Testing your camera for front/back focus

BabaGanoush

Have any of you used the FoCal or FocusTune software to do a microadjustment test of the autofocus system of your camera(s) and lenses for front/back focus? If so, what did you find? How did your cameras/lenses perform according to the test results? For those of you have D800 cameras, did you test for possible problems with the outlying left-most autofocus points, and if so, were there any discrepancies versus the central focus points of the camera?

DenverShooter

I run the FoCal software pro version and when you properly set it up and carefully use it it will accurately tell you the correct amount of offset for a given lens and repeatedly give you the same result (+/- 1).

The pro version gives you charts and graphs and lots of great data. The support has been very patient with me as I figured it out and always responsive.

This is a Great product. Those who are willing to put in the effort to get it exactly right will be rewarded with perfect calibrations every time and great pictures as a result...

Denver Shooter

Symphotic

I tested with a target (a card with the text "o u c a crack" and a very fine piece of wire on it) and a ruler when some macro photographs didn't come out right. I use a D800, one of the earliest, and have no left focusing point problem. I used the AF fine tune for many months, but when had a break in shooting I sent the camera in for adjustment and it needs no fine tune with any of my lenses now.

TaoTeJared

I use Lens align. No issues with the D800 at all. The only lenses that I really tune are the 1.4's and macros. None of them were terribly off, but just slightly. Not really anything you could notice unless you are very close and shooting wide open.

From what I understand, the left focus issue was more of the sensor being tilted a bit. I was NOT a focus point not being accurate. Nikon can adjust each focus area to bring it back in focus. Usually that happens with every body but evidently it didn't happen on some. People made more of a stink of it than what the issue really was.

BabaGanoush

Let me ask two follow-up questions. Have any of you read the whitepaper by Falk Lumo, "Testing the AF Sensor Array," which can be found at:

http://www.falklumo.com/lumolabs/articles/D800Focus/SensorArray.html

If so, what do you think of the results he plotted of Image Sharpness versus Fine Tune Adjustment in Figs. 12 and 13 of that paper? Do you find them surprising, as I do? If you have used either of the two software products I asked about in my initial post, do your results look like his findings?

TaoTeJared

Call me lazy, but I really don't feel like reading a technical paper nor learning a new test system today. What I speed read through it seems you need to know the program that the person is using.
Why are you pointing to the paper?

Ok so I read though it quick - still don't get it - you need to have a good working knowledge evidently of the program to understand the desirable and undesirable parameters to really .

A couple of things I saw:
-Tested at 1.2m (4ft) which is ridiculously close for a 14-24 and 24-70.
-14-24 and 24-70 are not "planner design" lenses meaning due to the design, at close distance and wide open you will see the focal plane "curve."
-Lenses can change focus accuracy due on distance although generally it is within the focal F-stop.

Key - he set out to prove he had a left-focus issue, and he did.

I don't find that different AF points read differently at all surprising. Fine-tune really is fine tune, and taking photos of paper taped on a wall could easily have an area "bubble" up in various areas, or swing in a breeze. Even walls can be bowed out at point enough to make a mm AF difference with focus points 2 feet apart.

Since I use to manage technical write-ups, red ink them and toss them back, I'm probably more critical than others.
All I saw was the notes on every single test they spoke to limitations and how limitations effected the outcome and it probably not accurate. ON EVERY SINGLE TEST. I would have to agree - they didn't test it correctly or realistically half-assed a test, like I do, to get close. I take issue that they took their half-assed work and then threw the results in a form and then called it a "whitepaper." Don't let that term fool you, it is not truly scientific, with repeatable results. You have to slack-ass in the same way as he did to get something similar. Good luck on that.

haroldp

I use Lens align with both the camera and target on tripods. I evaluate the target image at 200% on a computer screen to determine if the focus point should move. I have not tried the software approaches to evaluation as I believe my current method is precise enough for any practical photographic purpose that I need.

Macro photography users should focus in live view using the screen.

At 10 ft. the full range of fine tune available ( + or - 20 ) with the 24-120 moves the center of focus about two inches.

I recently 'fine tuned' the 24-120 f4 VR on my d800e and found that there was focus shift of about 4 units of adjustment between f4 and f5.6. I centered the focus between them so that it back focuses by 2 units (about 1/4 in. at 10 ft.) at f4 and front focuses the same amount f 5.6.

If I did not measure it, I think I could have used this combo for real photos for the next hundred years and never notice the difference.

I also tested the left focus sensor and it is right on.

