ygelman Loc: new -- North of Poughkeepsie!

Please help me figure this out. I was using a Canon 5DII in low light situation, but hand held, and in aperture priority.

I felt that auto focusing was not accurate (maybe because of the low light), so I went to live view with manual focus and pushed the zoom view button to get a sharp manual focus before snapping the shutter.

It turned out that the images with live view were all exposed brighter than when using the viewfinder which I'll call "normal".

The exposure times were longer when using live view.

When I got home, I took two images in the same way so that there was no possible change in conditions. I'm including those images in jpg here. I did not do any post processing; the only thing I did was to save the images to about level 5 to make the file sizes smaller.

I can send the raw by private message if that makes any difference.

Thanks in advance for your comments.

Normal exposure


Live view exposure

Camerabuff Loc: Liverpool UK

Try camera raw you can recover full exposure

RetiredPhotog Loc: South West Ohio

ygelman that is a strange situation. I see the white balance is off some but I wouldn't think that would cause any difference between live view or view finder. The only difference between the two views is using the mirror. One other possibility is light entering your view finder if it wasn't covered for the shot. That would have an effect on it more than anything. :idea:

Searcher Loc: Kent, England

My Nikons both came with an opaque "cap" to cover the viewfinder and prevent extraneous light from entering the pentaprism when camera is on tripod and the camera controlled remotely. There is a light sensitive something in the pentaprism which will affect exposure, this is usually covered by the shooters eye on the viewfinder. Try your experiment again but this time cover the eyepiece with something when using Live View.

I have just read that SOME cameras have the photodiode which measures incoming light mounted in the pentaprism above the eyepiece. Other camera appear to have the diode behind the mirror. My camera obviously have the diode in the pentaprism, perhaps yours is the same.

ygelman Loc: new -- North of Poughkeepsie!

Both RetiredPhotog and Searcher mentioned light coming through the viewfinder when using live view. I'll try covering it when I repeat the experiment tonight. But anyway, it doesn't explain why the exposure times were different, and why it differed the way it did. If light entered the viewfinder, wouldn't the exposure time be shorter?

Also, I just noticed that the focusing method changed: Centerweighted-Average in the normal case, and Evaluative in live view. Even though I didn't change it. Could that be a reason???

Searcher Loc: Kent, England

I don't have Live View but I did read somewhere that focussing methods can be restricted. I think that's a job for the manual unless someone who knows your camera can jump in.

RetiredPhotog Loc: South West Ohio

Definitely the focusing method can change the exposure settings. Spot focus will basically look at the light at the spot and averaging will look at the whole picture.

SharpShooter Loc: NorCal

Ygelman, if the metering mode changes when in live view, yes that will have an effect on light levels of the finished shot.
As to covering the viewfinder, I have read several fairly conclusive tests that find that light can not or will not leak into the viewfinder. I have never covered mine and probably never will.
In the test, I think they had the lens caps on and shined strong lights into the view finder for prolonged exposure times, and all was pure black.
Just some food for thought.
SS

big ed Loc: Dudley UK

Nikon provided a plastic cover for the viewfinder of my d3100 to use when shooting in liveview. I dont think they would give you this if it didnt make a difference......
Kev

Searcher Loc: Kent, England

Depends on how your camera is constructed - if the diode is below the focussing screen then no problems but if it is built into the pentaprism as quite a few are, then light can spill in.

ygelman Loc: new -- North of Poughkeepsie!

Update: 2nd round of experiments

In the course of taking many pairs of photographs, I did the following:
-changed the metering mode to Evaluative so that both shots used the same
-changed the lens opening to f/11 in case the original f/4 had caused problems
-changed exposure compensation to 0 rather than -1/3
-changed auto flash exposure to 0 (I never use flash anyway)
-changed to shutter priority and set it to 1/4 sec
-changed to Manual mode

For each pair, the result was always the same. Live view was always about 1 stop brighter than viewfinder. In fact, when using shutter priority, the shutter speed actually changed to slower by 1 stop.

When in Manual mode, I changed the shutter speed to bring the indicator in the "viewfinder" meter to zero for each shot. Still, live view was overexposed.

Finally, in Manual mode, after taking the 1st shot of the pair using the viewfinder, I let the "viewfinder" meter stay where I had set it before and changed to live view. The meter indicated very underexposed. And when I compared the two shots, they were finally the same!

So the results conclusively point to the light meter when using live view, since I (probably) eliminated every other cause.

Would others please try experimenting and report back?

Again, thanks in advance.

MagicMark

I really have no clue as to what might be causing that but here is my guess. It may be that either the image sensor is getting "hot" while in Live View or in some cameras the sensor gets a little more light through the view finder during Live View mode but of course if this last idea is true then you might have to cover the view finder while making the exposure. The only way to know for sure is to test my theory.

Again, I'm no expert on this subject and I could be totally wrong. Hope you figure it out and fix it. Good luck! :)

BermBuster Loc: Hi Desert S.Cal

I just tried this on my 5DMKII, and found the live view does come out brighter...unless I cover the viewfinder.
....low light, ƒ/4, iso 400, evaluative , aperture priority... my shutter speed was at 1 second - but on live view it went to 1.3sec, I covered the viewfinder and took another shot...shutter speed dropped to 1 second. on further tests, if I covered the viewfinder first, the exposure on live view matched.
You would think if light is entering thru the viewfinder, the exposure would be 'darker'.

ygelman Loc: new -- North of Poughkeepsie!

Hmmm. Three revealing comments. Two say that covering the viewfinder does matter. Could Sharpshooter provide some urls for the tests showing it doesn't matter? When I tried it myself, I got the same unequal exposure. . . I'll try it again more carefully tonight. Maybe individual cameras are different??? And, in BermBuster (thanks!), why would the procedural order of covering the viewfinder matter? -- if indeed it does matter. Still a puzzlement.

pinupphoto Loc: Pennsylvania

I'm not sure is Cannon has it, but my Nikon D-800 has it, a tiny toggle switch (which is veryEASY to miss!) that shutters off the eye-view hole. The Nikon literature explains EXPRESSLY, that it is to be used when shooting videos, or whenever using the viewing monitor.