Fill Flash Test - dome vs bender

For grins, I ran a test today as sunset approached - shooting in the shade at f8, 60 seconds, with a Canon R5, Canon 580EX flash, and a couple of different flash modifiers. Grumpy subject is about 6 feet away. I have an event tomorrow at about that time of day I am shooting, so thought this would be a good way to figure out what I'll use. In the bottom center is the configuration as I took each pic. I threw in shots with no flash, and flash but no modifiers as the first two shots for reference.








I think I prefer the last one - "dome" modifier with the flash straight up. I kinda anticipated I would like the "bendable panel" modifier, as it seems like it would have a bigger flash surface, so am slightly surprised by the result. And surprisingly, just popping that little "business card thingy" out on the 580 does pretty darn well for being so small and simple.

The last one exhibits the softest under-chin shadow. Which means the shadow under the nose and eyelids is probably softer too.
For this short of shot, it is often the appearance of those types of features that determine how effective the fill is.
If you had also tried using a large non-flash reflector held low and reflecting up at the subject, you probably would see some similarities.

Fantastic that you did some testing for yourself! With digital, so much can be tested 'for free' that there is no reason not to be able to learn from the testing.

For all shots the same fundamental principles apply...
Any light aimed upward is wasted power with zero contribution to illumnation of the subject...
...unless some of the light is reflected forward to the subject
The apparent AREA of the reflector forward (or light emitted directly forward by a translucent surface) is what is the key characteristic deterining the softness (freedom from cast shadow edges) of the light striking the subject...more AREA is softer

Based on those principles,

• I would expect photo 6 to project the softest light to the subject (a wrought iron plant stand, allowing you to better see cast shadow edges)
• Photo 7 modifier would be next softest, but waste a lot of flash power upward, using up battery charge because nTTL does not see as much coming back from the scene ...it suffers from the same criticism made for decades about the Lightsphere...don't bother outdoors, reserve its use to indoor ceiling bounce
• Photo 2 would have the harshest light
• Photo 3 would have the next harshest light
• Photo 4 and 5 have very similar harshness of light (less than in photo 3) and photo 5 would project just a tiny bit more intensity forward to the subject

To illustrate the benefit of more area, both these photos use a 'softbox' at 20' from subject...
35 sq.in. (5"x7")


320 sq.in. (16"x20")


Note the relative harshness of the shadows cast by the 35 sq.in. modifier.
(BTW, the larger softbox cast light that is more yellow, because the cloth is old and yellowed (30 year old softbox!) while the smaller softbox did not experience yellowing of its cloth


In view of the short timeframe until your shoot, I would use the modifier as shown in photo 6, for optimized battery drain combined with softness of projected light to the subject, but dial in FEC of about -0.66EV to -1EV so the artificial light is less obvious...which is why the lighting of photo 7 is more pleasing than photo 6, less light (but NOT any softer) is going forward in photo 7!

wiltw said:

Fantastic that you did some testing for yourself! With digital, so much can be tested 'for free' that there is no reason not to be able to learn from the testing.

For all shots the same fundamental principles apply...
Any light aimed upward is wasted power with zero contribution to illumnation of the subject...
...unless some of the light is reflected forward to the subject
The apparent AREA of the reflector forward (or light emitted directly forward by a translucent surface) is what is the key characteristic deterining the softness (freedom from cast shadow edges) of the light striking the subject...more AREA is softer

Based on those principles,

• I would expect photo 6 to project the softest light to the subject (a wrought iron plant stand, allowing you to better see cast shadow edges)
• Photo 7 modifier would be next softest, but waste a lot of flash power upward, using up battery charge because nTTL does not see as much coming back from the scene ...it suffers from the same criticism made for decades about the Lightsphere...don't bother outdoors, reserve its use to indoor ceiling bounce
• Photo 2 would have the harshest light
• Photo 3 would have the next harshest light
• Photo 4 and 5 have very similar harshness of light (less than in photo 3) and photo 5 would project just a tiny bit more intensity forward to the subject

To illustrate the benefit of more area, both these photos use a 'softbox' at 20' from subject...
35 sq.in. (5"x7")

320 sq.in. (16"x20")

Note the relative harshness of the shadows cast by the 35 sq.in. modifier.
(BTW, the larger softbox cast light that is more yellow, because the cloth is old and yellowed (30 year old softbox!) while the smaller softbox did not experience yellowing of its cloth


In view of the short timeframe until your shoot, I would use the modifier as shown in photo 6, for optimized battery drain combined with softness of projected light to the subject, but dial in FEC of about -0.66EV to -1EV so the artificial light is less obvious...which is why the lighting of photo 7 is more pleasing than photo 6, less light (but NOT any softer) is going forward in photo 7!

