PDA

View Full Version : Miss-behaving 20D AE


FlyingPete
28th of December 2004 (Tue), 19:00
Well, here I am away from home, just got a dial up connection to the web via dial-up (for those who have forgotten, you used to use a phone line to connect to the web!) to check my mail.

Anyway, doing lots of shooting as one should on holiday, although the 20D with a 28-105 (old version) gets pretty heavy after a while, don't know how you guys do it with your L glass! I did also have my one year old girl on my back at the same time as well though.

Anyway, I have been having some exposure fun with my 20D, esspecially with my CPL filter attached. A image that appears nice and clear with great blue skies etc in the view finder is comming out over exposed, by my reckoning around 2/3 stop (no exposure compensation is not set). The net affect is a loss of colour, normal coloured skies, and anything white being blown out.

Here is an example, only editing is resizing for web:
http://www.lowden.net.nz/Stuff/lake.jpg

Details are:
1/125s, F/8, ISO400, 65mm (28-105) Center weighted metering, no exposure compensation, WB auto. Image taken at around 3pm.

robertwgross
28th of December 2004 (Tue), 19:42
"Center weighted metering" is the one thing that I might suspect.

About the only time I use center weighting is when I have one distinct subject way out in a telephoto lens. In your shot, if the metering was centered on the medium trees in the center of the shot, then this might explain a few bright spots around the edges.

---Bob Gross---

PacAce
28th of December 2004 (Tue), 19:43
I don't think your 20D is behaving at all. The pictures averages out very well to a mid-gray (I checked using PS) so the camera meter is doing what it's supposed to be doing.

FlyingPete
28th of December 2004 (Tue), 22:58
I don't think your 20D is behaving at all. The pictures averages out very well to a mid-gray (I checked using PS) so the camera meter is doing what it's supposed to be doing.

How do you perform that check?

I supose it is just another example of when not to trust the internal meter. Perhaps the polarised sky and dark foliage caused the meter to over expose.

PacAce
29th of December 2004 (Wed), 07:36
How do you perform that check?

I supose it is just another example of when not to trust the internal meter. Perhaps the polarised sky and dark foliage caused the meter to over expose.

It's not a matter of trusting the internal meter. As long as your camera is functioning properly, you can be almost 100% sure that the meter will be working corretly. But you have to realize what it means when we say that the meter is working correctly. What it means is that, with EC set to 0, the meter will expose the shot to give you an overall exposure of an average midgray tone when using averaged metering. With Evaluative metering, things get a little more complex but it start out with the same average exposure and then tweeks it from there.

Now, the meter reading you get doesn't necessarily mean that that's the correct exposure you want. That why there is such a thing as EC and the histogram. You have to check the histogram to see if you like what you see and, if not, adjust the exposure with more or less exposure than what the meter is calling for.

BTW, to see the average tone of you image in PS, click on Filter > Blur > Average. This will turn you image into a single solid color, usually gray or somthing close to it. Next, click on Image > Adjustments > Curves and then click on any part of the image with the eye dropper. You'll see a dot on the curve line which coincides with the tone of the image. If you used the eye dropper on a regular image and pointed it at different spots on the image, you 'll be able to see the luminosity of the image at that exact spot where the eye dropper is. It comes in handy when you're trying to adjust an image, using curves, for certain tones without affecting the others.

P.S. Don't forget to cancel out of the Blur Average change before saving the image! ;)

FlyingPete
29th of December 2004 (Wed), 12:19
Hey Leo, thanks for all that! Up untill recently with my SLR, it has been take some shots over a day or two then get them developed (usually slide, so they had to be sent away) over a week or so, and get them back.

Because I no longer kept a diary of the shooting settings I used (I did this when starting out on my OM1 15 or so years ago) I would look at a shot and not really know what I did wrong.

Because I didn't see the results straight away, and there was a real cost associated with each shot, I vert rarely took risks or tried something completely new (bounce flash, EC etc).

Now that I am shooting and making mistakes I am learning very quickly again, the 20D does not seem as forgiving as my G3, or then am I trying to do more?

Yesterday I took some shots out in the volcanic plateu, as my subjects were distant (volcanos), there was quite a bit of haze. snow and a lot of sky so I tired an EC on -2/3 seemed to work quite well.

Jonny
29th of December 2004 (Wed), 12:56
by my reckoning around 2/3 stop

When people say 2/3 stop do they mean; A) 2 thirds of a stop or B) 2 to 3 stops.

PacAce
29th of December 2004 (Wed), 13:16
When people say 2/3 stop do they mean; A) 2 thirds of a stop or B) 2 to 3 stops.

What I would mean by 2/3 stop (and I'm sure Pete does, too) is A) two thirds of a stop. 2 or 3 stops is pretty drastic to be making for exposure compensation, although that's not to say that one can't. :)

Jonny
29th of December 2004 (Wed), 13:41
What I would mean by 2/3 stop (and I'm sure Pete does, too) is A) two thirds of a stop. 2 or 3 stops is pretty drastic to be making for exposure compensation, although that's not to say that one can't. :)

thought so, thanks for clearing that up for me.

FlyingPete
29th of December 2004 (Wed), 14:07
2 or 3 stops is pretty drastic to be making for exposure compensation, although that's not to say that one can't. :)

I did some more messing around with this and the CPL attached, to get the nice blue skies on landscapes, 1/3 - 2/3 seems to work well, 1 stop does nice deep blue skies, however at the expense of foreground.

Are there any 'rules' in certain conditions, such as snow, in the bush etc?

PacAce
29th of December 2004 (Wed), 14:48
I did some more messing around with this and the CPL attached, to get the nice blue skies on landscapes, 1/3 - 2/3 seems to work well, 1 stop does nice deep blue skies, however at the expense of foreground.

Are there any 'rules' in certain conditions, such as snow, in the bush etc?

No rules that I can really think of. It's more of learning from your own experiences, really. I've read a lot of books on wildlife photography and I could never remember much of what I had read when it came to applying the "rules" to a real shooting situation. It was only after actually shooting a couple of shots and then adjusting the EC until I got the picture just right that I was able to remember how much to adjust my exposure the next time I encountered a similar situation.

For example, the scene that you posted at the top. I would have shot that with -2/3 EC right off the bat because from previous experiences, I learned that when the middle third of the frame is foliage with dark shadows in it, the top third is the open sky (not overcast) and the bottom third is water and the lighting is from the side, the picture usually comes out over exposed by 2/3 stop.

So, the more you go out and take pictures and then analyze your results at home, tyring to figure out why the badly exposed pictures are that way, the better prepared you'll be for the next time.