7D wireless flash trigger can't expose

[ccp900]

ok i did some more tests and this time a bit more controlled.

cant post pictures though since isp is getting wonky.

anyways heres what i did.

1) got 2 pieces of white typewriting paper
2) got 2 pieces of typewriting paper and i filled it with black, only borders are white. so for our use we can consider this a black paper.

setup.

manual camera - so exposure wont change - camera also at 10-11 feet away.
settings are : 1/30 f4.0 iso800 evaluative metering
background is a light brown sofa.

i used 3 af points : extreme left, center, extreme right....

ok first pass

580ex on hotshoe - pointed forwards
1) Using center af point

a) center af point - on black - correct exposure
b) center af point - on white - correct exposure

2) Using extreme right af point

a) right af point on black - center on black - correct exposure
b) right af point on white - center on black - correct exposure
c) right af point on black - center on white - correct exposure
d) right af point on white - center on white - correct exposure

3) 580 ex on hotshoe bounced

a) using center af - af on black - correct exposure - i stopped here...i think we all know it works when its on the hotshoe.
second pass

Wireless flash pointed forwards - flash was moved 2 feet to the right of the camera
1) Using center af point

a) center af point - on black - correct exposure
b) center af point - on white - correct exposure

2) Using extreme right af point

a) right af point on black - center on black - correct exposure
b) right af point on white - center on black - correct exposure
c) right af point on black - center on white - correct exposure
d) right af point on white - center on white - correct exposure

3) using extreme left af point

a) extreme left af point on black - center on white - right on white - correct exposure
b) extreme left af point on black - center on black - right on white - correct exposure

3rd pass

Wireless flash - bounced - flash still at 2 feet to the right of the camera
1) Using center af point
a) center af point - on black - overexposed
b) center af point - on white - correct exposure

2) Using extreme right af point

a) right af point on black - center on black - overexposed
b) right af point on white - center on black - correct exposure
c) right af point on black - center on white - correct exposure
d) right af point on white - center on white - correct exposure

3) using extreme left af point

a) extreme left af point on black - center on black - right on white - correct exposure
b) extreme left af point on black - center on black - overexposed

so it appears that the problem manifests itself when
a) the flash is on wireless bounced configuration
b) i think what the camera is doing is pretty much how the engineers programmed ETTL2. im assuming that the af points are part of the flash metering zones being used by the system to determine correct flash exposure.

when the system sees that there is this black thing in one of the zones, it quickly thinks that this is probably the subject that i need to meter for because all the other points are constant so when it averages the flash exposure compensation it overexposes because of the black subject.

BUT when it also sees another white subject, meaning the camera now sees a black subject on 1 zone and 1 white subject on another zone and the rest constant, it averages down to the correct exposure because it thinks there are 2 subjects with different tonalities that it has to accommodate.

i think the solution is to just put that wireless flash pointed forwards...if you wanna bounce it just point it on the ceiling...come to think of it...i didnt test that configuration. i should have darnit!!

[tdodd]

Originally Posted by ccp900

i think the solution is to just put that wireless flash pointed forwards...if you wanna bounce it just point it on the ceiling...come to think of it...i didnt test that configuration. i should have darnit!!

Might not the solution be to identify the circumstances in which the camera regards the "subject" as dark, and prone to overexpose, and then dial in some -ve FEC? FEC is there to be used after all. That said, so far I have been unable to force an overexposure from my setup, despite following all the suggestions to cause the problem.

[siberslug]

Originally Posted by timbop

they told me the same thing; it's their canned answer. I'm not going to, I'll wait until they figure it out

I just bought mine, I'll just exchange mine and hope for a good one.

[ccp900]

Originally Posted by tdodd

Might not the solution be to identify the circumstances in which the camera regards the "subject" as dark, and prone to overexpose, and then dial in some -ve FEC? FEC is there to be used after all. That said, so far I have been unable to force an overexposure from my setup, despite following all the suggestions to cause the problem.

yep, youre absolutely correct. but of course we want things to come out correct since the system was supposed to be smart enough to handle the scenarios......also, its pretty much a guess when its gonna actually happen so you might end up with a great pic overexposed to smithereens and you cant salvage.

my own thoughts when it comes to bouncing to set the flash to manual so it just pumps out constant power everytime, youre bouncing the light anyway.

[apersson850]

I can clarify on a few things mentioned in the more recent posts above.

Yes, the master flash always fires when the picture is taken, even when set to "off". This is to sync the slaves, so they fire at the correct moment.
It's a very weak flash, but in reflective surfaces, like a mirror, you can see it.
The strength is such that it has absolutely no impact on the issue I've reported here.

For all those suggesting that it may be a setting, it may be flash in group C, it may be I don't know how many different things, please return to post #2 in this thread. There you see two images, taken with exactly the same setup and settings (lens, distance, aperture, ISO and all), except one has a 580 EX II as a master, the other the built-in flash. That ought to make it clear that it's which master you use that's the key to this problem. But once again, it doesn't seem to be every camera that's affected. It seems to be some timing issue, perhaps affected by component tolerances in the camera.
I could determine that in the overexposed image, the flash is firing at 1/1 power, not about 1/8 as it should have done, to give the correct exposure.
Since these two images are framed and focused the same way, it has nothinig to do with E-TTL, focus point priority or any other such thing either. Also, since it doesn't set the flash power correctly, there is no point in identifying setups where this will be a problem and then use FEC to fix the overexposure. It doesn't react to the FEC, so that doesn't change the overexposure, except for just a little when you reach -3.

