AE Lock on Canon and Flash

adbsgicom

Why does the exposure change if I don't recompose between using the AE lock and not using it if I bounce my flash? Here are two pictures, and you can see that the second one (where I pressed AE lock and got a test stobe) exposes brighter. Both are shot with the camera on manual (f/4, 1/60sec, ISO 200) about 2/3 stop under exposed ambient and the flash is set in E-TTL for +0 FEC.

No AE-Lock:


AE-Lock:


Any insights on exactly what is happening here would be appreciated. The metering mode doesn't seem to matter (tried all 4).

Snowgirl

My understanding is that AE lock doesn't work in Manual - but I could be mistaken (it has happened )

adbsgicom

It seems to be doing something, perhaps the firmware shouldn't allow it to do anything, but in some ways, I'm pretty sure it is doing something more right than wrong when used.

What I've noticed, is if I'm bouncing the flash and have it flagged to not spill forward, I may have to set the FEC up a stop more more in dim light. However, if I press the AE-lock button and fire the preflash, the I don't need to set the FEC on the flash (or if I do, I get an over-exposure).

adbsgicom

FWIW, the lock (* button) does lock the FEC based on the center, even in manual mode. What is odd is that regardless of my nominal metering mode, the use of the * button seems to cause the bounced flash to be just a hair brighter than when I shoot w/o it. Since I'm not changing my focus point, it seems a bit odd.

pyry

How fresh are the batteries in your flash? Is the ambient light made up of fluorescent tubes or energy saving bulbs?

This is just a first thought, but the only difference between locking the exposure first and not might be 1) allowing the flash to recycle from the metering pre-flash or 2) fluorescent ambient which isn't constant, it flickers.

adbsgicom

Batteries were very fresh, and in this case, the light was mostly from some halogens and sunlight. Even then, across all 4 metering modes, I had the same net effect which was wierd.

Qarik

adbsgicom wrote:

Batteries were very fresh, and in this case, the light was mostly from some halogens and sunlight. Even then, across all 4 metering modes, I had the same net effect which was wierd.

E-TTL measure distance if I am correct. My guess is the bounce is throwing things off.. if you are getting pre-flash in one instance and no preflash in another then one case the bounce flash is measuring the distance wrong. To test this you point the flash directly then I bet both exposures are the same.