Microadjustment is Strange

I just sold off a 1D3 and picked up a 1D4 and so I noticed a few quirks of MA that I thought I would share. I'm just tossing these out there for you to contemplate or comment on:

For starters, about half my lenses needed some MA on the 1D4. More interesting, the amount and in one case direction of MA needed on the 1D4 was different from the 1D3. Doesn't it seem like different bodies should at least show a general trend even if there is an offset?

More interesting to me, I've been using most of these lenses for 3-4 years now. They have all always worked perfectly (and continue to work perfectly) with my 5D classic which is not capable of MA.

What are your experiences with MA?

I found it mostly a waste of time because most of mine are zoom lenses and in most cases the camera was hitting close to if not as good as the lenses would focus through their range.

I was adjusting too often when repeating the tests with just a variation of the room lighting.

I do think MA is a good tool for minor adjustments but it just wont make a bad lens good nor bad technique better which more describes me than anything.

I could see on one body all lenses needing, let's say, all positive MA and all negative on another. I cannot see all positive on one camera and then a mix of negative and positive on another.

The only lens that I have that is not 0 MA, is the 85/1.8

I have also found, and one may think I did not do it correctly, is that when doing a controlled test to determine MA, that, in real shooting, the results may prove different.

I have yet to find an explanation for this.

Perhaps this may shed some light on what you are seeing.

http://www.lensrentals​.com …s-is-soft-and-other-myths

bohdank wrote in post #9742786
I could see on one body all lenses needing, let's say, all positive MA and all negative on another. I cannot see all positive on one camera and then a mix of negative and positive on another.

The only lens that I have that is not 0 MA, is the 85/1.8

I have also found, and one may think I did not do it correctly, is that when doing a controlled test to determine MA, that, in real shooting, the results may prove different.

I have yet to find an explanation for this.

I have a method that I consider to be very good. I shoot a flat, high constast target that is much larger than the AF point at 50X the focal length. Flanking the target are two items 6" before and behind the desired focal length.

I AF 8 times in a row starting alternately from MFD and infinity each time. The first four AFs are done once, the last for include a 'touch up' second half press.

I note basically how the lens performs (consistent or not) and if it tends to favor front or rear focus. I perform MA to get the lens to consistently hit even focus at the target as much as possible.

Zoom lenses can be tricky, though my 17-40 and 24-105L require no MA. My 70-200 wants different levels of MA at wide and tele, as does my 100-400.

In the end, you wonder if all this effort is worth much when you spend 20 minutes microadjusting your 100-400 on your 1D4. Then you slap the lens on your old 5D and rattle off a bunch of perfectly sharp shots.

I think it is a convenient way for a massive camera company to avoid many service calls while cutting there cost in there Quality Control.

I posted my method on another thread. After using charts and reviewing images on a PC, I thought there had to be a better way. I am convinced that this is the best possible method since it removes any subjective decisions on which is sharper. Best of all, I can do the test again and I always get the same unambigious results.

I use the back of a Western Digital 1.5TB drive. The printing on the label is extremely fine and clear edged. I stand it up on a table. When perfectly in focus there is no ambiguity.

Yes, Liveview, 10x.

The 45 degree angle is not needed and serves no purpose other than to tell you if you are front/back focusing. I only want to know that what I am focusing on is at it's sharpest. Now, depending what/how you shoot you may prefer certain lenses slightly front/back focus.

I have found that using this method, the results are repeatable...in other words, every time you do the test, the result is exactly the same. There is also no danger that the camera may AF on something other than you intended.

And best of all, no images to review.

JeffreyG wrote in post #9742667
What are your experiences with MA?

Lenses that are sharp on my 20D needed AF Micro adjustment on my 50D..

Two in particular needed AF Micro adjustment.. My Sigma EX 50-150mm f/2.8 (+15) and Canon EF 50mm f/1.4 (+5).. My Tamrons are OK..

Usually it's not the camera or the lens alone but the combination of the two..

mikeassk wrote in post #9742880
I think it is a convenient way for a massive camera company to avoid many service calls while cutting there cost in there Quality Control.

This.

JeffreyG wrote in post #9742864
I have a method that I consider to be very good. I shoot a flat, high constast target that is much larger than the AF point at 50X the focal length. Flanking the target are two items 6" before and behind the desired focal length.

I AF 8 times in a row starting alternately from MFD and infinity each time. The first four AFs are done once, the last for include a 'touch up' second half press.

I note basically how the lens performs (consistent or not) and if it tends to favor front or rear focus. I perform MA to get the lens to consistently hit even focus at the target as much as possible.

Zoom lenses can be tricky, though my 17-40 and 24-105L require no MA. My 70-200 wants different levels of MA at wide and tele, as does my 100-400.

In the end, you wonder if all this effort is worth much when you spend 20 minutes microadjusting your 100-400 on your 1D4. Then you slap the lens on your old 5D and rattle off a bunch of perfectly sharp shots.

When you were finding a different MA wide than tele, was the offset equal? IOW Would a MA with the lens set in the middle of the range average out for both ends? IE a 28-80 zoom MA set at approx 55mm average out close at both ends or was it unacceptable at one end or the other?

mikeassk wrote in post #9742880
I think it is a convenient way for a massive camera company to avoid many service calls while cutting there cost in there Quality Control.

+one (+1 was too short )

I seem to get different results each time I adjust my lens. One moment they are at 0, or no more than +/-2 and another they are much higher. Ive noticed that if I use the AF assist throws results off as well.


I just placed an order for LensAlign and a cheap tripod to save my sanity.

Try my method, first.

bohdank wrote in post #9742938
Try my method, first.

Are you saying you autofocus and then check focus using Liveview at 10x?

I do it both ways. Focus in Liveview at 10x, then Quck focus watching the lens (distance window) to see if it actually moves. If it doesn't move at all I know it's right on. If it does, I MA and do it all over again. When I think I have it right, I reverse the process. In Liveview I Quick focus and then see if if MF gives me a sharper image. If not, I'm done.