Unable to AFMA 70-200mm 2.8 is II

[gjones5252]

So i am so excited that i finally got this lens. Sad part is no matter what i do i cannot get it into focus. It always seems to be just a bit off. On closer inspection i noticed i had it at +20 the highest AFMA available and it cannot get accurate focus. It is just a slight bit front focused. I feel like for a 2500 lens I would want it to be acurate. As i am typing this i am wondering if its on when its at 70. The sample images are all cropped and shot at 200. Should i send it and the camera to canon? I have acquired what i consider to be  a set of bodies and lenses i am content with(minus 5d mark iii) and i have heard you can send all you equipment to canon and have it all calibrated to each other. If so what does it cost? Or am i being to picky?

[Don Haines]

Where do you think it was supposed to focus.... and what are all the focus settings? Could you repost with the settings and draw a circle around where you think it should have focused?

Also, one of my co-workers brought in his Rebel and complained that it would not focus..... turns out he had it set to spot focus and had selected the far left focus point, yet thought that the camera should focus in the center. it never hurts to verify the settings.....

[neuroanatomist]

Could you repost with the settings and draw a circle around where you think it should have focused?

Or better yet, open it in DPP, turn on the AF point display, and take a screenshot. 

One more test - compare live view AF with viewfinder AF - if the former is sharper, it's likely an AFMA issue, and if both are front-focused it's likely a lens quality issue.

Note that your test scene is not really ideal, either.  If you have a few lenses, I really think Reikan FoCal is the way to go for doing AFMA.

[PackLight]

Shooting at rocks on the ground isn't going to tell you how accurate it is.

For a caveman style AF test I would do the following at a minimum.

At the very least you need a definite isolated subject that you know your camera focuses on. Next a yard stick or something similar laying next to your target so you can actualy see the DOF that you get.

Settings use single point  AF in single shot mode with no surrounding points engaged.

Take a few shots and review. If it looks close but not perfect go to the next step and set up a more scientific test. Or buy focal and give it a try.

[neuroanatomist]

At the very least you need a definite isolated subject that you know your camera focuses on. Next a yard stick or something similar laying next to your target so you can actualy see the DOF that you get.

Yep - that was part of my point.  Here's what I recommend as a DIY setup, requires a cardboard box, a tape measure, a chopstick or pencil, and a printout of this focus target (use a tripod, but you could just set the camera on a book, as pictured). 

[dirtcastle]

Nice setup.

I just bought FoCal. It seems to get good results. Hopefully they will improve support for the 5D2 and 5D3. Manual mode is tolerable, but I sometimes felt like fidgeting with the controls and keyboard was undermining the delicate nature of the testing. A few times it told me the target had moved. Um, well... yes, because I'm hitting effing buttons during the test. 

[Don Haines]

Don't forget that the viewfinder has a diopter adjustment......

[PackLight]

Nice setup.

I just bought FoCal. It seems to get good results. Hopefully they will improve support for the 5D2 and 5D3. Manual mode is tolerable, but I sometimes felt like fidgeting with the controls and keyboard was undermining the delicate nature of the testing. A few times it told me the target had moved. Um, well... yes, because I'm hitting effing buttons during the test. 

Try reading a book or watching TV while the Auto test is running. You might get better results.   

[Mt Spokane Photography]

Don't forget that the viewfinder has a diopter adjustment......

The diopter adjustment has nothing to do with the camera AF.  It does help you see the focus screen sharply, but it does not matter where it's set as far as camera AF is concerned.

[dirtcastle]

Nice setup.

I just bought FoCal. It seems to get good results. Hopefully they will improve support for the 5D2 and 5D3. Manual mode is tolerable, but I sometimes felt like fidgeting with the controls and keyboard was undermining the delicate nature of the testing. A few times it told me the target had moved. Um, well... yes, because I'm hitting effing buttons during the test. 

Try reading a book or watching TV while the Auto test is running. You might get better results.   

That would be nice!

