80d micro focus adjustment results and observations
May 15, 2016
4
After getting my 80d a few days ago I decided to try adjusting micro focus for the lenses I have. This was my first time doing this with any camera and I started with the dot tune method. I used a joby gorillapod set on a chair, a focus chart taped to the wall and natural lighting via windows in the room (not exactly an ideal setup).
Using the dot tune method alone, here are the initial results I got:
EF-S 15-85 Wide end: -16 to 0. Initially set to -8.
EF-S 15-85 Tele end: -12 to +6. Set to -3.
EF 50 STM: -12 to +6. Set to -3.
EF-S 24 STM: -12 to +5. Set to -3.
EF-S 55-250 Wide end: -20 to +1. Set to -10.
EF-S 55-250 Wide end: -10 to +7. Set to -2.
After my initial attempt, I went for a walk with the 15-85. What I found was at the tele end it was as sharp as I had ever seen it, and gave the same results as live view. At the wide and middle parts of the zoom, it was not as sharp. I initially thought it might be a variance due to distance, but the good news was, distance did not seem to affect it. So, I came back home and tried again. After some experiments with both the dot tune method, and simply focusing and then manually checking focus in 10x zoomed live view I settled on a setting of -3 for the wide end as well.
Focus seemed to be inconsistent when the 15-85 was at its very widest setting (against the stop at 15mm), and I could end up with a correction result anywhere from -10 to 0. Zooming the lens in ever so slightly gave more consistent results. During this process I noticed that live view does not always get the focus as exact as I could manually adjusting with the 10x view. This could be due to some mechanical slop affecting live view, or it could be due to the relatively poor lighting and stability with my setup. Although I started out with IS off, I actually switched it on to get more consistent results. Live view focus error was mostly at the wide end of the zooms and the 24mm prime. Zoomed in, or with the 50mm, live view and manual focus gave similar results. Could be that the narrower plane of focus helped it get better results.
Having corrected the 15-85, I then went back and checked the other lenses a 2nd time. Based on my experience with the 15-85, rather than using the dot tune method, I started with the dot tune calculated value, then adjusted up and down while comparing focus results with live view zoomed in 10x.
The 50mm gave excellent results at -3. The narrow depth of field helped as it was easy to see the different just focussing with the lens at -4 or -2 in comparison to -3. In comparison, at the 15mm end of the 15-85, 1 step adjustments were indistinguishable.
The 24mm was not quite as easy to set as the 50mm, but still pretty easy overall. I concluded that while -3 was the closest to optimal setting, it actually needed something slightly more than that, but not quite as much as -4.
Finally, the 55-250 seemed to give similar results at the wide end with the MF adjust set to either -10 or -5. Both looked about as sharp as manually focusing. It could be that I was just getting tired at this point. The tele end was easier to set, and -2 seemed correct there.
So, I found the best method for this camera to start with the dot tune method, and then fine tune by adjusting MFA up and down slightly from the initial setting and check focus results on the live view display to find the mid point value that give the best results.
After all that, here are my final results:
EF-S 15-85 Wide end: -3.
EF-S 15-85 Tele end: -3.
EF 50 STM: -3.
EF-S 24 STM: -3. (actual is between -3 and -4, but closer to -3).
EF-S 55-250 Wide end: -10, but my confidence is low on this one.
EF-S 55-250 Wide end: -2.
Having done all this, I am very impressed with the consistency and accuracy of focus I am able to get with the 15-85 and both primes. Coming from an XSi it is quite unbelievable how sharp these lenses are with the 80d. I can also conclude that the camera needs an overall adjustment of -3, and so will start with that when attach any new lens. I actually tried the the MFA set to all lenses for -3, but then discovered that ignored any per lens setting, so have kept the MFA setting to be per-lens to have the flexibly of adjusting for individual lenses. I need more time with the 55-250. I am able to get sharp results, but am not yet convinced I have it set correctly at the wide end.
Just for fun, here is a 100% crop of one of the short focus distance test images I took with the 15-85 at 85mm. Admittedly this is not a scientific test of the focus accuracy as I picked this from a few pictures shot with focus tracking, but subjectively camera and lens did a good job of staying in focus at close to minimum focus distance with the MFA adjusted for mid to longer distances.
