[palmor]

Thanks to all for you input.

I have just uploaded 4 pef files on a public ftp site:

ftp://ftp2.landtrekker.org/landtrekk/pef/

I am anxious to hear from you

Best regards

Marc

The only thing I can think of (and I'm sure others will give their opinions) is that for each of those photos you used spot metering, and in each photo the center of the image was on a "bright surface", something that possibly reflected a lot of light (at least brighter then the rest of the photo). Just like shooting in snow if you meter off of something bright the camera will underexpose.

When using spot metering you should meter off the area that you want to get correct then AE-L your exposure, then take the photo. This will lock in your exposure for the correct part of the photo.

I think the only way to tell if you have a problem with the camera or just bad luck with what you chose to expose on is to setup some controlled tests and use the different metering modes to see.

Hope that helps
John

[jeffkrol]

Thanks to all for you input.

I have just uploaded 4 pef files on a public ftp site:

ftp://ftp2.landtrekker.org/landtrekk/pef/

I am anxious to hear from you

Best regards

Marc

Ok, just a start but the 2 I downloaded show some possible reasons for the camera to pick the exposure they did..... Just shows a bit of a learning curve difference and the fact the camera leaves a lot of thought up to you.
1 photo has a white sheet of paper, or building or something filling over half the frame.
As a camera (well DSLR's in particluar, and Pentax especially) has to boil all the data down to 1 value (EV) in order to set the shutter/aperature it has to make some decisions. Pentax is and does use a lot of old school approach to metering. It took a photo that internally averaged to 12% (use 18% if you must) which is really close to half black (18% grey= 50% black). Soooo the predominant white object will, as it should, come out grey. If you would have spot metered on the white area and then increased that exposure by +2 ish it would have come out as most people would have expected.
The other image I looked at had a flood of sunlight in the upper right quadrant and could easily overwhelm the meter.....
Sooo the 2 I looked at came out pretty much as I would expect... but then I'm easy to please.....
I

[jeffkrol]

What version of Pentax Photobrowser supports K20?
3.12 will show thumbnails but EXIF ect. is missing.......

[axl]

Looking at the first two shots (I wasn't able do download any others - and even on those two my Pentax Photo Browser didn't show any exif data) I'd say you had multisegment metering without any Ev compensation. As I wrote before, Pentax metering ALWAYS goes for highlights and tries to preserve them.
In first shot the sun is shining right into the lens, so I'd guess the camera metered off it, which would understendably resuld in heavy underexposure. Second shot has the sun hidden more behind the branches but the sky is still too bright so camera meterd of it.
Anyway shooting against the sun with any camera would result in underexposure on 99% of ocasions unless the USER didn't intervene!
So at least in case of the first two shots, don't blame camera but yourself. K20D as good as it is, is just the machine and it tried to keep as much detail as possible in very difficuld situation. These are the sort of shots where photographer's knowledge has chance to be put into practise...

[mrechte]

Thanks for your feedback.

Please could you mention the picture name your are talking about, it would help the understanding. I have not much understood the % given.

Pictures 353, 356, 368 and 387 were shot with spot mesure. I could understand the explanation for 387 wich has a white and bright center. But this is not the case of 368.

Also I posted some 2 (small) jpeg files in /jpg directory (same ftp site):

1109 is multi-zone mesured
1110 is spot mesured (on the face)

I would expect 1110 to be better exposed (considering my subject is on a bright white background).

Aloso the 905.pef and 906.pef is interesting: 905 is multizone wheras 906.pef is spot measured on the signpost. 906 is worse than 905.

I (or my K20D centre zone) is confused !

Pentax Browser is 3.50

Thanks again

Marc

[Mallee Boy]

Hi Marc,
I opened the first shot (the sign with sun in the background). Your camera AF area is set to multi segment, which means that is reading the entire screen and then choosing what it thinks is the best available exposure settings. What is killing that shot is the bright sun in the background, the camera reads that and says whoa buddy! way to bright, gotta underexpose to compensate.

That situation is about as hard as it gets to get a well balanced exposure.

Think you need to spend a bit more time with the good book (owners manual). A camera like the K20D does not guarantee better shots than a Point & Shoot compact, which will automatically make all sorts of instant adjustments to produce an ok to good image. DSLR's give you (generic termperator/photographer) the opportunity to move your photography up a level or two, they don't do it for you.

With a bit of homework and understanding your tool (the camera), and the principles of light/exposure you'll be taking great shots in no time.
Cheers
Grant

PS: also had a look at No 356, and much the same story. Although you have used spot metering, it depends on what 'spot' you focussed on, large black blobs on white surfaces are going to test the AF. In this shot you could have used flash.