Regards ... H

BabaGanoush

@TTJ: One of the dangers of speed reading is that one might misread the article. The test results shown in Figs. 12 and 13 in that whitepaper are for a focal length of 24 mm. A testing distance of 4 feet corresponds to 50X the focal length of 24mm, which is more than adequate for a proper test. It is true that the test results can depend on the distance to the focus chart, but even 25 X FL is sufficient. At this shorter distance, the outcome might be a little different than at 50 X FL or 100 X FL, but if one takes reasonable care the broad outlines of the camera/lens behavior do not change, at least that's what I can claim based on my own experience of the past few days using FocusTune as well as FoCal, both of which I bought last week to test my own equipment.

My interest in Lumo's article was prompted less by the issue of whether there are differences in the focusing of different focus points of the D800, and more by the general trend in image sharpness with microadjustment value (AFMA). The expected behavior of the autofocus system is that of a bell-shaped curve with a relatively narrow width, i.e., for a given camera and lens one expects a maximum in image acuity within a comparatively narrow range of choices of the AFMA value. But that's not what Lumo found. Instead, according to his Fig. 12, he found a broad plateau of nearly constant image sharpness over the entire range of microadjustment values from -15 to 0 (!), followed by a gradual decline toward more positive AFMA values. His finding is surprising because it implies a very different behavior of the autofocus system than one expects.

If this observed behavior turns out to be generally true of all D800 cameras, doesn't it imply that whatever microadjustment one dials in, as long as it's a negative one, or unless fine tuning in the camera is turned off, there is no difference in the image sharpness one can get, i.e., image sharpness is largely independent of the microadjustment setting? What does that say about the complaints people have written about on-line about having to dial in large microadjustment values to get their cameras to focus sharply? Were their complaints actually justified or were those people being led astray? And just why is it that Nikon's autofocus system/s show/s such a departure from the expected behavior in the AFMA tests? Do the tests suggest that Nikon's autofocus system ignores the microadjustment setting---which seems unlikely---or does the AF system do something more than just read the AFMA setting to achieve the sharpest focus?

Msmoto

I guess I am just too old to get into the technical issues. I would like to have someone tell me exactly how to get the camera to work the best. I have fine tuned with the AF Fine Tune, my longest lens combo, but have not done this with the others as they seem to be doing what they are supposed to. I will leave it to you experts to get this all figured out and then please just tell me what to do. Someone did this for the AF Fine Tune and it worked for me.

BabaGanoush

@Msmoto: One option is to do nothing, because it'd be a waste of time. Now that I've tested my D800 with my 70-200mm f/2.8 VR II lens, I'm turning camera fine tuning off. The results of my tests with FocusTune are very similar to those of Falk Lumo on a 24-70mm lens, namely, image sharpness is independent of the autofocus microadjustment value over a very wide range of this setting. As long as I avoid a large positive value of the AF setting, images of focus test charts as well as real-life photographs appear equally sharp. I'm about to test my 24-120mm f/4 lens. I'll be interested to see whether the results of the test once again indicate that fine tuning can be turned off.

Msmoto

@ BabaGanoush

This is exactly what I found with a couple lenses. Even the 400mm f/2.8 with the TC-20EIII at 60 feet was very difficult to distinguish the differences in fine tune. I did conclude a -6 worked best, but the differences were extremely difficult to see.

However, once I used this for the long combination, in my final photos, I believe I see a subtle difference, improvement in sharpness.

rmp

I must be lazier than anyon on Nikon Rummors because, I just want to find someone that will take my camera and lenses and adjust the so I do not need to adjust them at all. Does anyone know of such a calibration/adjustment service?

warprints

Send them to BabaGanoush !! Just kidding.

Msmoto

@ rmp

I can do this RMP....LOL Actually, the process is time consuming, at least for me, but straight forward. As for lazy...I did not do this until I started imaging the birds in my yard with the 800mm set up. Then I noticed a very slight back focus and after about an hour or so of fiddling and shooting test images on my chart. I settled for -6. I am certain someone in your area can help you with this. More LOL )

TaoTeJared

@BabaGanoush - I think taking one test of ONE D800 and assume it is the same for all is not smart. If the author did the test on 100 D800s then I might take note. Every camera has a slight difference from another one with focusing.

BabaGanoush

@TTJ: What you say is obvious and goes without saying...What I did in my earlier posts was to raise a possibility that I don't recall ever seeing mentioned in the many plaintive on-line complaints about D800 focusing. I guess the point was presented too subtly since you failed to see it. I'll be "smarter" next time and recognize that some readers may not catch on so quickly if the writing gets too terse.

Msmoto

The primary issue on this thread seems to be more of interest to the technical oriented folks who may be a bit more obsessive in the fine tuning of equipment, exposure settings, etc.

Then there are those of us who are more pragmatic in our approach and want to do the minimal. This does not mean shortcutting the process, but it does mean in some cases using the limits of the camera.

I always admire those who can read the technical articles and understand them. For me, my focus chart is home-made, duct taped to a light stand, and used in my garage which is long enough to configure about any lens I choose.

But, these hi-tech things are interesting.