I know this Speedlite is capable of better results than the disappointing series of photos I took at a Thanksgiving gathering. If not for Photoshop, half of them would have been unusable. I decided I had to do some serious testing to find settings that would yield consistent exposures in a variety of situations before my annual Christmas photo spree.
My camera is a Canon EOS 5D Mark II with a 24-105mm lens. The Speedlite allegedly can be used in point-and-shoot mode by setting the unit on ETTL and the camera on full auto. With this method I discovered the camera actually uses one setting regardless of distance: f4, ISO 400 at 1/60 second and the Speedlite adjusts automatically for distance. A lens aperture of f4 yields a very narrow depth of field; i.e.: only a small area of the picture will be in focus. I prefer to set the camera manually at f8 or f11, smaller apertures which result in greater depth of field. Allegedly, you can do this and the Speedlite will compensate, but in reality exposures are inconsistent, and maximum range is no more than 10 feet, a ridiculously short distance for a unit that's rated for up to 60 feet! Increasing lens aperture and ISO seems to have no effect beyond 10 feet.
I did a series of tests with the Speedlite and camera both set on manual and worked out a table of f-stop and ISO settings at 3, 6, 10, 15, 18 and 20 feet. This is the table:
3 ft. f22 200
6 ft. f16 200 or f22 400
10 ft. f11 200 or f16 400
15 ft. f8 200 or f11 400
18-20 ft. f5.6 200 or f8 400
I printed my table on a small reference card. Twenty feet is the greatest distance available in my house where I was doing all the testing, but these results indicate I could reach 40, 50, maybe even 60 feet with no problem. I took 11 pictures and all were near perfect. You can see the results at this link:
http://www.linwoodstreet.com/speedlite/manual/
The problem with this method is that it takes all the spontaneity out of my photography. My subjects will get impatient waiting for me to check the distance, look it up on my card and transfer it to the camera. They'll be yelling at me, "Will you hurry up and take the picture? We want to eat!" or "We want to start opening gifts!" I realized I had to find a means of getting accurate exposures with as little manual intervention as possible.
Next, I set the Speedlite up for external metering as described in the manual. You have to change a custom setting in the Speedlite, then set f-stop and ISO on the Speedlite to match the camera. The Speedlite will then achieve the correct exposure by varying the duration of the flash, longer for distant subjects, shorter for closeups. But it didn't work out that way. All photos taken with this method were considerably underexposed, much too dark. So it occurred to me that if I set the ISO on the Speedlite lower than on the camera, that should fool the Speedlite into extending the duration of the flash. This theory proved to be correct. After experimenting with several combinations, I settled on ISO 160 f8 on the Speedlite and ISO 400 f8 on the camera. I took 10 photos. One closeup was slightly underexposed and one mid-range shot slightly over-exposed, all others near perfect. This is acceptable and I'll go with this method for the time being. I don't object to making minor corrections now and then with Photoshop. You can see the uncorrected results of my final test at this link:
http://www.linwoodstreet.com/speedlite/external/
I was never satisfied with flash photos taken with my Nikon D70s with built-in flash. Colors were flat and lifeless, sometimes muddy. The Canon Speedlite produces more brilliant colors, but sometimes harsh and contrasty. More pleasing results can be obtained by bouncing the flash off a white or light-colored ceiling, which I do whenever possible. The ability to rotate the Speedlite head upward at an angle is a valuable feature. It can also be rotated sideways to bounce off walls. Summing up, it's a versatile and powerful flash, a very good performer if you have the patience to experiment exhaustively until you find the right combination to yield consistently accurate exposures.
A review of my camera can be seen at this link:
http://linwoodstreet.blogspot.com/2009/09/first-impressions-of-5d.html