[flyer]

Salut! Marc
Les membres du forum pensent que tes mauvais résultats sont dues à un manque d'expérience. Le posemètre du K20D (comme celui de tous les appareils photos) est calibrée pour une lecture à 18% gris.Donc, lorsque tu utilise l'option "spot" et que tu pointe ton appareil sur un objet qui est plus brillant qu'un gris neutre de 18%, ton appareil compense pour ajuster l'exposition à 18%, ce qui fait paraitre les photos sous-exposées. En mesure "spot", les tons de chair ou le vert (herbe) sont de bons substituts à un gris neutre. Tu pointe ton appareil sur un de ces tons et, pendant que tu pointe, tu poussse le bouton AE-L, ce qui va mémoriser l'exposition pour te permettre de recomposer ton cadrage et de prendre la photo. Si j'étais à ta place, je commencerais en me servant surtout du mode de mesure matriciel. Je crois aussi que tu ferais un bon investissement en allant dans une librairie pour te procurer un bouquin sur la manière de gérer l'exposition. Si tu as d'autres question et que tu aimes mieux communiquer en français, tu peux m'envoyer un message personnel. Bonne chance et ne laches pas!
P.S.: Le K20D est un excellent appareil, mais, comme tous les outils de précision, il demande une certaine expérience pour en tirer le maximum.
Yves

[Canada_Rockies]

Thanks to all for you input.

I have just uploaded 4 pef files on a public ftp site:

ftp://ftp2.landtrekker.org/landtrekk/pef/

I am anxious to hear from you

Best regards

Marc

I downloaded the first file there, and your underexposure was caused by the sun in the view in that image. In a case such as that, you will need to switch to spot metering, and meter the sign.

As a hint, if you have any doubts at all about an image, use the bracketing function described in your manual. That way one of the exposures will be close.

Enjoy! It's a long road to learning the whole camera. I've been on the road since about 1957, and am still learning.

[mrechte]

I am happy that you spend some time downloading and checking my pictures.

Please note that I have beeen using SLR for 20 year and especially models with spot meter (introduced with Olympus OM2-spot) so the problem of understanding what is backlight shooting, etc. is not mine.

What I learn from the above is the (Pentax ?) 18% exposure. I take note.

Please take a bit more time to compare set of pictures I uploaded that are to me inconsistant:

905.pef to be compared to 906.pef

1109.jpg to be compared to 1110.jpg

Whatever factory % the light meter is set , you will agree that it should be the same % whatever mode one uses (spot versus multi-seg), right ?

So how can one expect that the same frame, with a major white and bright background and a smaller darker subject is exposed more with a multi-seg mesure than with a spot mesure on the darker subject ?

Thanks again
Merci flyer pour l'explication !

[mrechte]

I downloaded the first file there, and your underexposure was caused by the sun in the view in that image. In a case such as that, you will need to switch to spot metering, and meter the sign.

As a hint, if you have any doubts at all about an image, use the bracketing function described in your manual. That way one of the exposures will be close.

Enjoy! It's a long road to learning the whole camera. I've been on the road since about 1957, and am still learning.

Thanks for your comment.

This is exactly what I did, please compare 906 with 905 (same scene) 906 has been shot with spot mesure and is worse than 905 !!!

[dopeytree]

A cameras meter looks for a middle grey value so in a senario like this it is trying to balance out the very bright sunlight causing underexposure. In this situation i would punce in a +1/2 stop to a full stop? Its all about playing around.

[mrechte]

To ease the picture downloading, I converted the PEF files into smaller JPG files (with EXIF data). They are located on the same ftp site:

ftp://ftp2.landtrekker.org/landtrekk/pef/

[falconeye]

I agree with what most people have posted here. But I agree with the OP as well. Some (few) of the samples are underexposed for no obvious reason.

E.g. #0356

It could be due to the 18% grey rule, though. Because with spot metering, the big black area wouldn't count and exposure would be ok.

Of course, this seems to be sunlight from behind and there is another possibility we shouldn't exclude:

Light from behind right into the viewfinder!

@mrechte: are you wearing glasses?


And of course, spot metering doesn't work with split screens (I mention this because you are coming from analog SLRs...)

[mrechte]

Thanks falconeye for your judicious remarks. It is true that I wear glasses. I just walked outside to do some testing (for a change there is a good mid-day sun light today !). If I move away from the camera there is 1.5 IL biais in the measure (with the view finder in the sun light). But with the glasses in contact with it, I have not managed to biais the measure.

Also what is your opinion about the 2 couples of pictures 905/906 and 1109/1110 ? For my opinion I should have got opposite results (higher exposure with the spot measure).

Many thanks

[TonyC]

This is not an issue that would ever concern me because I never use auto exposure on a camera. I use a trial shot followed by a histogram check and after that I work by 'feel' with the occasional histogram check.

However to be fair to this poster this month's UK Professional Photographer magazine slates the K20D for (inconsistently) underexposing in bright situations. They said sky almost always needed +1 stop compensation. Other than that they were positive about every aspect of the camera.

Curiously the photo in the review that is supposed to be an example of this 'problem' shows a perfect (in my opinion) exposure of a very difficult subject. A white waterfall partly in the sun and partly in shade. There is full detail visible in both sun and shady areas - a remarkable performance in my opinion.