Click to expand...

Thanks for the insight! I wish I would have also shot a test of the lightsphere facing forward, but without the dome... so the flash is able to shoot straight away through the center. Just curios to see how much less harsh that would be than straight on flash. One thing I didn't test is the impact on distance - I have some memories that shooting the lightsphere straight up has noticably less reach than shooting with it forward.

bud007 said:

Thanks for the insight! I wish I would have also shot a test of the lightsphere facing forward, but without the dome... so the flash is able to shoot straight away through the center. Just curios to see how much less harsh that would be than straight on flash. One thing I didn't test is the impact on distance - I have some memories that shooting the lightsphere straight up has noticably less reach than shooting with it forward.

Click to expand...

Certainly, leave the lid/dome on even though aiming the head forward, tonight! The lid/dome increases the apparent area while use of the same modifier without lid leaves the area the original size of the flash lens/reflector.

I once made this in jest, only to prove the point that one did not have to spend much money

...and it is collapsible for transport! Here is a shot taken with it...

...vs. without (straight flash)

bud007 said:

For grins, I ran a test today as sunset approached - shooting in the shade at f8, 60 seconds, with a Canon R5, Canon 580EX flash, and a couple of different flash modifiers. Grumpy subject is about 6 feet away. I have an event tomorrow at about that time of day I am shooting, so thought this would be a good way to figure out what I'll use. In the bottom center is the configuration as I took each pic. I threw in shots with no flash, and flash but no modifiers as the first two shots for reference.

Click to expand...

First of all, congrats to the test and for the general idea testing your set-ups. That is something which I often explain in my workshops, too.
Knowing the strengths and limits of the equipment you are using - extremely important.

I am using fill-in flash regularly, both for film and digital.
And your shot No. 2 is probably the most important shot of the whole series, because it demonstrates something which is very important to know in general:
This shot is the less attractive of all flash shots in your test series, because it is overexposed. It is much more than a "fill-in" flash that improves shadow detail.
The reason for it is that most camera manufacturers decrease the amount of light by only 2/3 stops in the fill-in / balanced light function.
As the shadows in a scene are the areas with -1, -2 and -3 stops, this amount of light reduced by only -2/3 stops is simply too much.

With a perfect fill-in light you want to give the shadows just so much more light, that you have the needed visible detail in the shadow zones. But the shadows should remain shadows, it is not intended to fully remove them and put them on Zone V.
Therefore I can highly recommend from my experience to test the behaviour of your system, and make tests without modifiers, but with manually reduced flash power of -1 stop, -1 1/3, -1 2/3, -2, -2 1/3, - 2 2/3 and even -3 stops.
Look at the results, at the shadow detail. And you get a feeling for the amount of light needed for different scenes as well.

When I use fill-in flash with my cameras and flash system, I am working mostly in the range of - 1 2/3 to - 2 2/3 stops manually reduced flash power. It depends on the used medium (digital, positive film, negative film, special film characteristic), and on the aesthetics, how much more detail I want in the shadows in a certain scene.

Best regards,
Henning

I would prefer to use direct flash like photo #1 but increase the ambient light exposure and decrease flash exposure a lot.

Event went well. Sunny day here and the event started 2-3 hours before sunset, and is set outdoors on a wooded ridge. I was thinking I might not use the flash much for an hour or two, but I noticed when I got there there was this intense dappled sun everywhere (leaking through the rather heavy tree canopy). Hard to find a spot where people didn't have sun splotches all over them, and it's not the sort of environment (i'm walking around doing candids of the goings on) where I could really take the time to arrange and pose people too much. Anyway, I used the fill flash from the get go to fight those sun/shadow splotches and it more or less worked pretty well at that task.

Couple of other musings from the night...

For grins at the end of the evening (an hour past sunset), I did some 25,600 ISO natural lit shots. This is the first time I've used DXO RAW to clean up noise, and I was pretty impressed. A few of those images made the cut, and definitely helped capture the spirit of the evening.

