24-70 AF all over the place!

JJPhotos

I think my 24-70 will be making a trip to Nikon UK as I've just run a few tests in FoCal and it's -1 @70mm, -8 @50mm and +17 @24mm.

Before I put it in the post however, can anyone on here think of a reason that there is such a wide gap in the numbers? I mean -1 and -8 I understand and can work with in a zoom lens but going from -8 to +17 is a bit much.

Cheers,

Jamie

TaoTeJared

All zooms are like that. It's normal and that is part of the reason you never see a zoom below 2.8. Ideally the focus would be a bit more off on the wide end (as that is the most DOF) and closer on the long end. The middle range of the lens should be a bit better than half way between the two - just as your lens is.

I think the better question is how much does each -/+ # actually equal? 2nd question is, how much DOF do you have at each focal length? 3rd question is, does it make any difference?

24mm, f2.8 @ 10 feet = 11 feet of focus depth and 5 feet is about 28".
50mm, f2.8 @ 10 feet is like 2 feet of focus depth and 5 feet is about 6".
70mm, f2.8 @ 10 feet is about 1 foot of focus depth and 5 feet is about 3"

.

I'm guessing the auto fine tune only will works with the middle 50% of DOF. At 24mm (f2.8 @ 10 feet = 11 feet ) that is 5 feet. Then if you say there are 40 steps to tune focus, that is about 1.5 inches per step. At +17, that is about 2 feet. If you figure the DOF for any focal point is 1/3 in front (approx 3'9"), 2/3 in back(7'6"), you are well within the DOF.

Msmoto

Just a note...the 400mm f/2.8 VRII with the TC-20Eiii at about 60 feet, each number + or - of AF Fine Tune is maybe 1-2 cm. I found it very difficult to do this in spite of using an actual scale in inches at 45° to the lens axis. However, once I had dialed in a -6, the lens seemed to be right on when focused at about 15 feet on my bird feeder.
So, the compromise in a zoom is to figure where you want it the closest, most often at the long end as TTJ has said, and do you calculations there. If you do not like the results in the end, notch it back or forth as you prefer. It took me about one and a half hours to do this, redoing, dozens of exposures, etc. I did use a simple procedure of simply using the card from the camera and looking at the images on the card, never actually downloading to the computer.

TaoTeJared

Sorry MsMoto - I was editing while you posted - I did do the calculations and re-wrote it.

JJPhotos

Thanks guys. I'll persevere with the lens. Amusingly (and I'm prepared to acknowledge that my method may be faulty though I don't think it is) on the D7000, FoCal suggests using -16 @ 70mm with the same lens!

BabaGanoush

I started a thread on the same subject on January 29, "Testing your camera for front/back focus," which is currently on the next Web page, so you might scroll down and take a look at the various posts there. I was testing my 70-200mm f/2.8 VR II lens on a D7000 and a D800. I was using Michael Tapes's new software, FocusTune, but I ran the same test images through FoCal in a couple of cases and got the same answers as with FocusTune, so the two analyses are consistent to that extent.

My focus tests for both cameras showed a wide range in the formal results for the AF microadjustment value corresponding to peak sharpness. For the D7000, peak sharpness at FL=70mm occurred at an autofocus setting of -16, increasing monotonically to -4 at FL=200mm. For the same lens on my D800, the numbers were -16 at FL=70mm, increasing monotonically to +4 at FL=200mm. At FL < 150mm, peak sharpness on the D800 fell at negative (minus) microadjustment values, at FL > 150 mm the peak was at positive values, and at FL = 150mm the Sharpness versus AF Microadjustment curve was literally flat over the entire expanse from -14 to +4. More recently I've tested my 24-120mm lens on the D800. The range in the microadjustment settings for maximum sharpness was a lot narrower, -6 at FL=24mm, increasing monotonically to 0 at at FL=120mm. As a compromise, I've set the default AF microadjustment value for the D800 to -3. I'm also using -3 for the D7000, but that may change after I've tested two other zoom lenses I use on that camera. The reason I'm using a fixed microadjustment setting of -3 is that I've found there's not much difference in the actual image quality at the different microadjustment settings. Despite the wide range in the formal test results for the 70-200mm lens, for example, I see very little difference in the image sharpness at an AF setting of -16 versus one of -4.

Turning the fine tuning completely off is another option. I don't see any appreciable difference in the sharpness of the test images I've taken with fine tuning turned off versus those I've taken at an AF setting of -3. I plan to do more focus tests in the next few weeks and then, after the weather clears, I'll try to confirm the results with some real-world photographs. At that point I'll make a final choice.

DJBee49

Very interesting.

I have tested all my lenses using FocusTune and have stored the results on my D800. I must admit that I have not done any comparative 'with and without' tests though and cannot say that I have noticed any difference. I went through this just for peace of mind really; it is just one more variable to get rid of. It is beginning to sound as if the increments are so small as to be virtually un-noticeable. Having tested mine, it was reassuring to have confirmation that none of my lenses were wildly out of acceptable tolerances though.

Has anyone been through all this and really noticed a significant change in sharpness? I do not mean here those who have been unfortunate enough to have bodies with significant alignment faults but with reference to lenses on 'normal' bodies whose corrections are within the range of adjustments on AF Fine Tune.

Paperman

I only tried once to fine tune a lens - my 50mm f1.8D . Though it is a prime, after 30-40 shots and going from maybe -15 to +15 for each shot, I gave up as every time I shot , I got a different result. I just decided that the AF systems are not "that" perfect and not worry about it as it did not affect my photography - the DOF being enough to cover the lens focus error. Maybe it is because it is a cheap lens.

I just wonder if any of you have gone back to repeat the tests once you thought you had the ideal calibration. I have a feeling you may be getting different results every time you do the tests.

DJBee49

Strange! My 50mm f1.8D was also the only one that needed significant adjustment (-10). I was a little surprised at this as I always thought the lens to be very sharp, even wide open. I have saved this value on the camera and it seems fine- but then it did before too! I did repeat the tests on a few of mine and they were pretty consistent.

I was slightly puzzled by the spread of focus points when taking the samples for one set up though but not having any data on the actual difference between shots it is not possible to say how significant these might be. I suspect though that the error in distance from shot to shot on the FocusTune graph is very small and we may be worrying too much!