The Light Leak Issue

[mrmarks]

If anything is used outside of its intended function, it will probably result in some unintended or unusual outcome. For example, if the reverse gear is used to try stop a car, this will result in a less than desirable driving experience. In one of the tests done by a user, a long exposure was taken at night on a tripod and when the LCD light was turned on, the exposure time decreased. Perhaps, the LCD display needed to be lighted up to check the settings, but after it turns off automatically in a few seconds, then an exposure can be taken. There is no need for the LCD to be lighted up when an exposure is taken. For long exposures, it is common practise to cover the camera body with a light-proof black cloth to prevent stray light from entering the VF and causing flaring effects in the exposure. I doubt Canon will do anything about this "issue" and at the most Canon will just send out an application note?

[neuroanatomist]

Perhaps, the LCD display needed to be lighted up to check the settings, but after it turns off automatically in a few seconds, then an exposure can be taken. There is no need for the LCD to be lighted up when an exposure is taken.

But what if you wanted to repeatedly verify your settings during the exposure by viewing the LCD during the exposure?!?  Don't go saying that's not realistic because you can't change the settings at that point or that pressing a button on the camera would result in a blurry image.  It's more realisitc than taking pictures with no lens...   

[mrmarks]

Perhaps, the LCD display needed to be lighted up to check the settings, but after it turns off automatically in a few seconds, then an exposure can be taken. There is no need for the LCD to be lighted up when an exposure is taken.

But what if you wanted to repeatedly verify your settings during the exposure by viewing the LCD during the exposure?!?  Don't go saying that's not realistic because you can't change the settings at that point or that pressing a button on the camera would result in a blurry image.  It's more realisitc than taking pictures with no lens...   

LOL! Actually I don't even use the top LCD as I check all my settings in the rear LCD in quick mode all the time. I can't be bothered even if the top LCD is removed.

[smirkypants]

LOL! Actually I don't even use the top LCD as I check all my settings in the rear LCD in quick mode all the time. I can't be bothered even if the top LCD is removed.

+1 ...  I seriously wouldn't notice if the LCD were gone.

[mrmarks]

Canon EOS 60D
I tested 2 bodies and found no issue once I placed the folded black fabric over the viewfinder. No change with backlight on.

Canon EOS 7D
I tested 2 bodies and found no issue once I placed the folded black fabric over the viewfinder.  No change with backlight on.

Canon EOS 5D Mark II
I tested the one body I have in stock. This camera exhibited the same issue as the 5D Mark III in direct sunlight, however the backlight didn’t affect exposure. I also did the flashlight test that is mentioned below and nothing changed in exposure.

Canon EOS-1D Mark IV – Body #1
There was an ever so slight shift in exposure with the viewfinder cover down.  10 seconds out of the sun, 8 seconds in the sun. I stuffed the viewfinder with the black fabric and there was no change. The variance in exposure went away if I wrapped the LCD with the black fabric and only left enough space to see the exposure number. No change with backlight on.

Canon EOS-1D Mark IV – Body #2
This body exhibited no movement in exposure. No change with backlight on.

-----------------------------------------------------

Could someone check which side is the LED (inside the LCD) on all the tested bodies? In the 5D3, the LED is on the left side, closer to the metering sensor. Could it be that all the other bodies have the LED on the right side? Just a hyphothesis. You could check the LED position by looking sideways into the LCD.

[justsomedude]

I just did a similar test, taking two outdoor pictures with my new 5D MK iii. The one without the lcd light on is metered for 10" exposure duration and turns out fine. When I use the lcd backlight the meter shows 6" exposure duration and the shot gets underexposed by almost one 1 stop.
I acknowledge these tripod mounted night shots with long exposure duration are not the ones I do regularly, but to me the light leak problem seems very real.
So, should I send the camera back?

Is it normal practice (for you or anyone else for that matter) to have the LCD backlight on when metering? If your answer is no, then just use it as you normally would and don't waste your time and potentially money returning it. If yes, then may be you need to review your practices and/or make manual adjustments.

As I noted previously - some astrophotographers use the LCD backlight during metering to observe the exposure settings.  Well - at least I do.  This way I can use the settings to calculate a longer exposure for a manual/bulb exposure.  And there's really no way to see that LCD display without activating the backlight (or by using a headlamp - which we're learning can also effect the metering).

So if we can't use the backlight under pitch black conditions out in the field - then one could certainly resort to the rear LCD image preview/info screen.  However, this is a huge hindrance in my opinion.   If the camera is on a tripod in a funky position, where access to the rear LCD is totally obstructed, you would literally be screwed. 

I don't think everyone is realizing how much of a serious problem this is turning out to be (at least for nighttime landscape photographers). 

[justsomedude]

From the video previously posted by "highnfar," here are screen caps of images taken with the backlight on during metering, and backlight off during metering....



 

[cinema-dslr]

Canon EOS 60D
I tested 2 bodies and found no issue once I placed the folded black fabric over the viewfinder. No change with backlight on.

Canon EOS 7D
I tested 2 bodies and found no issue once I placed the folded black fabric over the viewfinder.  No change with backlight on.

