Recommendations and experience doing AFMA on a 5D3

[davidmurray]

I'm looking to hear recommendations and experience to do with doing Auto Focus Micro Adjust for lenses with a 5D3.

I've never done AFMA and am looking to understand what is the most popular, accurate, and easiest method(s) to do this.

Your views?

 

[Valvebounce]

Hi David.
Reikan FoCal pro.
The camera model does not really come in to this, any Canon camera that can be tuned has the same basic interface and issues.
Repeatable results are the real requirement for making these adjustments, and some of the alternative methods (dot tune) have given some of us less than satisfactory non repeatable results.
FoCal needs loads of light and a firm support including the floor*, to get good results but it is worth the trouble. I use it on my 7D and 7DII, unfortunately most of the more recent cameras require assisted mode where you have to make adjustments as the software can not change the AFMA setting any more, it was fully automatic on the 7D.

*A wood joist and plank floor will require that you do not move around close to the tripod as the floor will flex affecting the result.
HTH

Cheers, Graham.

 

[AlanF]

FoCal is probably the best method, and I have used it for years, including with the 5DIII. Focal has only just updated in the last few days to include the 5DIV and so I used a simple method for AFMAing my 5DIV last week. Pin a chart to the wall - I use one from Bob Atkins site - and take 6 or 7 shots at every five AFMA numbers, and see where there is optimal resolution. Then home in with a fine analysis around the peak to tie it down. The camera was between 15-30 times the focal length of the lens from the target. It's importance to take several shots at each AFMA number as there can be variation in focus between shots. The whole procedure.

Others use a mechanical target like Lensalign. Some claim a method called "dot tune" works for them. It never did for me. FoCal does give you lots of interesting information about your lens in addition to the best AFMA, and I'll upgrade to the latest version with the 5DIV to satisfy the geek in me.

 

[davidmurray]

Thanks guys for giving your experience.

I had been on the fence re buying software for doing AFMA, but your comments helped me to make the decision.

Tried it on one lens so far. Very easy to use, and interesting to see the results.

Cheers
David Murray.

 

[Valvebounce]

Hi David.
As I mentioned, FoCal (if that is what you bought) likes lots of light, if you get inconsistent results, you probably don't have enough light, I have 3x 500w halogen worklights at about 2m from the target, 2 (1pair on a single stand) on one side from slightly above, 1 on the other side from below, plus normal workshop lights, but I think the halogens overwhelm those. I have found this gets me repeatable results with all but my Σ150-500 which is a f6.3 lens at 500mm so I add another 500w light, I'd use all 4 for the other lenses but for the heat they generate! :'( That's sweat not tears. ;D

Cheers, Graham.

Thanks guys for giving your experience.

I had been on the fence re buying software for doing AFMA, but your comments helped me to make the decision.

Tried it on one lens so far. Very easy to use, and interesting to see the results.

Cheers
David Murray.

 

[bjd]

FoCal is probably the best method, and I have used it for years, including with the 5DIII. Focal has only just updated in the last few days to include the 5DIV and so I used a simple method for AFMAing my 5DIV last week. Pin a chart to the wall - I use one from Bob Atkins site - and take 6 or 7 shots at every five AFMA numbers, and see where there is optimal resolution. Then home in with a fine analysis around the peak to tie it down. The camera was between 15-30 times the focal length of the lens from the target. It's importance to take several shots at each AFMA number as there can be variation in focus between shots. The whole procedure.

Others use a mechanical target like Lensalign. Some claim a method called "dot tune" works for them. It never did for me. FoCal does give you lots of interesting information about your lens in addition to the best AFMA, and I'll upgrade to the latest version with the 5DIV to satisfy the geek in me.

When I tried it the automatic adjustment did not work with the 5Dmk3, does it work with the 5Dmk4 now?
Cheers Brian

 

[AlanF]

Unfortunately, the "automatic" procedure is only semi-automatic for the 5DIII and now the 5DIV. They can't program the change of AFMA and so you have to dial it in yourself.

 

[drjlo]

Hi David.
As I mentioned, FoCal (if that is what you bought) likes lots of light, if you get inconsistent results, you probably don't have enough light

I use 5D III and was wondering what happens when one carefully AFMA's using optimal (lots!) lighting but when the AFMA'd camera/lens combo is actually used in non-ideal (darker) environments to auto focus and shoot photos? Are AF results going to be consistent/same as when AFMA'd in lots of light..

