7D - 580EX II - Bounced Wirelessly - ETTL failure?

[Andy0x1]

I just got off the phone with someone at Canon. They were very helpful, but unable to repdocue the problem using the same equipment combination.

Hopefully the replacement will resolve this.

 

[Andrew Payne]

I've updated my Wiki page with some new test images that show the problem, see: http://www.payne.org/index.php/Canon_7D_Wireless_Flash_Exposure_Problem

Comparing the images a little more diligently, I see the problem with my 430EX and my 580EX as well (previously, I just thought it was with the '580).

I also think that the degree of overexposure is very dependent on how the scene is illuminated by the flash. Ceiling bounce seems to be the worst case.

-andy

 

[Aboy]

No flash was mounted on camera. All 3 were on stands. Sorry.

 

[apersson850]

As far as I understood from Canon, it's not all cameras that are affected by this error. If you are lucky you get one that works.
--
Anders

 

[Andy0x1]

Update - After speaking with Canon tech support, they were unable to re-produce my problem and figured I may have had a bad copy of the 7D. I exchanged the camera earlier today for a new one.

Unfortunately... the new one has the same ETTL problem... I guess it isn't too terribly surprising as they were probably both from the same production run / batch.

Ending on a positive note, I no longer see a 'hot' pixel when playing back SD video on this body.

-A

As far as I understood from Canon, it's not all cameras that are affected by this error. If you are lucky you get one that works.
--
Anders

 

[apersson850]

Canon said to me that I could go to the dealer and ask for a replacement, but they also were fair enough to admit that it may not solve the problem, as obviously there are several (have no idea how many) cameras that are haunted by this problem.
--
Anders

 

[Andy0x1]

The new Firmware 1.1.0 appears to have resolved this for myself and others.

Thanks everyone.

 

[ScratchDisk]

Great news - I was beginning to worry that it was somehow hardwired in. I hope we hear from Anders and that he is sorted too. Have Canon given any more info about this problem? I am very keen to hear all the detail on it that I can find. Tx for posting - I bet it has lifted a little depression! Cheers, Donald

 

[apersson850]

I'm not so sure.

It's not that my camera doesn't work with the 1.1.0, since it does so far. Have only had it in there for a day, but so far, so good. Seems to do just as well (or sometimes bad) as we have come to expect from E-TTL II.

BUT, and that's the big BUT, I shot some test images prior to installing 1.1.0, since I wanted to be able to compare the result with 1.0.9 with whatever the new firmware would give. But apart from one totally overexposed image, the camera exposed fine! So the new firmware seems to have healed my camera even before I got it installed.
Talk about the magics of programming!

Had I not shot these test shots just prior to upgrading, I would have praised the new firmware as well. Now I just don't know if it's a coincidence or really Jesus that has returned.
--
Anders

 

[Russell Williams]

The slave flash(es) are extremely sensitive to being able to see the on-camera flash directly. In particular, if a slave flash is behind the camera, even if there's plenty of reflection off walls so that the slave gets triggered, I usually get severe overexposure. If the slave is in front of the camera with the sensor aimed at the on-camera flash, the exposure is fine.

Canon's system is apparently not as robust in this regard as Nikon's.

 

[Russell Williams]

Interesting -- if I hold the slave directly in front of the on-camera flash with the sensor pointing back at the camera, the problem goes away for me.

 

[apersson850]

Well, when there's an unobstructed line-of-sight between the camera's internal master flash and the slave, and the slave is well within the range Canon specifies for indoor use, it should work.

If it doesn't, there's something wrong.

Had the slave been around a corner and further away than Canon says it would work with, even if within sight, then I wouldn't complain if it didn't work. After all, in that case they said it wouldn't work.
--
Anders

 

[Andy0x1]

While I beleive you that you were unable to reproduce the problem on 1.0.9 as of late - I question that you were using an 'affected' lens combination, as for whatever reason, it does seem to make a difference.

Using a combination that failed with 100% certainty on both 7D's I had no longer fail with 1.1.0

-A

 

[apersson850]

I used lenses which gave the problem earlier. With 1.1.0, the problem remains when using FEL.

Anyway, this is now a moot point. Canon Sweden has today told me that Canon Japan have issued a service bulletin, where they address this problem. Affected cameras have to go to one of Canon's service centers to have the flash controller module in the camera replaced.

This service bulletin is very recent, so all service centers may not be up to speed on it yet. Neither do they perhaps have the necessary part.
--
Anders

 

[ScratchDisk]

Tx for posting. Does this info appear on any Canon Website yet?

 

[apersson850]

I don't know. They called me on the phone.
--
Anders

 

[barrosinc]

1.1.0 Solved this issue for me.

Thanks!

 

[ScratchDisk]

Hi, Thanks for this extra info. (My 7D arrives tomorrow!)
By 'this issue' do you mean, something along the lines of;

" ..If I use the internal (built-in) master, it works almost always (98%) with the slaves aimed directly at the target, but almost never (2%) with bounced slaves.. "?

..and for you, under those exact circumstances, with the new firmware, is your multiflash, ETTL/Autoflash exposure (slaves in bounce) 1) reproducible to 98% or better and 2) accurate to within 2/3rds of a stop or better? I could live with those. Tx for considering my questions. Cheers

 

[barrosinc]

I had the problem if the flash was bounced i was getting overexposure (a lot!)

Now if i bounce I dont get that issue... works flawlessly

 

[ScratchDisk]

Thanks very much. My 7D arrived early - with 1.0.9 on board. With an 17-55 2.8 EF-s, I dont seem to be getting flash exposure problems with built-in as controller and 550 bouncing as its slave. So far anyway. I suppose I should try an 580(11). I don't plan to upgrade the firmware just yet - more testing first. Tx