My standard habits these days... I have the ISO at 800 and I'm in Program mode, and using back button focus. More and more, especially when I think I'm gonna have to rush a shot, I've found myself experimenting with Auto mode on the (Canon) 5R, which means I'm letting it take a crack at figuring out what to focus on, in addition to ISO, aperture and shutter. One of the shots tonight that blew me away was of a young girl getting her face painted, which was done in Auto, indoors, with the flash. Her head was turned a quarter, and the 5R found her close eye and nailed the focus, and because it was wide open, everything on the fair side of her face was soft... resulting in really cool look (well... at least to me...). Here's the eye and nose part of the pic:



Oh, and I also shot a roll of film, but I won't talk about that here

Glad to hear that worked for you. BTW, that posted photo has some severe color artifact showing up around her eye

I just ran a test to illustrate that there is nothing special about Canon flash 'in fill mode' vs 'full contribution' mode...

Indoors, illuminated only by window light, incident meter set for ISO 800 read 1/60 f/2.8 for 'proper' exposure

• I set a Metz 64MZ on Auto mode mounted on Canon 7DII, and 1/60 and/or f/2.8 on lens, I ran the camera thru M, Av, Tv, and P ...in every case the flash fired.
  I set the Metz 64MZ on eTTL mode still mounted on the Canon 7DII, and 1/60 and/or f/2.8 on lens, and ran the camera thru M, Av, Tv and P...and in every case the flash fired.
  In every case, the flashmeter read the flash output at subject position, and in all cases the metered amount of light was virtually identical...f/5.6 + 0.5EV
  
  The flash output was ALWAYS the same, no matter what Mode the camera was in, and no matter if eTTL calculated the exposure or the flash photosensor calculated the exposure!
  And in spite of the fact that there was sufficient ambient light for 'full exposure' and the flash was obviously providing fill to the window light, the meter reading was the same and indicated that ambient+fill would be +2.5EV overexposed.

Outdoors, illuminated by partial sunlight, incident meter set for ISO 100 read 1/200 f/10 for 'proper' exposure

• I set a Metz 64MZ on Auto mode mounted on Canon 7DII, and 1/620 and/or f/10 on lens, I ran the camera thru M, Av, Tv, and P ...in every case the flash fired.
  I set the Metz 64MZ on eTTL mode still mounted on the Canon 7DII, and 1/200 and/or f/10 on lens, and ran the camera thru M, Av, Tv and P...and in every case the flash fired.
  In every case, the flashmeter read the flash output at subject position, and in all cases the metered amount of light was virtually identical...f/11
  The flash output was ALWAYS the same, no matter what Mode the camera was in, and no matter if eTTL calculated the exposure or the flash photosensor calculated the exposure!And in spite of the fact that there was sufficient ambient light for 'full exposure' and the flash was providing 'fill' to the sun no matter what Mode the camera was in (M, Av, Tv, P), the meter reading was the same and indicated that ambient+fill would be +0.33EV overexposed from the no-flash shot.

So the flash programming from Canon does compensate the flash output so the shot is not significantly overexposed by the double contribution of light (ambient + flash), while use of flash in the Auto photosensor mode did not have such a decrease in flash output. And the camera in Manual mode caused no different behavior than all of the exposure automation modes in the camera, in terms of flash output.

Since Canon in the past has included features called NEVEC (Negative Evaluative Exposure Compensation) and called AFR (Auto Flash Reduction), it is appropriate to determine if NEVEC or AFR was influencing the flash output...the ambient-only exposure indoors is EV9, and outdoors is EV14

NEVEC table


NEVEC+AFR table



And we can see that AFR does not influence shots at EV9, but there is a NEVEC effect that applies at ISO 400 EV8, but no effect from either was expected at EV14

For those who never heard of either,
Here are the photos from the outdoor test. Manual no flash, Manual with flash, Tv, Av, Program mode (all using eTTL flash) from left to right. Anyone see a difference in the brightness of the wall behind, that would indicate a difference in eTTL output based solely upon which automation mode for exposure had been set (with flash providing primary illumination source vs. 'fill' illumination?

I use my Nikon SB800 either with a dome or as a bounce flash. One needs to be conscience of the wall color when using bounce flash with color film.