 

[Valvebounce]

Hi drjlo.
I think the difference is that the software is doing intense examination of the detail differences between two slightly different focus AFMA settings (to take account of manufacturing variations) after the camera locks focus whereas the focusing in low light is down to what the camera can achieve with the contrast and detail that prevail where you point it after locking in the AFMA adjustment.
My experience is that after AFMA the camera can lock focus reliably in low light.

Cheers, Graham.

Hi David.
As I mentioned, FoCal (if that is what you bought) likes lots of light, if you get inconsistent results, you probably don't have enough light

I use 5D III and was wondering what happens when one carefully AFMA's using optimal (lots!) lighting but when the AFMA'd camera/lens combo is actually used in non-ideal (darker) environments to auto focus and shoot photos? Are AF results going to be consistent/same as when AFMA'd in lots of light..

 

[scottkinfw]

Hi David.
As I mentioned, FoCal (if that is what you bought) likes lots of light, if you get inconsistent results, you probably don't have enough light

I use 5D III and was wondering what happens when one carefully AFMA's using optimal (lots!) lighting but when the AFMA'd camera/lens combo is actually used in non-ideal (darker) environments to auto focus and shoot photos? Are AF results going to be consistent/same as when AFMA'd in lots of light..

In my experience with the software over the past few years, you need lots of light-outside sunny day is great. Some have found that flickering of lights may interfere with results.
As long as you have the AFMA dialed in correctly, that will be set. If there is any degradation of focusing in the dark, it will have nothing to do with the AFMA, but rather the limitations of the camera focusing system.
Great camera, enjoy.
Scott

 

[Mt Spokane Photography]

Hi David.
As I mentioned, FoCal (if that is what you bought) likes lots of light, if you get inconsistent results, you probably don't have enough light

I use 5D III and was wondering what happens when one carefully AFMA's using optimal (lots!) lighting but when the AFMA'd camera/lens combo is actually used in non-ideal (darker) environments to auto focus and shoot photos? Are AF results going to be consistent/same as when AFMA'd in lots of light..

Auto Focus has a wide range of results, because there are so many variables. Focal sets the optimal AFMA value for the camera / lens combination using a neutral color and enough light so that it can detect tiny differences in focus accuracy.

In the real world, even the color of the subject may affect autofocus, so using Focal is a good thing, but you are still at the mercy of general autofocus variations due to many factors.

 

[takesome1]

Hi David.
As I mentioned, FoCal (if that is what you bought) likes lots of light, if you get inconsistent results, you probably don't have enough light

I use 5D III and was wondering what happens when one carefully AFMA's using optimal (lots!) lighting but when the AFMA'd camera/lens combo is actually used in non-ideal (darker) environments to auto focus and shoot photos? Are AF results going to be consistent/same as when AFMA'd in lots of light..

I find it varies.

However if you are not focusing in good light then you will likely be worse in bad light.

I have seen the differences when using the lens align in various situations and lighting.
But it has never been something I saw a need to compensate for in actual use.

 

[Refurb7]

I use Reikan FoCal. It does the job very well. Just be sure that you do it at the recommended distance for each lens.

 

[YuengLinger]

I have not tried Reikan. I know several CR members like it a lot.

But, OP, you also asked for the easiest, and I'll tell you what it is.

My results are very good. I love shallow DoF, and I have fast primes, including an 85mm 1.2L II.

All I do is set up a target (currently using SpyderLensCal), take several shots with my camera on a tripod, wide open, good light...Then I check the image on the camera's LCD, zoomed in. I wear reading glasses to see as sharply as possible. Based on the results, I adjust the AFMA in camera, try again, and have it zeroed in usually within a couple of minutes.

I do use LiveView AF first as a baseline to make sure I'm spot on with that before going to phase detect through the view finder.

I can imagine the eye-rolling, the scoffing, the claims that this isn't precise enough, but it works consistently on two bodies, a half dozen lenses, zooms and primes. When I use my lenses for portraits, focusing on an eye, I'm spot on, even pixel peeping.

So, in a nutshell, just find a good target and use the LCD display to check and make adjustments.

It has to be the simplest method. Maybe not the best down to a millimeter, but, wow, it's fast and reliable!