Anyone got a rational explanation why Canon underexposes?

Jon wrote in post #2394853
Unfortunately, to do this on the CMOS sensors Canon uses won't work, since reading the sensor is essentially a destructive operation.

You will have to explain that a little more Im afraid.
The 20Da had a live preview option, so its possible for data to be drawn off the sensor in a continuoes fashion.
What is the "destructive operation" that prevents data being pulled off and analysed during an exposure?
I understand of course if this involves lots of detailed electronic discussion that is beyond resonable explanation, and you may state so and leave it at that.

Basically, reading the sensor drains off whatever's in the sensor sites. So the camera needs to start over again from there building up a charge. Think of CMOS, in it's simplest form, as dynamic RAM. The computer has to restore the charge to the memory site every time it's read. With RAM, it's just a 1 or 0; with a sensor, there's more to deal with.

If I understand the 20Da's LCD display correctly, you had to switch between viewing and taking. And, like P&S digitals, the display would "lag" if you were moving the camera - it was more for focussing than for framing, so the blocky and slow refresh wouldn't be as noticeable.

Just an interesting side note, when I use the Shadows/Highlights tool Photoshop Elements it defaults to a +25% brightness increase in the highlights. It may be because the camera exposure is dark, or maybe more people "prefer" the lighter look. But cosworth must not be the only one thinking this. I don't worry about it though, if I have time to worry then I should be out taking pictures instead.

-Ryan

My 30D exposes perfectly it might be a bit dark when using spot on a small subject. But in those cases i usually lock of something that will trick it

But the 400D/XTi might as people have said because their pixels are so small and have a greater chance of blown out highlights and smaller Dynamic Range.

Jon wrote in post #2395498
Basically, reading the sensor drains off whatever's in the sensor sites. So the camera needs to start over again from there building up a charge. Think of CMOS, in it's simplest form, as dynamic RAM. The computer has to restore the charge to the memory site every time it's read. With RAM, it's just a 1 or 0; with a sensor, there's more to deal with.

Gotcha!
I understand it now.
I can deffinitly see how that would cause a problem during an exposure