My camera too works with some setups, under some conditions, sometimes, with the internal flash as a master, but most often it fails.
Meanwhile, I'd like to say thank you to those who have spent time testing for this issue, to get a feeling for how many cameras are plauged by this problem.

[int2str]

Originally Posted by apersson850

My camera too works with some setups, under some conditions, sometimes, with the internal flash as a master, but most often it fails.

Strange that yours is so different. I had one or two shots fail in hundreds attempted, yet yours fails "most often"?

Is this doing stages tests or "real world" shooting?

[apersson850]

I've given up "real world", since it just spoils pictures I want. So now it's mostly testing taking photos in situations that could have been real world.

I keep on repeating that it's different from one camera to another. Which is why I think it has to do with manufacturing tolerances being a tad too large for something.

[windpig]

Somebody early on said they were seeing issues with a slower lens so I came home tonight and compared 24-105mm f/4 to 24mm f1.4
No comparison, 2 speedlights set to A:B 1:1 bounce were fine with the faster lens, the slower prime was over exposed throughout any thing over 100 ISO.
Subject had black and white, focused either places. I think it's like the AF system, faster lens, better focus.

[apersson850]

Right now, I get the same good flash exposure with two flashes as slaves (A and B), regardless of whether I use the EF-S 17-55 mm f/2.8 IS USM or use the EF 24-105 mm f/4L IS USM.

Don't ask my why it suddenly works?

[Andy0x1]

OK, so about this ISO 100 stuff.

I tested it with my 10-22 lens at f4.5/22mm

Exposure looked OK.

I set the flash (580EX II) to manual "full dump" 1/1

Exposure looked the same.

So to conclude, ISO 100 doesn't fix anything, it only seems to as the flash isn't powerful enough to over-expose in a bounce scenario.

I tried to repeat my testing with the 50mm 1.4, however with that lens attached, every bounce shot turned out decent no matter what settings I used.

Why the 50 is unaffected is beyond me....

I loosened the 50 so it wouldn't report to the camera, and guess what

Now the onboard flash began to contribute to the picture (but the overall exposure was still decent)

tightened it back up, and the onboard flash stopped contributing.

(it was configured to only fire the slave)

So any of you testing your setup that are unable to re-produce the problem, try a different lens and make sure your using a faster ISO 400+ in a small room, 1600+ in a large room.

OK AND GET THIS

The 50mm 1.4 set at f/5.6 ISO 1600 looked decent
The 28-135mm set at f/5.6 50mm ISO 1600 over-exposed.

With the lens loose as to not report to the body, it still over-exposed.

Yet on the prime it didn't (however the onboard contributed when it wasn't supposed to)

I would provide pictures, but I don't know where to host them.

-A

[ccp900]

well, i think the thing here is that we are really seeing different symptoms but same result. apers has it real bad because his copy seems to always want to overexpose.

i can live with mine since i seen to overexpose when there are dark items in the frame....i took some real world shots of my son....i was at iso 800 and accidentally bumped the camera to P. there was some overexposure due to flash but not nuclear, and it was because the flash was approximately 3 feet away from him or 4 feet pointed directly. so in this scenario no system will be able to expose properly since at iso 800 im betting the minimum flash dump will have to be at a distance farther than 3 feet...the flash just cant go any lower. please correct me if im wrong.

ALSO....my copy reacts to FEC set in the camera menu. i dialed down 1 stop less and the flash responsed. so again a different case from Apers.

Heres the good news....Apers said that the new firmware update fixed his camera....SO HOORAY for canon.

they are backing up this camera so much that you just gotta love the giant company flexing a bit of muscle.

[Rio Sundoro]

To Anders, I read your post here [http://photography-on-the.net/forum/....php?t=777238]. Does the new firmware 1.1.0 really fixed this issue?

[Andy0x1]

Here is the 50 1.4 at: 50mm / f/5.6 / ISO 1600
http://i927.photobucket.com/albums/a...1/IMG_0597.jpg

And here is the 28-135 at 50mm / f/5.6 / ISO 1600
http://i927.photobucket.com/albums/a...1/IMG_0598.jpg

All else the same.

[Andy0x1]

Update: Firmware 1.1.0 fixed it. At least it seems to have. granted I only took one picture, but I know for a fact that one would have blown out before.

Sweet deal.

Thanks everyone.

-A

[saberg]

I can confirm that 1.1.0 fixed the problem. With the previous version my 7D with 430 ex as slave was pretty much unusable. 100% of all photos taken from more than 5-6 ft. would be seriously overexposed. I've shot 20-40 photos since the update and not a single photo has come out overexposed. Thanks Canon for the quick fix!!!