New 24-70 II, maybe AFMA will help?

[Shawn L]

Just got a new 24-70 II and I must say that my initial images are, well, disappointing. I was hoping for something along the lines of the ridiculous sharpness of the 70-200 II. Sadly, what I'm seeing is much closer to my 28-135.

If you look at this .cr2 file in DPP (http://home.comcast.net/~sal6/BY5R6357.CR2), you'll see the focus point is the white and purple cluster of flowers. However none of the flowers in that cluster are what I would consider remotely sharp -- this was taken at 70 mm, 1/320 sec, with f/2.8, ISO 100. With those, I don't think it's the lack of IS.

All of the photos I've taken are like this, without regard to shutter speed (even up through 1/2000 @ f/2.8 (ISO 150)) -- also,  this is actually one of the sharper ones.

Maybe AFMA will help. But I'd (naively?) expect an AFMA problem would mean that something, somewhere in the image would be sharper. I suppose it's possible that there's nothing at the right depth...

Thoughts?

Shawn L.

[wopbv4]

Hi ,

looks to me as serious case of front focusing.
I believe that you took this picture with the camera angled down.

Anyway, the leaves at the bottom of the picture (5 o'clock) are sharper then leaves behind it. Also purple flowers at 7 o'clock look a bit sharper then flowers at the centre. All this points to front focusing.

Please do a simple test. Put a ruler at a 45 degree angle and focus on the middle.  This should clearly show if you have a front focusing problem.

Hope this helps

Ben

[Shawn L]

Good idea. Thank you, Ben.

Shawn L.

[neuroanatomist]

Please do a simple test. Put a ruler at a 45 degree angle and focus on the middle.  This should clearly show if you have a front focusing problem.

If I had a nickel for every time I see this advice...

Focusing on an angled ruler is not effective.  Say you put the AF point over the 10 cm mark on the ruler.  You know you're aiming for that line, but the camera doesn't, it's going to lock onto the region of highest contrast under the selected AF point, and that AF point is actually larger than the little box that represents it in the VF.  So, most likely it will be locking onto the edge of the ruler which is passing through the AF point, somewhere along that edge, not where you want it to focus.

There's a reason that the commercial AFMA tools have a focus target that's oriented parallel to the image sensor, with an offset angled ruler to judge the DoF, but you aren't focusing on the angled ruler.

Actually, there's a much simpler test to see if you need AFMA.  Select a flat target with high contrast and lots of detail, oriented parallel to the sensor.  Go into Live View and use contrast detect AF (called Live AF, do not use Quick AF - the mirror should not flip during focusing).  Take a couple of images that way.  Then, turn off Live View and take a couple of shots with phase AF.  If the Live View shots are sharp and the phase AF shots are less sharp, AFMA will help.  If the Live View AF shots are not sharp, it's more likely a lens issue.

[Random Orbits]

Please do a simple test. Put a ruler at a 45 degree angle and focus on the middle.  This should clearly show if you have a front focusing problem.

If I had a nickel for every time I see this advice...

Focusing on an angled ruler is not effective.  Say you put the AF point over the 10 cm mark on the ruler.  You know you're aiming for that line, but the camera doesn't, it's going to lock onto the region of highest contrast under the selected AF point, and that AF point is actually larger than the little box that represents it in the VF.  So, most likely it will be locking onto the edge of the ruler which is passing through the AF point, somewhere along that edge, not where you want it to focus.

There's a reason that the commercial AFMA tools have a focus target that's oriented parallel to the image sensor, with an offset angled ruler to judge the DoF, but you aren't focusing on the angled ruler.

Actually, there's a much simpler test to see if you need AFMA.  Select a flat target with high contrast and lots of detail, oriented parallel to the sensor.  Go into Live View and use contrast detect AF (called Live AF, do not use Quick AF - the mirror should not flip during focusing).  Take a couple of images that way.  Then, turn off Live View and take a couple of shots with phase AF.  If the Live View shots are sharp and the phase AF shots are less sharp, AFMA will help.  If the Live View AF shots are not sharp, it's more likely a lens issue.

+1.  Stop signs work well and they're plentiful.  Also make sure you're using a high accuracy focus point (usually the center).

[Dylan777]

Please do a simple test. Put a ruler at a 45 degree angle and focus on the middle.  This should clearly show if you have a front focusing problem.

If I had a nickel for every time I see this advice...

Focusing on an angled ruler is not effective.  Say you put the AF point over the 10 cm mark on the ruler.  You know you're aiming for that line, but the camera doesn't, it's going to lock onto the region of highest contrast under the selected AF point, and that AF point is actually larger than the little box that represents it in the VF.  So, most likely it will be locking onto the edge of the ruler which is passing through the AF point, somewhere along that edge, not where you want it to focus.

There's a reason that the commercial AFMA tools have a focus target that's oriented parallel to the image sensor, with an offset angled ruler to judge the DoF, but you aren't focusing on the angled ruler.