Alas... my 5Ds only work in manual mode. But still, I'll take it; it's a lot easier than the setup Neuro diagrammed.

[PackLight]

Nice setup.

I just bought FoCal. It seems to get good results. Hopefully they will improve support for the 5D2 and 5D3. Manual mode is tolerable, but I sometimes felt like fidgeting with the controls and keyboard was undermining the delicate nature of the testing. A few times it told me the target had moved. Um, well... yes, because I'm hitting effing buttons during the test. 

Try reading a book or watching TV while the Auto test is running. You might get better results.   

That would be nice!

Alas... my 5Ds only work in manual mode. But still, I'll take it because it's still a lot easier than what Neuro outlined.

Yes it easier than the old stick and stone methods.

I don't have the 5D III (I read it has to be done in manual) but my 5D II works in Auto fine, but mine is finicky. What I found is a couple of things, if it is not finishing the run it could be one of a few things. Distance to target, sometimes if it doesn't finish the cycle adjusting a bit forward or back seemed to help. If the camera isn't square to the target it will have problems. Lighting, if you get a lower EV sometimes it has problems. The only lens I had that I couldn't get to run in Auto was the 300mm f/2.8 but I was trying to do it at 20x focal length distance away.

[dirtcastle]

Nice setup.

I just bought FoCal. It seems to get good results. Hopefully they will improve support for the 5D2 and 5D3. Manual mode is tolerable, but I sometimes felt like fidgeting with the controls and keyboard was undermining the delicate nature of the testing. A few times it told me the target had moved. Um, well... yes, because I'm hitting effing buttons during the test. 

Try reading a book or watching TV while the Auto test is running. You might get better results.   

That would be nice!

Alas... my 5Ds only work in manual mode. But still, I'll take it because it's still a lot easier than what Neuro outlined.

Yes it easier than the old stick and stone methods.

I don't have the 5D III (I read it has to be done in manual) but my 5D II works in Auto fine, but mine is finicky. What I found is a couple of things, if it is not finishing the run it could be one of a few things. Distance to target, sometimes if it doesn't finish the cycle adjusting a bit forward or back seemed to help. If the camera isn't square to the target it will have problems. Lighting, if you get a lower EV sometimes it has problems. The only lens I had that I couldn't get to run in Auto was the 300mm f/2.8 but I was trying to do it at 20x focal length distance away.

I was confused when my 5D2 didn't work with auto mode. I'll give those tricks a try. Thanks!

[scottkinfw]

You are not too picky.

I have similar problems with my 5D3 and my L lenses.

I sent my lenses, along with my 5d2, 5D3, and 50D to get calibrated.  I got the equipment back from Canon, and was very disappointed.  To make it worse, I did this prior to an African Safari, and got way more duds than keepers.

If you send them in, be very careful on checking out the gear before you take it out on an important shoot.

sek

[scottkinfw]

Here are a couple of tips.

Be sure tripod is very stable.  Lock all of the rotation points.  I also use masking tape or Frog Tape to further secure against movement.  This really helps.

sek

Nice setup.

I just bought FoCal. It seems to get good results. Hopefully they will improve support for the 5D2 and 5D3. Manual mode is tolerable, but I sometimes felt like fidgeting with the controls and keyboard was undermining the delicate nature of the testing. A few times it told me the target had moved. Um, well... yes, because I'm hitting effing buttons during the test. 

[Don Haines]

Don't forget that the viewfinder has a diopter adjustment......

The diopter adjustment has nothing to do with the camera AF.  It does help you see the focus screen sharply, but it does not matter where it's set as far as camera AF is concerned.

Agreed, but I have run into people who do not know about it and wonder why the viewfinder is never sharp..... There are a bazillion settings on a camera and many people only know what a few of them do. Quite often, when things go wrong, it is a settings problem.... This focusing problem could be lens, could be body, or the camera could be set to auto select the focus point, it picks a point other than the center, yet the user mistakenly believes it is still set to focus on center point only..... That's why I asked what the settings were....