contrast detect increases exposure over phase detect..?

[paulph]

Hi - I'm using a D3100 and have noticed that a shot through the viewfinder (phase-detect) is always 1-2 stops faster (aperture or shutter-speed) than the self-same shot using Live-view (screen, contrast detect)! I've experimented with priorities, settings, compensation, focus modes and so on, and it remains the case. I think it may be the the different ways of gathering info by the two systems. I've not found a 'scholarly' article, though.
Any ideas? Thanks!

[Dave Humphries]

Hmmm - an interesting question Paul,

The Nikon cameras, being 'proper' DSLRs, will be forced to use an alternate metering method while the mirror is 'tucked up' for Live View' because the normal metering sensor is switched out of the optical path, as is the viewfinder, while LiveView is in use.

If you can't find a 'cause' in a different menu setting for metering (e.g. average vs spot*) when in LiveView (LV), and it bothers you, it might almost be worth querying it witha a Nikon Service Centre.

* I suspect the metering area in LV is probably fixed, if you can discover what this is (perhaps it is in the full manual), then set the normal metering to the same mode, and conduct some tests on an evenly exposed blank wall or similar and see if there is still a difference. If there is, it is probably just a calibration issue that a service centre can fix easily.

I might try comparing on my D5000 next time it is out of the bag.

Hope that helps - and welcome to the CiC forums from ....

[paulph]

Thanks, Dave. I'm certainly curious about it. Might have a nip downtown to ask at the shop. I'd be interested to know if your D5000 does the same, or if there's a general explanation.

[dje]

Hi Paul

I can't give you a "scholarly" reply but here are a couple of comments.

The terms Phase detect and Contrast detect refer to the Auto Focus mechanisms and sensors rather than metering. The fact still remains however that metering in Viewfinder mode uses a separate metering sensor so there is always the possibility of calibration errors between the two metering situations - Live View or Viewfinder.

One or two stops seems very severe - do you mean full stops or partial stops. I've found for example on my Canon 600D that the two modes might differ by one shutter speed setting in aperture priority mode. eg from 1/50 to 1/40 sec.

If it's as bad as you say, a visit to your friendly Nikon dealer could be in order.

Personally I rarely use Live View mode - only for manual focussing in low light conditions.

Dave

[paulph]

Thanks, dje - you're right on too, it is one shutter speed difference, sometimes two. I'm continuing to experiment. I also noticed that even when set to AF-A (viewfinder info) the system always sets itself to AF-S in Live View. Maybe that's both normal and unrelated to the exposure behaviour!

[enaiman]

What happens if you cover the viewfinder and try live view again?
I know that sometimes, stray light coming through viewfinder might cause different exposure settings.
I have never experienced that but this is what others are saying ...

[paulph]

Thanks, enaiman - I hadn't thought of that, and it could well be a factor! I'll try it.

[Ady]

stray light coming through viewfinder might cause different exposure settings.

Canon certainly recommend you use the viewfinder cover when using live view for this very reason, I can't confirm if this is true without testing as I never use live view.

Cheers,
A

[Manfred M]

Canon certainly recommend you use the viewfinder cover when using live view for this very reason, I can't confirm if this is true without testing as I never use live view.

Cheers,
A

This is also true for Nikon; there is an cover that clips onto the camera's hot shoe for lower end cameras and a built-in cover for higher end ones. I don't do a lot of live-view shooting, but I have certainly noticed that my exposures are much darker if I forget close off the viewfinder.

[paulph]

Thanks, fellas. I'm in email discussions with Nikon, they seem a bit mystified too!

[Dave Humphries]

If you described it to Nikon as "2 stops" (which is like; 1/50 to 1/200), when it's actually only one or two clicks (probably only 2/3 of a real stop), that possibly did get their interest.

Bias from light coming through the V/F is a good call (whoever suggested it first), although to my mind, it might be more sensitive to that when in normal mode auto-exposing if your eye isn't over the V/F because it is on a tripod. I might have expected the V/F metering sensor to be turned off when the mirror is up (and the main sensor is providing that information) - at least, if properly designed.

Good luck,

[paulph]

From Nikon tech-support:

"The reason for the exposure difference is because the angle of view between the viewfinder and the LCD screen. Since the angle of view is changed between the LCD screen and the view finder, the shutter speed changes while in aperture priority."

[dje]

From Nikon tech-support:

"The reason for the exposure difference is because the angle of view between the viewfinder and the LCD screen. Since the angle of view is changed between the LCD screen and the view finder, the shutter speed changes while in aperture priority."

That doesn't seem to be a very informative answer to me !! Perhaps they are referring to the fact that the Viewfinder exposure sensor is up near the viewfinder and the light reaching it travels a different path to that reaching the main sensor. The two should be calibrated to read the same but I guess this may not be entirely accurate.

Dave

[Manfred M]

From Nikon tech-support:

"The reason for the exposure difference is because the angle of view between the viewfinder and the LCD screen. Since the angle of view is changed between the LCD screen and the view finder, the shutter speed changes while in aperture priority."

I agree with Dave's thoughts on this. While the positioning of the sensors is different in location and how the light is read, the difference in most cases should be virtually imperceptible, assuming of course that the light meters have been calibrated properly. SLR photographers (and camera makers) have been using slightly different metering strategies for years between camera makes and models, we should be looking at fractions of a stop here, not the 1 -2 stop difference in the initial question. My own experience is the same as Eugen's; the only time I see a significant difference in my exposures is when I forget to cover my viewfinder when using live view.

[paulph]

Thanks for the comments, folks. I agree, the camera should meter the same, regardless of the viewing mode and light path. I found the Nikon tech-support a bit dismissive. I plan to experiment with friends' cameras to see if this is a common occurrence.