[AggieDad]

I know there are standard angled focus adjustment charts, but at there moment I do not have one – and won't for a while as AggieMom just approved a new monopod and tilthead.

I wanted to check the focus of my Sigma 150-500 mounted on a K-3ii in preparation to heading out to see the whooping cranes on the Texas coast, so I had to think of a way to do it.

My thinking was that if I tested every focus adjustment from -10 through +10, I really wouldn't need to know if the lens front or back focused. I could just pick the best setting. So I set out to do this.

I used the license on my car (the Texas license has nice, clean black letters on a white background) and set my camera up 65 feet away on the tripod. I centered the camera image on the center of the license. I used mirror up and reset the the lens focus to about 3 feet between every shot so the lens focused each time. All shake reduction (camera and lens) was off. Photos were taken at 500mm at f/6.3, 1/800th, ISO 200.

Looking at the resulting images at 100%, 200%, and 300%, I could see only a small amount of difference between the 21 images, and from -4 to +4 there was little, if any, difference.

Was my original thinking valid? If so, would these results be normal (i.e. so many settings with so little difference)? If not, what would be a good test?

Many thanks,

Don

[aslyfox]

pm sent

[CarlosU]

Two things to consider:

- The closer your target is, the less depth of field you have so the differences between AF adjust settings become more evident. You shot at 65 feet, I would try at least half of that distance.

- There is some AF variability between shots (at the same AF adjust setting), so to get reliable info from your test you should do various shots (from 6 to 10 maybe?) at each AF adjust setting. Even more, I would do half of them with the lens coming from infinity focus and half of them from closer than the target.

[pschlute]

The flaw with this technique is that you are making a subjective analysis of the image because it has no depth. The point about using a focus chart is that you can easily see front or back focus without making a subjective judgement.

Print out the chart from the end of this article http://www.kscameraclub.org/docs/pdfs/focus_test_chart_edited.pdf

Use natural light if possible and widest possible aperture. Centre AF point only. CarlosU's point about doing 10 shots at each setting is spot on. You will find doing this testing that your camera autofocus is not as repeatedly accurate as you thought it was !

[DeadJohn]

If you're adjusting without a readily available chart, a license plate could work. Setup your tripod so you are at a 30 to 45 degree angle to the plate to see back- and front-focus. Some plates are printed flat while others are stamped with the letters protruding; flat would be better for testing.

Laptop monitors make good focusing targets IMO. Positioned the screen at 30 to 45 degrees relative to the camera. When you zoom in you can see individual pixels to determine whether the camera is focusing to close or too far.

[DW58]

License plates look pretty crisp but they're really not. I won't swear by this but I think it may help giving you a starting point on distance between camera and target. Click on the link and go to distance tool. When you're ready take a shot with the camera, on a tripod, AF fine tune at zero. Then plus 1, 2, 3. Then minus 1, 2, 3. Look at those images on your computer at 100%. You'll be able to tell if any are looking sharper than when set at zero. If plus or minus 3 looks best shoot some more adjusting to to 4, 5, 6. Etc. Plenty of printable targets available on line for free. Or you can spend a lot of money on one of the kits available.

Edit: Forgot to mention to keep camera and target on the same plane. More light the better. ISO 100 is best. Widest aperture. If you don't have a remote the 2 second delay is your next best bet. Adams post below makes a lot of sense.

http://michaeltapesdesign.com/lensalign.html

[Adam]

If you aren't seeing any issues in real-world photos, there is no need to apply adjustments IMO.

Otherwise, as others have noted, decrease your working distance to be as close to the minimum as possible, shoot wide-open at the focal length you most often use, and minimally use a printable focus chart.

[Jeff Lopez]

You should grey card your camera before anything..AF depends on proper contrast and ambient lighting..

[Photobill]

To Jeff Lapez
Can you expand on "Grey Card" your camera I have a K3II and have not heard of this?
Thanks

[Alex645]

set my camera up 65 feet away on the tripod. I used mirror up and reset the the lens focus to about 3 feet between every shot so the lens focused each time. All shake reduction (camera and lens) was off. Photos were taken at 500mm at f/6.3, 1/800th, ISO 200.

Was my original thinking valid? If so, would these results be normal (i.e. so many settings with so little difference)? If not, what would be a good test?

Just to add to the feedback:
If your test target was closer, youʻd see a bigger difference because of reduced depth of field, but ultimately, Iʻd put the test target at an average distance your subject is anticipated in your use of this lens. So if 65 feet is in that range, that makes sense.

Yes, SR off, wide open f/6.3, but 1/800" is slower then Iʻd want to do tests with a 500mm lens. Iʻd try 1/1500" at ISO 400 to safely eliminate blur out of the results.

With ʻmirror upʻ were you AF with live view or your OVF? With live view, AF Fine tuning is not needed. With your OVF, you should be seeing a difference. If you canʻt see a difference between your -10, N, and +10, then that tells me either you were in live view, or your test target doesnʻt have enough detail to show and test the differences.

Instead of the license plate, can you tape printed matter to a fence and prop a yard or meter stick (or a PVC pipe or broom handle with incremental marks in front and behind the plane of the fence) at a 45 degree angle from the sensor/film plane?

And of course, it is possible your AF fine tune needs no correction with this lens, but there is a flaw in your testing method if you see no difference between -10 and +10.

[AggieDad]

Many thanks to everyone for your thoughts and ideas. I will be working on this as soon as we return from a short road trip to see whooping cranes.

It is not a critical issue (it was on the K-5iis), so I can take my time and try everyone's suggestions. It makes a good at home activity.

I'll repost when I have new results along with what works and doesn't.