Flash Gurus, help needed for Sigma ETTL

I have noticed FEC doesn't seem to work on SIGMA 500 DG SUPER on a 400D body. (Can't test on 350D, it's broken...) I've set the camera to manual, flash to ETTL mode and tested against a 580EX. The ambient light is at least -2 stop than correct exposure, so it should be a factor. I've set FEC to -1, 0, and 1, and the 580EX gives very obvious different in exposure, where the SIGMA 500 DG SUPER seem to always give the same exporue... Please see the attached photos. Is this a known bug?

canon -1 stop

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

that is odd. can you adjust FEC on the flash? or are you adjusting it on body?

FEC on the flash will override what is set on the body. Check that to be sure your sigma isn't set to +1 or some such thing, which will make everything bright no matter what your body is set to. If the Canon is set to 0 and the Sigma set to anything other than 0, the in-body FEC will appear to be operational on the Canon flash and defective on the Sigma.

cdifoto wrote in post #4443601
FEC on the flash will override what is set on the body. Check that to be sure your sigma isn't set to +1 or some such thing, which will make everything bright no matter what your body is set to. If the Canon is set to 0 and the Sigma set to anything other than 0, the in-body FEC will appear to be operational on the Canon flash and defective on the Sigma.

well done master flasher...youre right....

Thanks everyone for helping out. I'm baffled. I did not set FEC on the flash for the posted photos. Actually, I tried another round by setting the FEC on the flash, a special symbol will appear on XTi's screen. But the end result is the same, I got unreliable exposures. Is there any compatibility issue with new camera like XTi? Anyone using Sigma flash an a newer camera (newer than XTi) observe this problem?

cdifoto wrote in post #4443601
FEC on the flash will override what is set on the body. Check that to be sure your sigma isn't set to +1 or some such thing, which will make everything bright no matter what your body is set to. If the Canon is set to 0 and the Sigma set to anything other than 0, the in-body FEC will appear to be operational on the Canon flash and defective on the Sigma.

I have a 400D and Sigma DG 500 Super and setting the camera's FEC works as expected. Make sure flash is in ETTL mode with its FEC set to zero. Also set camera special function for ETTL to AVERAGE rather than EVALUATIVE.

The problem with these shots is that the ambient is the predominant light source. Try 1/200 f/8 ISO 100 and see how FEC works.

Curtis N wrote in post #4446636
The problem with these shots is that the ambient is the predominant light source. Try 1/200 f/8 ISO 100 and see how FEC works.

Thanks Curtis, you are always helpful as always. Appreciate the helps from everyone. The ambient is not the predominant light, though. I took care to make sure the ambient light is about -2 stops underexposed. In other words, if I turn of the flash, the picture is almost completely black. (the ambient is a light torch turned to its lowest)

I'm thinking maybe the sigma flash can't shut off fast enough? I'm using F1.4 and high ISO. Will give it a try with F8 and ISO 100.

Thanks for everyone's help. It turns out if I set ISO to 1600 and F1.4, altough ambient light is extremely low and doesn't contribute to the exposure, the flash duration required is so short that the Sigma cannot do it. It will always over expose as in the image. The 580EX is nice enough to do it.

The Canon Speedlites also indicate a minimum distance on the distance scale. The Sigma doesn't, so you have to guess at it.