Canon EOS 5D Mark II
I tested the one body I have in stock. This camera exhibited the same issue as the 5D Mark III in direct sunlight, however the backlight didn’t affect exposure. I also did the flashlight test that is mentioned below and nothing changed in exposure.

Canon EOS-1D Mark IV – Body #1
There was an ever so slight shift in exposure with the viewfinder cover down.  10 seconds out of the sun, 8 seconds in the sun. I stuffed the viewfinder with the black fabric and there was no change. The variance in exposure went away if I wrapped the LCD with the black fabric and only left enough space to see the exposure number. No change with backlight on.

Canon EOS-1D Mark IV – Body #2
This body exhibited no movement in exposure. No change with backlight on.

-----------------------------------------------------

Could someone check which side is the LED (inside the LCD) on all the tested bodies? In the 5D3, the LED is on the left side, closer to the metering sensor. Could it be that all the other bodies have the LED on the right side? Just a hyphothesis. You could check the LED position by looking sideways into the LCD.

that was what i was suggesting. on my 7d the amber backlightled is on the right side.
So it seems they made a placement error.
It may be a small to nonexistent problem but still should be easy to fix for canon.

[JLJL]

Serious Light Leak Issue with the 5D mark III!

Light leaking trough top LCD causes underexposure. This is true for the for other sources, e.g. sunlight!

http://www.akphotodenver.com/images/backlight.jpg

D81 0512

http://translate.google.com/translate?sl=nl&tl=en&js=n&prev=_t&hl=nl&ie=UTF-8&layout=2&eotf=1&u=http://nl.fotovideo.nu/nieuws/27104/test-5d-mark-iii-light-leak-bug

[epsiloneri]

I don't think everyone is realizing how much of a serious problem this is turning out to be (at least for nighttime landscape photographers).

I don't think it is. Please see my earlier reply to you (#38).

[AUGS]

From the video previously posted by "highnfar," here are screen caps of images taken with the backlight on during metering, and backlight off during metering....



 

Yes, I agree I can see a visible difference between the images.  But which image is the correct exposure for the scene and what you are trying to achieve?  Without a review of the histograms and images, neither may be correct, and for night photography nor should they be assumed to be correct.  (FWIW, the image on the right looks over-exposed, but its almost impossible to tell from such a small grab - and I do know the left is the LCD affected exposure.)
I'm not saying there isn't a problem, there clearly is but it looks to be manageable.  In my experimentation, I've found that the only time the exposure is affected is when the LCD panel in is brighter light than the image to be captured and how much it is affected depends on the intensity difference.
With this knowledge in hand we are forewarned and hence forearmed and can work around it - albeit we shouldn't and I wish we didn't have to, but it is what it is at the moment. 
My workflow for night photography will not change:
1. Compose the image and set focus.
2. Set a high ISO, Aperture Priority, biggest aperture for the lens, set autobracket +/-1 stop.
3. Review the 3 histograms and images.
4. Switch to Manual mode, with same ISO, same aperture, and adjusted time based on review of previous images and the result I want to achieve.
5. Confirm adjusted exposure by creating another image or adjust again.  Determine Bulb exposure time for desired ISO and desired aperture.
6. Set bulb setting, start exposure, grab a coffee, plug in the iPod and wait.

The good news is I've confirmed (on my 5D3 at least) that the light leakage does not get to the image sensor when the mirror is up, so the activation of the LCD or the use of a head-torch or similar during the exposure doesn't affect the image. 

[bp]

People are still screaming about this?  Seriously?  Yes yes, this is a HUGE PROBLEM!!!!!!!!!!!!
IF you're doing near pitch-black landscape photography
AND you're relying on the meter to figure out your exposure for you
AND you've put the camera in a place where you can't use the rear LCD
AND you're not willing to wait to execute the exposure until after the LCD backlight turns off

I don't do much nighttime landscape / astrophotography - but when I have, I've never once relied on the camera's metering system.  That's what M mode and a few test shots are for.   Why all the drama?

[Martin]

Please could you check metering in normal use? I had very annoying problem with 5D2 which underexposed by 2/3 stop all the time. I did a white(gray) wall test with all different mattering modes and histogram against white or gray card was off to the left by 2/3 stop. it's not normal behaviour as it should be set dor 18% gray, so the histogram peak should be perfectly  in center. Please check if  5d3 is also affected by this issue.

[kozakm]

FYI: 1D X is ok, no light leak problem (tested)

[javier]

I am glad you recognized that light leaks through thumbs!  Good catch.  Since my 5D2 acts the same way, im gonna say NO ISSUE here!

Now, back to wedding photography.

What about the 2/3 stop off in normal use of the camera.

Javier, If you think it is the camera, return it.
My friend and I do NOT have 2/3 stop off in normal use.

I told you....  http://www.usa.canon.com/cusa/registration/professional/professional_cameras/digital_slr_cameras/eos_5d_mark_iii

Also another guy find the camera is getting to much light through the view finder, so my conclusion is...
Canon 5D Mark III - Light Leak via Viewfinder

the sensor is getting to much light "leaks" and when you take the picture the picture just use the light coming through the lens and that is why the pictures are underexposed.


so start to pack your camera for a replacement, I already sent mine for a refund 3 days ago, and now I am waiting to see how this ends to buy a new version.