pictures come out dark on bright sunny day?

[glennb]

Had an issue the other day with my pictures coming out too dark. Was using Aperture setting and it was a nice bright sunny day! Im pretty sure I was on matrix metering and shooting in raw.55-300dx and 35mm, not all come out like the below pics but most did. They look like they were taken under the shade. any advice would be great, is it me or is it the camera or lens, possible causes. Cheers This pic,f4.5,1/3200sec,ISO200,auto white balance
2013 06 22_1782_edited-1 by Glennb76, on Flickr. This next pic,1/1250sec,f11,ISO100
2013 06 22_1757_edited-1 by Glennb76, on Flickr even this landscape 1/500sec,iso200,f11
2013 06 22_1793_edited-1 by Glennb76, on Flickr This is how I thought it would of come out like, this after a play in photoshop
2013 06 22_1757_edited-2 by Glennb76, on Flickr

[Glenda]

White birds are difficult. Usually best to use spot metering on the bird - I assume you used matrix or evaluative metering. Did try to check on your Flickr page but you have the image set at private. With the camera metering the whole scene the pelican will usually be dark. Like how you've processed this though.

[ameerat42]

Somehow your camera metered the highlights - see the white feathers - and underexposed everything else.
The exposures are way too low for those conditions. Anyway, a good recovery in Photoshop. No camera
mentioned, so can't be specific on its settings. (Can't get EXIF.)
Am.

[Mark L]

As Lplates said, it's the metering mode you're using. Spot metering on the bird will expose for the bird. The rest of the scene is brighter than the bird and the camera is exposing for that brightness.
Knowing this, if you used the same metering mode again, you could dial in a slower shutter speed (as you where using aperture priority). The camera would say you're over exposing, but that's what you want to expose the bird properly with such a bright background.
Last photo shows the benefit of RAW.

[glennb]

Thanks guys ill try and put more of the meta data up , wouldn't allow me to copy and paste , camera was d5100 still confused why the land/seascape was also underexposed.
Edit now open up flicker for public

[Glenda]

Maybe you were pointing the camera at the sky which is lighter and the camera metered off that. I know my Nikon always tends to underexpose. I did read somewhere that Nikons did this rather than blow out highlights. I tend to use spot metering 90% of the time but more particularly with birds and I always expose to the right - I've found this gives me the best results. You've just got to experiment and see what suits your camera best.

[ricktas]

Looking forward to getting access to the EXIF, that is going to tell us a lot. easiest way is to make your album on flickr a public one rather than a private one.

[arthurking83]

the other important question is, did you use any filters of any kind? UV, ND, CPL, etc?

[mongo]

Mongo finds it hard to believe that these results were due to exposure meter mode alone.

the third image had a lot of dark blue in it which would have tended to brighten the image in ant metering mode except spot metering and only if pointing to sky in that image. Even with filters, the camera metering should have adjusted to give a longer exposure.

the problem looks more like your compensation exposure on your camera may be set to about -1.3 to -2.0 EV. Check that your exposure compensation is set to "0".

It could also be a combination of some of the possibilities we have all mentioned in this thread. Mongo will be intrigued to learn what it is eventually.

[glennb]

Looking forward to getting access to the EXIF, that is going to tell us a lot. easiest way is to make your album on flickr a public one rather than a private one.

It should be open to the public now, the only data I can find on flicker is under additional info which just gives the main settings. Also I didn't use any filter on the 55-300 and uv on the 35mm, problem was on both lenses.

[glennb]

Ok, I have uploaded more pics to my flickr and is open to public .Also seems I used centre weighted and pattern/multi-segment most of the time with a few at spot metering. But all still inconsistent, and EV is at 0 for all.
here are a few more pics and you can get the exif data on flicker.


2013 06 22_1791_edited-1 by Glennb76, on Flickr



2013 06 22_1788_edited-1 by Glennb76, on Flickr



2013 06 22_1745_edited-1 by Glennb76, on Flickr



2013 06 22_1778_edited-1 by Glennb76, on Flickr

[arthurking83]

OK, so filters aren't the issue.

I just checked the exif data on #3 in the first set and exposure compensation is set to 0!

Why it exposed so dark is a bit of a mystery to be honest. and it seems to have been shot with the 35/1.8 which (from memory) usually exposes a bit on the brighter side.
ie. T stop on this lens is better than the kit lenses tend to be.


So far I'm at a bit of a loss.

[ameerat42]

Well, what about the processing? (???)
Am.

[rackham]

The pelican is in a couple of instances backlit, that may have contributed a bit.

