Canon Confirms Light Leak Issue in the 5D Mark III

[Digital Guy]

Wow, how sensitive are the sensors becoming?

http://www.petapixel.com/2012/04/13/...e-5d-mark-iii/

I don't think they'll be able to fix this with a firmware update.

[bomtek]

Having had a good read about the subject it doesnt seem that it will be an issue to 99% of the users.

Those that do take photo's in the middle of the night (spies, terrorists, CIA etc) just need to be aware of the issue and how to circumvent it.

The one issue that is concerning is to the guys who take long exposure landscapes using ND filters whilst the sun is still out.

Thanks for the post!

[Gavin van Dam]

The could just add LOMO and HOLGA to the pictures styles in the firmware

[Henk Coetzee]

Having had a good read about the subject it doesnt seem that it will be an issue to 99% of the users.

Those that do take photo's in the middle of the night (spies, terrorists, CIA etc) just need to be aware of the issue and how to circumvent it.

I disagree. Let's just say I was a Canon shooter and I had just put down my R31k on a new 5D Mk III. You're telling me that there's a problem which I need to be aware of and know how to circumvent. Somehow it doesn't work for me.

The one issue that is concerning is to the guys who take long exposure landscapes using ND filters whilst the sun is still out.

Which is one of the 5D series niche markets. The other issue is going to in the studio with lots of lights at lots of different angles - the 5D's other niche market.

The one silver lining here is that recovering from a problem like this can do a company a lot of good, as it gives them a chance to interact with their clients. It will be interesting to see if the issue is a design fault which could be quite major for Canon or a manufacturing issue which affects a batch of cameras and can be quite easily repaired/replaced. It won't be the first time there's been a batch of cameras with a problem, or the last.

It reminds me of what happened to some of the early production Pentax K5s. Nikon may have had the same problem with the same sensors, but their reaction wasn't as clear, leading to lots of grumbling on the interwebs.

This sort of problem is really just symptomatic of a perceived need for new and better products to be brought to market too quickly for proper testing.

[bomtek]

I disagree. Let's just say I was a Canon shooter and I had just put down my R31k on a new 5D Mk III. You're telling me that there's a problem which I need to be aware of and know how to circumvent. Somehow it doesn't work for me.



Which is one of the 5D series niche markets. The other issue is going to in the studio with lots of lights at lots of different angles - the 5D's other niche market.

The one silver lining here is that recovering from a problem like this can do a company a lot of good, as it gives them a chance to interact with their clients. It will be interesting to see if the issue is a design fault which could be quite major for Canon or a manufacturing issue which affects a batch of cameras and can be quite easily repaired/replaced. It won't be the first time there's been a batch of cameras with a problem, or the last.

It reminds me of what happened to some of the early production Pentax K5s. Nikon may have had the same problem with the same sensors, but their reaction wasn't as clear, leading to lots of grumbling on the interwebs.

This sort of problem is really just symptomatic of a perceived need for new and better products to be brought to market too quickly for proper testing.

Not important Henk, I didn't expect you to agree with me!
Secondly, if you've ever done any studio work, you will know that most togs use light meters so the issue is irrelevant.
Thirdly, the only time its an issue when taking landscapes is with direct sunlight. The golden hour doesnt generally have this problem.
Lastly, if Canon isn't able to fix it and you're not prepared to work around it, don't fork out R31K.




and of course all of this is tongue in cheek, so don't get your nickers in a knot

[Dodging]

Those that do take photo's in the middle of the night (spies, terrorists, CIA etc) just need to be aware of the issue and how to circumvent it.

They've all moved over to 'The Dark Side', so they don't need to worry

[visagieh]

Lastly, if Canon isn't able to fix it and you're not prepared to work around it, don't fork out R31K.

Sorry, if you didnt know about the problem, a full refund would be in order. Come to the dark side. This is a real @#$% design issue. Wonder why they havent done a product recall.

Just watch the movie clips.

Hannes

[Chris F]

how serious is this problem ?

are potential buyers actually canceling their orders now ??




I am interested in star trail photography - is this problem actually going to impact on the quality of such photos ?

[Henk Coetzee]

Lastly, if Canon isn't able to fix it and you're not prepared to work around it, don't fork out R31K.

