MkIII, AF test, and Microadjustment

So I printed out the commonly accepted "test sheet" and stated to "calibrate" my lenses. I found the 70-200 f/2.8 to be spot on. Move on to a 150mm Macro and I had to adjust it +20 just to get close. I thought it was a little weird, but figured maybe something was funny because it was a Sigma lens and not Canon. So ok, I accept that and move on. However, then the 50mm and 24-105 also have to be adjusted to nearly +20 to make them work with that focus chart.

I start to think somethings funny, so I just start focusing on things farther away, perpendicular to the camera with details that I can zoom in on. Sure enough, now the 24-105, 50, and 150 macro are all much softer than they ever were in the past. Microadjust them back to 0 and theyre all tack sharp again. Put on the 70-200 and adjust it about +7 and now its sharper than it has ever been before.

Has anyone else ran into problems with that focus chart? Or even using a ruler or something else like Canon says? Seems to me, best results are had by focusing on something in front of you, not at an angle, and then zooming in to see how sharp it is, no?

to my eyes, all of my lenses perform well. I wouldn't touch that adjustment with a ten-foot pole.

mark

MDJAK wrote in post #4065527
to my eyes, all of my lenses perform well. I wouldn't touch that adjustment with a ten-foot pole.

mark

I felt the same way at first but thought Id just measure the front/back focus to see if mine were doing it. While adjusting them using that 45° chart made everything worse, now that Im doing it with what I feel to be the proper way, it has made all my images sharper.

For most of my lenses, some experimentation revealed that the adjustments weren't bringing any positive benefit. 0 no worse than anything else in the +/- 5 range. However, my 70-200 2.8L IS was a seven headed demon to adjust. It took all weekend and a lengthy bit of fidgiting, but I finally got something that was better than the default. For my lens, the trick was a bunch of test shots (like 50+) as each adjustment position. In between each one I'd run to focus out to the stop at infinity or the near limit, let it focus, then take the shot. After taking a bunch of these and recording the MA for each, I'd grade them on acceptable focus with a yes/no. Once I put them in a spread sheet and looked at the data side-by-side, it became quite obvious that a +8 (or something) was the best performing setting. This was in stark contrast to my 24-105 4L IS, which adjusted to a +1 with perhaps 7 tests, and the correct setting was quite conspicuous after only a few shots.

Moral of the story is that I guess some lenses have more "personality" for the MA than others, even when compared to one of the same model. Just food for thought for anyone that may have had similar problems. I would, however, like to hear what your process was (not just the OP's), particularly if it was something radically different than the "standard" 45° and set it off focus. I did mine with a (printed) ISO12223 chart hanging on a door at the other end of a hallway in RAW.

Imaginos wrote in post #4065791
For most of my lenses, some experimentation revealed that the adjustments weren't bringing any positive benefit. 0 no worse than anything else in the +/- 5 range. However, my 70-200 2.8L IS was a seven headed demon to adjust. It took all weekend and a lengthy bit of fidgiting, but I finally got something that was better than the default. For my lens, the trick was a bunch of test shots (like 50+) as each adjustment position. In between each one I'd run to focus out to the stop at infinity or the near limit, let it focus, then take the shot. After taking a bunch of these and recording the MA for each, I'd grade them on acceptable focus with a yes/no. Once I put them in a spread sheet and looked at the data side-by-side, it became quite obvious that a +8 (or something) was the best performing setting. This was in stark contrast to my 24-105 4L IS, which adjusted to a +1 with perhaps 7 tests, and the correct setting was quite conspicuous after only a few shots.

Moral of the story is that I guess some lenses have more "personality" for the MA than others, even when compared to one of the same model. Just food for thought for anyone that may have had similar problems. I would, however, like to hear what your process was (not just the OP's), particularly if it was something radically different than the "standard" 45° and set it off focus. I did mine with a (printed) ISO12223 chart hanging on a door at the other end of a hallway in RAW.

I did the 45° test and it was horrible. While I was slow here at work, I put the camera on a tripod and just had it focus on books all the way across the store. Then Id zoom in on the viewfinder to and check border sharpness ( as well as I can ). All the previously calibrated lenses using the 45° method were horribly soft so I recalibrated accordingly, and found all the borders to be sharper than I had ever seen them.

Just for the fun of it, I checked them against the 45° test again and while the "Focus here" line is still tack sharp, the bottom numbers stay within acceptable focus much longer than the top numbers.

I did these same things too. I erased all my AF microadjustment profiles back to 0 and now it's all good. Don't do it unless it is necessary.

jdizzle wrote in post #4065851
I did these same things too. I erased all my AF microadjustment profiles back to 0 and now it's all good. Don't do it unless it is necessary.

too bad really. i hadn't gotten around to microadjustment yet (and will still try it out for myself) but its a bummer people haven't gotten better results

I did the 45 test with the chart. I had varying results that didn't match what I saw in the field. I substituted fishing lures (with good sharp treble hooks) at a 45 angle (hanging over the chart). This work a heck of a lot better. My logic is that a 3d subject with contrast is easier to focus on than solid white or solid black, even if it does have graduated lines. May not be good science, but it works for me.

Cathpah wrote in post #4066169
too bad really. i hadn't gotten around to microadjustment yet (and will still try it out for myself) but its a bummer people haven't gotten better results

Ya. Don't touch it if you don't need to. I think Pekka had done some Micro AF adjustments on his 85 L if I can recall but, it was only a minor adjustment.

Cathpah wrote in post #4066169
too bad really. i hadn't gotten around to microadjustment yet (and will still try it out for myself) but its a bummer people haven't gotten better results

Fine results. Like I said, Ive been able to tune it sharper than its ever been...Its just that I think the 45 degree test is not very good.

AdamLewis wrote in post #4066889
Fine results. Like I said, Ive been able to tune it sharper than its ever been...Its just that I think the 45 degree test is not very good.

did you try the focus cube test? I think that's what worked for CDS or Pekka on here (can't remember who). I'm still travelling for anohter couple of months till i'm home in a place where I can take the time/have the environment in which I can keep a focus sheet or focus cube flat, straight, and clean.

Hi Guys

Sorry to appear a bit thick, but am I missing something - are you saying that there is a micro adjustment which enables us to make corrective adjustments to our lenses or cameras? If so, does it apply to the 40D?

Ray.Petri wrote in post #4067947
Hi Guys

Sorry to appear a bit thick, but am I missing something - are you saying that there is a micro adjustment which enables us to make corrective adjustments to our lenses or cameras? If so, does it apply to the 40D?

no, only the 1D MKIII at present (1DsMKIII when available has it too).

Thanks Toby

I new I should have convinced my wife that it was necessary for her get me a Mk3.

i tend to find that the older the lens is, the more likely it is to need af adjustments. a used 16-35 and used 28-70 (two of them), both need varying adjustments from -15 to +15 (16-35 is a **** to adjust... just too much dof). 70-200IS needs nothing though.. as well as a 300, 50mm and 85 1.8