[glennb]

Well, what about the processing? (???)
Am.

Sorry Am what processing are you talking about? No post processing done except change to jpeg for flicker, straight from camera to photoshop organiser to edit for jpeg to flicker. Exif data can be seen on flicker now.

[ameerat42]

OK, so you do not have any problems in that area! OK, back to square one.
Am(puzzled).

[mpb]

I think as rackham said the main issue is backlit birds with the front of the bird in shadow.
The metering system is trying to average the exposure in the area being metered when in multi segment/patern metering. Therefore because most of the scene is in sunlight and not in shadow so it will darken the shot to compensate.
Some of the shots have been cropped so we can not see the full view that the camera was seeing.
The landscape is dark probably because it was shot with spot metering, so it is really sensitive to what was actually metered, ie if the spot was on the sky them the shot is probably as expected.

[arthurking83]

The landscape image was indeed spot metered.
This could account for the dark exposure. If the central focus point was chosen as the spot, and the meter catered mainly for the sky section of the horizon, then this would account for the dark exposure.

Pixel luminance values of the slight cloud mass along the horizon are roughly in the 140 ballpark, so this makes the most sense.

And if mpb's comment that the images have been cropped in any way is true, then this also makes it harder to assess the exposures.

Also I noticed that the OP uses Ps Elements(ACR7.1) .. in my experience with ACR via Lightroom 4, is that it usually tend to render a raw file more bright when compared to the rendering on the camera and on the Nikon software I use.
If you have access to Nikon's ViewNX2, I'm wondering how different the image rendering looks between the two softwares(assuming that you shot in NEF format!!).

[glennb]

The landscape image was indeed spot metered.
This could account for the dark exposure. If the central focus point was chosen as the spot, and the meter catered mainly for the sky section of the horizon, then this would account for the dark exposure.

Pixel luminance values of the slight cloud mass along the horizon are roughly in the 140 ballpark, so this makes the most sense.

And if mpb's comment that the images have been cropped in any way is true, then this also makes it harder to assess the exposures.

Also I noticed that the OP uses Ps Elements(ACR7.1) .. in my experience with ACR via Lightroom 4, is that it usually tend to render a raw file more bright when compared to the rendering on the camera and on the Nikon software I use.
If you have access to Nikon's ViewNX2, I'm wondering how different the image rendering looks between the two softwares(assuming that you shot in NEF format!!).

Hi mpb and arthurking, No cropping has been done, I was 3 to 4 meters from it, so pic is what the camera is seeing. No post processing done except change to jpeg for flicker, straight from camera to photoshop organiser to edit for jpeg to flicker.(but disregard pics with my water mark on it on my flikr page they have been edited) I will try Nikon viewNX2 and see if theirs a difference. cheers Guys

Also is there a particular test I can do to see if I have a exposure problem with the camera?cheers

[arthurking83]

At this stage, I don't think testing for exposure accuracy is important!

FWIW: I just loaded a couple of images(NEF) into Rawtherapee(raw converter) and they came out a touch less than -1Ev darker looking by default than the same raw files look in ViewNX2.

But with this particular image, the exposure level in LR4 is about the same as that seen on the camera, and hence also in ViewNX2.

So again we have one level of variance in raw file interpretation between some raw file converters, and not others. But I've seen many instances where LR4 and ViewNX2 differ quite a bit too.
Nikon's software(ViewNX2 and CaptureNX2) will always render the image as per the settings on the camera, as it takes into consideration the camera's Picture Control settings .. which many other raw converters don't.

As for a quick'neasy exposure test ... a plain white(ish) wall .. reasonably lit, and just try to expose it neutrally(ie. no exposure compensation). Your WB can be set later on if you choose raw file type in the camera.

Test for all three metering modes, and you may notice very slight differences in exposure between all three modes .. and also test at wide open aperture settings as well as a mid range setting(eg. f/5.6 or something)
Make sure that what ever you're looking at, the centre pixels in these white wall test images should be registering approximately 128 in RGB luminance levels. Doesn't have to be exact tho.
In many instances, you will get some vignetting at the periphery, and luminance levels may be significantly lower.

If your walls aren't very white, then a sheet of A4 copy paper can substitute well enough for such a non scientific test to see if you have any issues with your gear. You don't necessarily need to have the entire white copy paper as the image frame, just a major portion of it in the centre, especially for spot metering. You then use this white paper as your white point to set a whitebalance value.
Note that domestic CFL lights are notoriously inaccurate in their colour rendering properties(very low CRI values) .. if you have CFL lighting in your home/office/etc.