My issue would be if I'd already forked out my R31k. It's not the issue of fitness for purpose, it presents a huge opportunity for Canon to either make a really good name for themselves - a recall of some sort and repair/replacement or to trash their name as a company which doesn't stand by its products.

Don't know what Canon SA has to say but Canon USA is making all the right noises.

and of course all of this is tongue in cheek, so don't get your nickers in a knot

Like I said. It's not relevant to me at all.

Sorry, if you didnt know about the problem, a full refund would be in order. Come to the dark side. This is a real @#$% design issue. Wonder why they havent done a product recall.

It's not a Nikon is better issue. All manufacturers will, at some point, have a problem with a product. It's how they respond to this.

[Chris F]

Henk and this is why I am asking. I am gearing up for the Mk III. This does "dampen" the mood though .... but I expect that I wont ever even notice this. Still, knowing about the issue one would want some peace of mind ....

[Forgiss - Sean Nel]

Played with it, tried it, tried to get to influence a shot, ranging from ISO100 to ISO 12,800 and from 1/4000th of a second to 30 seconds... The only thing I didn't do is switch my maglight on and pressing it against the LCD while doing High ISO long exposures...

I am sorry, but I just do not see the problem... I can get it to change the exposure in the metering, but once you have a lens on there, the situation drastically changes... I am sure there are people out there that have gotten the effect on images, but I have not.

I am not saying it doesn't exist, I am just saying I haven't gotten it to influence my shooting in any way. It may have made a change, but I didn't see it or realize it...

[Chris F]

Thank you Sean ..... guess what I am buying myself for my birthday in September ...

[BrianSA]

Yes, this is the real point isn't it ? All reports of this oddity show it to be exhibited with a body cap fitted. I have not seen any report anywhere of the quirk showing up with a lens fitted. Not sure why anyone would worry about it, or even want to have it fixed really. I certainly find it difficult to take pictures with the body cap on

[visagieh]

All reports of this oddity show it to be exhibited with a body cap fitted. I have not seen any report anywhere of the quirk showing up with a lens fitted.

Brian, if you go to the OP, it looks like the second clip has a lens attached. Its not just direct sunlight, but also the build in light of the LCD.

How bad it is I cant tell, Sean has first hand experience.

My point is when one pay so much, I would expect a constant reading from the camera.

[BrianSA]

Ah, OK. I just went back to read the bit where I saw this info, and same results with lens fitted and lens cap on. That one showed no problem with the lens cap off. Still, Canon have said there's a problem and there'll be an announcemnt soon, so we'll find out their answer soon enough.

[Carel1966]

I’ve tried to simulate this problem, both with the LED indoors, and direct sunlight outdoors, and my reading does not jump at all. I did not try with a lens cap fitted, but don’t see the point of doing that.

[Forgiss - Sean Nel]

I’ve tried to simulate this problem, both with the LED indoors, and direct sunlight outdoors, and my reading does not jump at all. I did not try with a lens cap fitted, but don’t see the point of doing that.

Hahahaha... I have never tried to shoot with my bodycap on, or my lenscap on (well, not intentionally) I can see a problem happening with somebody shooting star trails, or something like that, but then, the setting would be way past what the LCD can display. I just think this is a non issue.

As for Canon "Confirming" the problem, that is also not quite accurate, they have confirmed the "phenomenon" exists but not that it's a problem (fact is a few other cameras do this, and not only canon cameras, the 1DmkIV does this, and the 5DmkII does this)

From the Canadian LensRentals Site:

...The final test I did with the 5D Mark III was take the lens cap off and meter off a relatively neutral wall. I turned on the backlight and nothing changed. I then pressed a 130 lumen flashlight up against the top LCD and the exposure didn’t change. I also ran the flashlight around the camera body and absolutely nothing changed...

[Fanie]

Sounds like a storm in a tea cup.

[Grahame Lane]

If Sean is happy then I am happy

[Digital Guy]

I tried this on my 60D and yep its also affected but more noticable is when you shine a torch inside the viewfinder. In the old film days when you were'nt behind your camera you had to close the viewfinder door to stop light entering the camera.

I'm not so sure that this is a non-event. Please explain to me, how, when sunlight hits your top LCD screen or enters thru the viewfinder and the exposure jumps 3-4 stops, how is that not going to affect my final exposure. It has to and will affect it.

Apparently this affects the Nikon cameras as well according to some forums. Can Nikon users confirm or deny this.