Actually, there's a much simpler test to see if you need AFMA.  Select a flat target with high contrast and lots of detail, oriented parallel to the sensor.  Go into Live View and use contrast detect AF (called Live AF, do not use Quick AF - the mirror should not flip during focusing).  Take a couple of images that way.  Then, turn off Live View and take a couple of shots with phase AF.  If the Live View shots are sharp and the phase AF shots are less sharp, AFMA will help.  If the Live View AF shots are not sharp, it's more likely a lens issue.

+1....someone has mentioned this method few months ago. I tried and  yes..I can see the diffirent on my 16-35 II - AFMA or not AFMA.

[Shawn L]

Ah, thank you, Neuroanatomist et al.

I did a quick test before work with a yardstick tilted at what I hoped was 45 degrees. The results were a bit inconclusive as nothing felt any sharper than my intended focus area, and that still felt soft.

I'll try the other test tonight or tomorrow.

Cheers.

Shawn L.

[neuroanatomist]

I did a quick test before work with a yardstick tilted at what I hoped was 45 degrees. The results were a bit inconclusive as nothing felt any sharper than my intended focus area, and that still felt soft.

If you're going to set up a yardstick test, a setup like this (using a focus target such as this one) is the best way to do it.  This is basically a DIY version of the commercial AFMA tools. 

[Shawn L]

Okay, so mine was nowhere near sufficient, then

Shawn L.

[K-amps]

I tried a 24-70ii, ran some tests and the good folks on this site help me understand it was not worth $2300  I returned it.

I might wait till there are more copies to choose from but it seems my copy was not the best out there.

If you are not blown away by a $2300 lens; then return it and get something that does.

[nightbreath]

Actually, there's a much simpler test to see if you need AFMA.  Select a flat target with high contrast and lots of detail, oriented parallel to the sensor.  Go into Live View and use contrast detect AF (called Live AF, do not use Quick AF - the mirror should not flip during focusing).  Take a couple of images that way.  Then, turn off Live View and take a couple of shots with phase AF.  If the Live View shots are sharp and the phase AF shots are less sharp, AFMA will help.  If the Live View AF shots are not sharp, it's more likely a lens issue.

Live View magnification can also help. When I do AFMA for primes I follow next workflow:
1. Make sure camera is set on a sturdy tripod.
2. Use Live View with 10x magnification to focus on the target (contrast detect).
3. Move focus ring to minimum focusing distance, focus using regular focusing (phase detect), make a shot.
4. Move focus ring to maximum focusing distance, focus using regular focusing (phase detect), make a shot.
5. Compare, make adjustments, check again.

The only issue I'm having is that the difference between AFMA values (e.g. "-1" and "0") is twice more than I'd like it to be. In my testing I see the same focus shift for two neighboring values, while contrast detect hits somewhere in the middle (i.e., at "-0.5").

[Old Shooter]

Shawn - I'm not trying to spend your money - but you are shooting a 1DX with a 24-70II... If you like shooting wide open - I would seriously recommend FoCal - it made a huge difference with my 5DIII... If you buy it - after you AFMA - run the aperture sharpness test at both ends... I was really surprised by the results... Cheers!

[Mt Spokane Photography]

Shawn - I'm not trying to spend your money - but you are shooting a 1DX with a 24-70II... If you like shooting wide open - I would seriously recommend FoCal - it made a huge difference with my 5DIII... If you buy it - after you AFMA - run the aperture sharpness test at both ends... I was really surprised by the results... Cheers!

Yes, first thing to do with a new lens (after you snap a few images just to see it work) is to run a AFMA and FoCal is the best tool, its reliable and accurate (if you read the instructions and follow them)
Then you will know that your lens is focusing accurately.
The other quick and dirty method is to put it on a tripod and use AF to snap a image of a flat target that is parallel to the sensor.  Then use liveview and live autofocus to do the same.  If the live AF image is sharper than the first image, AFMA is needed.
Be aware that its sometimes very difficult to visually detect tiny differences in sharpness, but Focal can detect them, which is why so many use it.
 

[Shawn L]

Old Shooter & Mt Spokane:

I have been looking at getting FoCal, my hold up has been that I don't own a tripod and I'm not sure I need one other than for this. Perhaps, it's time to rent one, though

Thanks, everyone.

Cheers.

Shawn L.

[Mt Spokane Photography]

Old Shooter & Mt Spokane:

I have been looking at getting FoCal, my hold up has been that I don't own a tripod and I'm not sure I need one other than for this. Perhaps, it's time to rent one, though

Thanks, everyone.

Cheers.

Shawn L.

The purpose of using a tripod is merely to remove the possibility of camera movement as well as making sure that all the test shots are taken from the same place and the same angle.
You might be able to construct something using a 1/4-20 bolt to attach the camera, but it might be more trouble than its worth.
If you can borrow a good tripod and head, that might be ideal.  Sometimes you can find one locally from Craigslist.  If you get a cheap one, hang a weight under the center to stabilize it, a camera bag full of lenses works well.  You can usually screw a hook into the bottom of the center column.