Canon 40D questions: exposure and ISO

To meter the scene one of the things the camera may* do is to compare
the meter's data to a scene from a set (I've heard that it's 100's or
1000's) of stored image types. For example, a bright area with
darker area across the bottom is probably a landscape shot.
What F1_Fan writes here does ring a bell with recent p&s cameras, but as far as I know, the 40D doesn't work in this way. According to Canon whitepapers, the evaluative metering in question is just matrix metering with some "secret sauce" in the algorithm.

The camera breaks up the scene into 35 zones and meters each of those zones separately to come up with a "best" exposure based on the average light readings and the dynamic range... with emphasis possibly given to the area that's in focus. The "secret sauce" (the part of it that they tell us about, anyway) is in the fact that it uses the distance information reported by the lens to determine what's "important" in the scene.

http://www.usa.canon.com/uploadedimages/FCK/Image/2007/White%20Paper/40D/EOS_40D_WhitePaper_070817.pdf

Here is an excerpt from the 1DsMk3 whitepaper about how evaluative metering works in that body:
--------
Source:

http://www.usa.canon.com/dlc/controller?act=GetArticleAct&articleID=666&productID=109&articleTypeID=125

The new evaluative metering algorithm is based on the algorithm for the previous 21-zone and 35-zone metering systems. With the optimized 63-zone metering sensor and improved algorithm, more consistent and correct ambient and flash exposures are obtained with less influence by the subject. The basic concepts for the evaluative metering algorithm are:
1. Metering is weighted on the linked AF point.

2. If there is a very bright object in the picture, the exposure will be increased.

3. In backlit scenes, the exposure will be increased. With dark backgrounds, the exposure will be reduced.
--------

geek
 
What F1_Fan writes here does ring a bell with recent p&s cameras, but
as far as I know, the 40D doesn't work in this way. According to
Canon whitepapers, the evaluative metering in question is just matrix
metering with some "secret sauce" in the algorithm.
You may be right and I may be wrong... from what I've read I thought Canon's DSLRs did a simplified version of the P&S system.

A P&S has access to the entire scene including colour information so the scene can be easily matched to an internal database. I thought I remembered reading that the multi-zone exposure sensors in DSLRs were used to do something similar... maybe in that case the comparison is against many less image types.

Maybe I'm just wrong :)
 
... but thinking about it a little more, I'm pretty sure the DSLRs wouldn't work this way... in a p&s, the sensor is always active and exposed (when the camera's on). In a Canon DSLR, the sensor isn't exposed to the scene until the shutter opens, which happens after the exposure has already been determined. So I'm pretty sure this extra step of comparing scenes with a database of stock sample scenes isn't happening. I'd think it's more likely that some statistical functions are run on the light samples from the 35 zones, and exposure settings are chosen to maximize dynamic range while prioritizing whatever zones are in the plane of focus.

Dang, that's a lot of work for a sub-second response time. I think I'm going to give my 40D the day off!

--
geek
--
A 'must watch' for forum participants everywhere!
http://www.albinoblacksheep.com/flash/posting
 
Thanks to JimH and just a web geek for your more than helpfull answers.

I was in the same situation as tesebas after I upgraded from a 350D. Most of my pictures looked underexposed compared to what I was used to.
What confused me is that underexposing was not consistent.
Now that I now more about metering, I understand why.

I'm still a little surprised not to have experienced anything like this after using a 350D for 30 months and the fact a friend took similar pictures (same time, same place) and got better results (even if somewhat overexposed with a 400D!). I sent my 40D to have it checked and will come back when I have the results.
 
Thanks to JimH and just a web geek for your more than helpfull answers.
I was in the same situation as tesebas after I upgraded from a 350D.
Most of my pictures looked underexposed compared to what I was used
to.
What confused me is that underexposing was not consistent.
Now that I now more about metering, I understand why.

I'm still a little surprised not to have experienced anything like
this after using a 350D for 30 months and the fact a friend took
similar pictures (same time, same place) and got better results (even
if somewhat overexposed with a 400D!). I sent my 40D to have it
checked and will come back when I have the results.
Same here... with the inconsistent underexposure and with the overall underexposed look... Very rarely (almost never in 3000 clicks) that I got a overexposed pic with my 40D. Also, I have a friend that has a 350D and upgraded to 40D - same experience (he still grabs the 350D when he wants a better exposure).

I would really appreciate if you could please let me know (here or through email) if anything changes when your 40D comes back

--
Cheers,
Gene
 
No problem. It won't before another week or so though.
How can I email you?
 
Same here... with the inconsistent underexposure and with the overall
underexposed look...
One thing that can cause underexposure is when light enters the viewfinder. If the camera is on a tripod the finder needs to be covered and if the camera is handheld your eye should be held tight against the finder. I saw a guy shooting a 40D the other day without the eyecup installed. Not a good idea with regard to accurate exposure.

I was getting under-exposed images with my 40D when it was new but the problem has vanished. Not sure what setting I changed but now it works how I expect! Make sure the wrong viewscreen is not accidently selected in the custom functions. I know this will throw the exposure off but that wasn't my problem.

--
Mike Mullen
 

Keyboard shortcuts

Back
Top