Anyone with both 40D and 30D?

As you can see from my sig, I have both the 40D and the 30D.

And in my experience, the 40D focuses faster, but in one shot mode it misses focus quite a lot more often in low light with fast primes. It doesn't consistently backfocus or frontfocus, and focus with slower lenses (e.g., the 24-105L) is generally reasonbly good even in low light (or, at least it was some time ago...maybe I should recheck it). But even on those lenses it has missed from time to time. At the same time, there are some lenses, like my 85 f/1.8, with which my 40D seems to get dead-on accurate focus seemingly every time.

My 30D, on the other hand, appears to be rock solid with pretty much all my lenses, at least when focusing using the center point. It will nail focus every time with the same lenses that my 40D has trouble with.

My question is: is this typical behaviour for the 40D? If not, then it means I need to send mine into Canon for calibration and a good once-over. It also leads me to ask: what techniques does Canon Irvine use to test the 40D's autofocus system?

And finally, have any of you had such focus reliability issues with the 40D, sent your camera into Canon Irvine, and gotten back a significantly improved body?

My experience has been that the 30D focuses appreciably slower, the 40D focuses much faster and sort of polls the boxes much faster.
The downside to the speed and what seems to be the sensitivity of the 40D is that you the shooter have to be "better" at holding off for the shot.
It's almost like the 40D is going "focus here focus here focus here" where as the 30D is saying, "I think I will focus here, ok I am almost ready to focus HERE!"

Once I figures out the nuance of the faster and more sensitive 40D I haven't had a problem, spending more time with the 40D and then going back to the 30D the 30D seems "wrong" now.

the 40D outfocuses my 30D in all situations with all lenses in all the poor lighting conditions I work in. It's a real winner. But really, for day to day shooting, the 30D performs more than well enough.

Well, the reason I'm asking is that my warranty period is almost up and I need to decide very soon whether or not to send the camera in to get this sorted out. If what I'm seeing is just the normal 40D behaviour then there wouldn't be any point in doing so because they'd just claim that the camera "meets factory specifications".

But if what I'm seeing isn't how the 40D is supposed to behave then obviously I should send it in.

And the reason I'm asking people who own both the 30D and the 40D is that the 30D is the only camera I can currently compare mine against, so I need to know if people who own both have seen the same relative behaviour.

It sounds, however, like the 40D should be better than the 30D in all situations, so I'm beginning to think that I should send mine in.

Basically, the 40D AF is superior to the 30D in every way.

I haven’t experienced a single situation where a 30D would have been superior. My worst lens regarding AF consistency is the EF 50/1.4 USM. If the subject does not fill the frame and at the same time does not offer very good contrast, the 50/1.4 tends to focus on a spot somewhere else in the frame that has better contrast – no matter how far this spot may be away from the selected AF-point. This behaviour seems to be even worse on the 40D, but not by much. If, on the other hand, the focus hits the right spot, it is even more accurate with the 40D compared to 30D.

I guess the reason why we experience even more focus inconsistency on same ultra fast lenses is the improved centre AF sensor on the 40D, that is ultra high precision for lenses providing f/2.8 or faster aperture. Canon also added sensor elements for improving the detection of extremely defocused subjects. All this might be optimised for f/2.8 lenses.

Lenses that are significantly faster, obviously will introduce some problems generally, but most noticeable on a fast lens that has pronounced weaknesses wide open, such as the 50/1.4. The AF relies on images with well defined transitions, hence good contrast, and can stumble across smeared structures due to lens imperfections such as spherical and chromatical aberrations.

Let’s have a look at the EF 50/1.4 USM:
- Contrast is low wide open
- Chromatical aberrations (CA) and purple fringing (PF) are pronounced wide open

The bundle of light rays leaving the rear element of the lens is certainly much wider than on a slower lens. So, while the AF is optimised for f/2.8, it sees now a lot of additional light rays (coming from the borders of the lens) hitting at more extreme angles. This might overstrain the capabilities of the separator lenses in the AF system, ultimately leading to false focus lock.

In an attempt to improve the AF even further with each new generation, one of the design goals is to achieve focus lock even in low light and especially on low contrast subjects. With a super fast lens that provides much more light rays, hitting the (AF)sensor at somewhat unexpected angles, and that on top of that does not perform too well wide open, i.e. the light rays that have to pass the borders of the lens do not exactly hit the correct spot in the sensor plane, the camera AF could lock onto some sort of “ghost image” that relates to a completely different area of the frame.

Since the 40D is optimised to lock on images with even lower contrast than the AF modules before it could manage, and attempts to recognize even severely blurred images (heavily defocused) to initially grab the focus on, it also can be fooled more easily.

Just to sum it up:
In my experience, the 40D performs better than the 30D with an EF 85/1.8 USM. With the EF 50/1.4 USM though, 40D is better (more accurate) if it gets the correct focus lock, which is likely to happen on larger targets with nothing else in the frame that provides better contrast, but might fail more often that the 30D to grab the right focus spot if the before mentioned conditions aren’t met.

I've also read that some people have found the 40D to start hunting for focus in more situations than older cameras did.

I sometimes see this with my 40D as well. Now and then, I've taken out the 400D, to compare with, as I thought "Was this that bad before?". But then the 400D does the same, or can't focus at all, in the same situation. So I go back to the state of thinking "Well, this isn't optimal, but it's at least trying".

Check this https://photography-on-the.net/forum/showthre​ad.php?t=376605