[bschriver11]

Interested in the answer as well. I recently took some photos at the county fair at night along with fireworks in AV mode since the Ferris wheel was changing the lighting every 5-10 seconds and i wanted to keep it exposed properly with the fireworks and when using the viewfinder the exposure was much darker than when i used live view and i'm sure there wasn't enough difference in the lighting to make that much difference in the exposure.i used center weighted average for the exposure and kept the Ferris wheel near the center, it almost looks like live view is using full matrix metering rather than the center weighted average selected or at least that the best way i could describe it.My camera is almost new and i see no problem with any photo in any other way.

[UncleVanya]

What lens are you using? Is the lens one with an aperture ring?

[bschriver11]

i used the DA 35 F/2.4 lens.

[StephenHampshire]

Light entering via the viewfinder when eye not applied to it will bias the exposure reading. With mirror up, the optical path from viewfinder is blocked and only light from through the lens is metered. Always cover the viewfinder when using remotely on a tripod (i.e eye not at viewfinder) Whether this effect could explain an 8 stop difference is a moot point....

[bschriver11]

Dark photo using viewfinder and bright exposure using live view, as you can see the lighting really didn't change.It was dark behind me so i'm still confused as to the exposure difference plus i always just put my hand in front of the viewfinder if it is bright from behind.Another test will maybe shed some light so to speak.

[clackers]

What were the exposures for those two shots, BShriver?

I'm a little puzzled by your strategy.

You have a lot of area around the wheel, so presumably, you wanted to show it in its environment.

You would want to expose for the carpark and/or sky, right? You expose for that, it's a DSLR, you're in charge. It's how you do fireworks.

Once you walked up to the wheel itself so you could fill the frame, you would expose for its highlights. Again, you wouldn't let a camera make an assessment over the whole image, only the lights are of interest. Take a spot reading of a cluster of them, or do a guess and snap/chimp until you're happy with the result.

[FotocamG]

Hi neptune!!

Thanks for following the thread! Right now I am under the same issue. Did you fix the problem?
While I wait your reply... I am going to try to dissasemble my K-50 to fix the aperture motor... if that doesn't do the trick... I'll be back!!

[UncleVanya]

Dark photo using viewfinder and bright exposure using live view, as you can see the lighting really didn't change.It was dark behind me so i'm still confused as to the exposure difference plus i always just put my hand in front of the viewfinder if it is bright from behind.Another test will maybe shed some light so to speak.

So first of all - I have taken shots at that same location I think! (Or maybe not? Mine were at the "Island" in PIgeon Forge TN) That wheel is lighted with different patterns that have vastly different metering implications. I elected to turn OFF all automatic exposure during my shots of that wheel because the meter is fairly easily tricked with that thing. I happened to have my Panasonic GX-7 and the 12-35 f/2.8 lens with me at the time rather than my K-3. The GX-7 is a micro 4/3 camera and meters in what is essentially live view all the time. Even so I had wide variations of exposure on different patterns on that wheel until I switched to manual. I don't know if I kept some of the others but here are two that I did keep:





Sorry these lack EXIF data - I was remote without a laptop and used my phone to transfer the pic via wifi and then did my edits in Pixlr and uploaded from there to Flickr. None of this used RAW just my in camera JPG. I do have the files still I think and could dig into them a bit to see if I had any blown exposure ones but honestly I think unless you set the camera in spot metering mode you will have trouble with autoexposure on that subject.

Is this happening in more traditionally lighted scenes? Say daytime in the park?

[mikesbike]

Light entering via the viewfinder when eye not applied to it will bias the exposure reading. With mirror up, the optical path from viewfinder is blocked and only light from through the lens is metered. Always cover the viewfinder when using remotely on a tripod (i.e eye not at viewfinder) Whether this effect could explain an 8 stop difference is a moot point....

This is very true, and is an unknown factor in assessing these different experiences between shots taken with the eye covering the VF and nothing covering it. If there was virtually no lighting coming in from behind the camera to enter the uncovered VF, there would be no influence. If there is a chance of influence, it is necessary to either cover the VF when either using live view or on a tripod without an eye to the VF, or use the VF first to set exposure, then either hit the AE-L to freeze the exposure, for a quick shot using live view, or switch to Manual mode and using the VF first to set exposure by hitting the green button (Hyper Manual) then go forward with live view shooting.

For that night scene, what does that ferris wheel look like? Does it look like it has spokes, or does it look like it has red, white, and blue jaggies in a circle filling its entire profile? If not, it is because the camera "decided" to expose for the more shaded, less lit areas of the scene, therefore using a slower shutter speed, which would make the wheel look like it is spinning a a fast speed, hence the artificially whirling jaggies. The buildings then look normal, but not the wheel?? Whoever was actually there and took the shot should know.

There is no way to properly do a night exposure for moderate- to dimly lit buildings, and other objects and people in a scene to look normal, while at the same time expose for a very bright moving object to look normal. When taking night street scenes, for example, with store fronts, buildings, and people standing at a bus stop, if you get all that pretty accurately portrayed, the passing cars will have streaking headlights and be blurred.

Shoot in Manual mode, use spot metering to choose which part of the scene you want to bias your exposure for, then as clackers says, modify it according to viewing your results until you get close to what you are after. Use the Pentax Hyper System, for Hyper Manual operation- just hit the green button when training the spot meter on an area to set an exposure reading. Move the camera to recompose and frame your shot. Then modify from there.