Quick 50D AF Microadjustment

Just got my refurbed 20, er, 50D today (thats gonna take time). I was excited to finally see if my lenses were focusing properly, and nearly each one was not! My quick method was to use an engineers ruler (triangular with lots of hash marks), angle it and set the camera for max depth of field. I then focused straight on the ten or whatever number I chose. Seemed to work well with all but the 35mm f/1.4L, it was right on the numbers for the ruler, but was off for portraits? So many variables there so I just went for the middle ground for now with that one. My others lenses needed as much as 11 units either way. Anyway, I'm going to do a full test tonight, with tripod and remote. But for now, it seems to have sharpened up my lenses for sure.

gregpphoto wrote in post #8534144
Just got my refurbed 20, er, 50D today (thats gonna take time). I was excited to finally see if my lenses were focusing properly, and nearly each one was not! My quick method was to use an engineers ruler (triangular with lots of hash marks), angle it and set the camera for max depth of field. I then focused straight on the ten or whatever number I chose. Seemed to work well with all but the 35mm f/1.4L, it was right on the numbers for the ruler, but was off for portraits? So many variables there so I just went for the middle ground for now with that one. My others lenses needed as much as 11 units either way. Anyway, I'm going to do a full test tonight, with tripod and remote. But for now, it seems to have sharpened up my lenses for sure.

Well, first of all I'd recommend you read one of the other threads on this topic (down in the left-hand corner). Especially those that explain why a target such as the one you used is totally wrong for MFA.

hollis_f wrote in post #8537192
Especially those that explain why a target such as the one you used is totally wrong for MFA.

And hows that? A gradiated ruler, how can that not be useful for determining exactly where the camera is focusing?

Lets take what we know. We know AF works on contrast (black hash marks on a white ruler), and we know that depth of field causes only one spot to be in focus at wide open apertures. So then, explain how this method, while not as precise as the others (and it was never intended to be), isn't proper in your eyes?

gregpphoto wrote in post #8539089
And hows that? A gradiated ruler, how can that not be useful for determining exactly where the camera is focusing?

Lets take what we know. We know AF works on contrast (black hash marks on a white ruler), and we know that depth of field causes only one spot to be in focus at wide open apertures. So then, explain how this method, while not as precise as the others (and it was never intended to be), isn't proper in your eyes?

Quite simple, really. The AF sensor covers quite a bit more area than the little red box in the viewfinder. So it's also going to cover more of your ruler than just the bit you're focussing on. So you've no way of knowing which of those nice, contrasty, hash marks it decides to focus on. If you want the camera to consistently focus on the same plane each and every time then you need a target that is parallel to the focus plane and covers the AF sensor completely. As I mentioned before - the other threads contain numerous suggestions on how to combine such a target with an angled ruler to give the optimum solution.

It's not just me who thinks this. Try reading Chuck Westfall's words on the topic.