Canon 5D Mark III Light Leak?
- Thread starter Canon Rumors
- Start date
- Status
Very unlikely. Plus, what is "over/under-exposed" is very subjective.Etienne said:
Personally, I prefer to slightly "under"-expose relative to what my 7D and my 5D3 "think" is the correct exposure, because I think it results in better colour saturation. I find colours tend to "wash out" slightly at the exposure levels that the camera indicates when exposure compensation is set to 0.
IMNSHO, there's NO WAY that in real-life shooting scenarios, this effect could cause an exposure deviation as large as 2/3 of a stop.
Upvote
0
Etienne said:
That's not entirely accurate. Saying it "underexposed" would imply that the Mk3 made a mistake, but that's not what he said. What he said was "To make it even harder to do direct comparisons the 5D mkIII seemed to be around a stop to 2/3 of a stop darker than the D800 at any given ISO setting." It's not really surprising that two cameras with different metering systems interpret the same scene differently - but he never inferred that one camera had it right while the other was wrong. Just that in comparison, one tended to darker than the other.
Also, one of the comments below the article, a user reports that in HIS comparisons, the Nikon tended to be darker than the Canon.
Upvote
0
Read this elsewhere.
Read it and think about it.
This leak thing with cap on is BS.
These lens cap on tests are ludicrous beyond belief.
They involve light levels far beyond the meter's range of operation. The meter is designed to produce an image that is 18% grey. Does anyone seriously think that the meter reading in these tests, with or without the LCD light, are registering 18% grey.
Add to that the fact that the EV scale is non-linear and it's hard to see how a rational, informed person could get so hysterical about this. The quantity of light required to shift exposure by 1 stop at these levels becomes a tiny fraction of a stop at real exposure levels - probably less light than leaks into the viewfinder past your eye - and much much less if you happen to look through the viewfinder wearing glasses. 8)
Read it and think about it.
This leak thing with cap on is BS.
These lens cap on tests are ludicrous beyond belief.
They involve light levels far beyond the meter's range of operation. The meter is designed to produce an image that is 18% grey. Does anyone seriously think that the meter reading in these tests, with or without the LCD light, are registering 18% grey.
Add to that the fact that the EV scale is non-linear and it's hard to see how a rational, informed person could get so hysterical about this. The quantity of light required to shift exposure by 1 stop at these levels becomes a tiny fraction of a stop at real exposure levels - probably less light than leaks into the viewfinder past your eye - and much much less if you happen to look through the viewfinder wearing glasses. 8)
Upvote
0
altenae said:
Thanks for asking. I take most of them with the cap on, I like the look and avoids loosing time with post processing...but seriously, I've taken aurora pics where it was pitch black outside and used to check the time by lighting the LCD (while the pic was being taken). Likelly under that scenario there could have been some shift had I used the Mark3....just speculating.
Upvote
0
C
chriswatters
Guest
When taking pictures under those conditions, do you use the camera' light meter to set the exposure, or do you set it manually? If you set the exposure manually, how would this issue affect you?fegari said:
It sounds like this effect is only observed when there is so little light that the sensor is blindly guessing at the exposure. What is the "correct" exposure that the camera should be reporting when the camera has the lens cap on and viewfinder closed?
Upvote
0
I'm pretty sure many will blame Canon now for taking underexposed photos.
M
magus424
Guest
fegari said:
And how, exactly, would this affect your long exposure shots? I just explained that this has no effect when a lens is on instead of the body cap, so unless you're shooting with a cap on, it won't change a thing
Etienne said:
No. See above.
fegari said:
No, there would've been no shift.
Upvote
0
D
Devinder
Guest
so to help you blokes and my own sudden heartattack calm down. I ran two tests.
With lens cap off, (i believe this is how those things are meant to be used)
regular bright light from sun light room + and minus lcs (and a pen light to boot) --> no change
all blinds down closed no lights in room vs lcd and then pen light! --> no change.
ps: i tried with lens cap on and almost gave my self a heartattached because well, in that setting your so called light leak does indeed happen.
So in conclusion, i don't see this as an issue. Just like a power (diamondtipped)chainsaw test failure on integrity isn't an issue.
With lens cap off, (i believe this is how those things are meant to be used)
regular bright light from sun light room + and minus lcs (and a pen light to boot) --> no change
all blinds down closed no lights in room vs lcd and then pen light! --> no change.
ps: i tried with lens cap on and almost gave my self a heartattached because well, in that setting your so called light leak does indeed happen.
So in conclusion, i don't see this as an issue. Just like a power (diamondtipped)chainsaw test failure on integrity isn't an issue.
Upvote
0
I don't think the sensor is blindly guessing - I think it has just stopped searching for a proper exposure setting. Heres my thoughts why:chriswatters said:
I did this crazy test inside a neoprene bag with a lens attached and the lens cap on in a very dark room and got 10 seconds using Aperture Priority. I was truly expecting to see an under-exposed warning (>30 seconds) in such an extremely dark environment, not a repeatable 10 second exposure setting like others are seeing (see videos and CR Guys notes). Why didn't I get an under-exposure warning in such a low light condition? Why isn't the exposure time flashing at 30 seconds, but rather stopping repeatably at 10 seconds? So I started to think about the metering algorithm. My guess is the algorithm increases exposure time (2x) and look for some corresponding change in meter reading (2x) (as you would expect in normal use (you know, some light, lens cap off) until it found the optimum exposure), and if it doesn’t get a reading increase change after a few increments it stops its search. 10 seconds at ISO800 seems to be the algorithm limit in the absence of light to the meter. At ISO6400, the time was 1 second. So it seems likely.
Upvote
0
This thread is a very good example of how forums can over analyse things. If you are doing shots where it is completely dark such as night sky and you are trusting any camera'a meter, then in my opinion you need to learn to shoot at an advanced level. I do some star trails photography sometimes in the countryside of Northern England, and I shoot only in bulb setting. You can use the back screen to show what aperture etc things are set at then press the display button again to turn it off before you press the shutter using timer or cable release.
The nitpicking annoys me. You can find a way that any electrical item can produce a fault. Shoot something other than lens caps, and learn to make your own exposure decisions in manual mode. Shoot compare the histogram if shooting in daylight and reshoot if necessary. Simples.
The nitpicking annoys me. You can find a way that any electrical item can produce a fault. Shoot something other than lens caps, and learn to make your own exposure decisions in manual mode. Shoot compare the histogram if shooting in daylight and reshoot if necessary. Simples.
Upvote
0
- Status
Similar threads
