispersing into the larger area of the room, then dissipates between 70 and 105mm..

If you note, the highest EV value is at 14mm region (where it's almost bare flash explosion), after that it's basically spreading around the room... don't forget these are diffusers doing their job, so by the time it's 105mm, it's gone pretty soft..

There's the Power of flash, Density and its Speed at play here..