ADI only works with D lenses. I don't think that 50mm f/1.7 was made as a D lens, but even if it was the majority of them will not be. Mine certainly isn't.
I have to agree. I have never been able to tell any difference either. In theory ADI should be better in 'difficult to meter' situations - very light or dark or reflective subjects etc. I have never ran any tests on it, but in general use I have never noted it to be any better.
Actually, I take that back. Promted by this thread I have just pointed my camera at an object behind glass, such that the reflaction of the flash would be in shot. ADI got the exposure about right. TTL underexposed as it over compensated for the reflection of the flash. It does work.
If you use any sort of light modifier on your flash (e.g., a diffuser) you do not want to use ADI, as the camera has no way to know that you are altering the effective output of the flash. ADI only works with the flash aimed straight ahead at the subject when mounted on the camera. In theory, you should not see much difference with most subjects. It is intended to compensate for unusual circumstances that could affect the flash metering, such as a highly reflective surface (e.g., shooting with a mirror behind the subject the reflects the flash back towards the camera). ADI is not the primary flash metering method, even when on, but rather is a supplement to the TTL flash metering to help compensate for unusual situations. ADI is automatically disabled if you use the flash in bounce mode, or off the camera.
I believe that it will default to TTL if ADI is selected in the menu, but a non-ADI lens is attached. I also think that I read somewhere that it will do the same if an external flash is set to wireless or bounce (both of which will not work with ADI), but I could be wrong about that.
