FoCal or USAF resolution chart for AF fine tuning

If you would like to post, you'll need to register. Note that if you have a BCG store account, you'll need a new, separate account here (we keep the two sites separate for security purposes).

alanlwilder

Member
Supporting Member
Marketplace
After reading your e-book "Secrets To The Nikon Autofocus System-Mirrorless Edition", I purchased FoCal to best tweak AF on my 600 AFS II with my D850. Prior to FoCal, I used an old USAF resolution chart to calibrate AF fine tuning at f/4 obtaining a value of -6 as seen here:
-6 AF FT.jpg
You can only see EXIF info for this image if you are logged in.

Using the FoCal AF calibration, it suggested +2 as the best value as seen here:
+2 AF FT.jpg
You can only see EXIF info for this image if you are logged in.

As you suggested it's best to check the settings and attached are my results. Picking the SKU labels on the transformer as my viewfinder's central sensor subject, the USAF chart gives me a better result when focused on the labels where the numbers and SKU lines are better focused at the -6 setting. If you look at the wooden pole supporting the transformer, the + 2 setting is better focused as expected because it is behind the transformer. My question is this, do I use the +2 setting or the -6? I didn't know if results from linear resolution in a USAF chart is equivalent to the sharpness needed in shooting the eye of birds or wildlife for tack sharp results. The sharpness difference between the -6 and +2 is a little more noticeable on my computer with LR that what I posted here.
 
Last edited:
My question is this, do I use the -2 setting or the +6? I didn't know if results from linear resolution in a USAF chart is equivalent to the sharpness needed in shooting the eye of birds or wildlife for tack sharp results. The sharpness difference between the -2 and +6 is a little more noticeable on my computer with LR that what I posted here.
Gotta say, neither of those look to be in sharp focus assuming you were focusing on the transformer labels. Did you shoot these at a sufficiently high shutter speed and in bright enough light to keep ISO low?

There's lots of methods for performing AF Fine Tune but personally I get my most consistent and repeatable results using the Dot-Tune method on a well lit and flat target. The USAF target is fine but the key is to shoot it in bright light at representative distance with the lens well supported but still using a high shutter speed as in 1/1000" or even faster with a 600mm lens.

Here's a good link to the Dot-Tune method:

And like any testing, it's best to repeat the test several to many times to check for consistent results before changing your camera settings.
 
I too think there is something up with that lens. It not just missed focus, but it looks soft overall. Normally when you're front / back focused, there something tack sharp in the image - even if it's not what you intended. This looks like something is wrong optically to me. I'd test using Live View AF just to see if you can get a tack sharp result before going forward. If you can't, then no amount of fine tuning will help.
 
The ISO was 800 and shutter was 1/400 so with a 600/4, more light would definitely help. The image shown was moderately cropped as the target was about 300 feet away. However, despite the noise, -6 was definitely sharper to the point where the SKU bars and small #s can just be read at 300% in LR. At +2, it was a little soft to be certain.
 
The ISO was 800 and shutter was 1/400 so with a 600/4, more light would definitely help. The image shown was moderately cropped as the target was about 300 feet away. However, despite the noise, -6 was definitely sharper to the point where the SKU bars and small #s can just be read at 300% in LR. At +2, it was a little soft to be certain.
More light and faster shutter speed will help but it sounds like you're shooting from quite far away as well. There are tools like this one that indicate good distances for focus fine tune testing based on sensor size, lens focal length and target size. Somewhere around 40 to 60 feet is a more typical testing distance as shown by this calculator: https://support.fo-cal.co.uk/article/27-af-calibration-target-distance-tool

Besides shooting through a lot of atmosphere at 300', that distance also makes critical determination of best focus difficult because a 600mm f/4 lens has roughly 6 feet of Depth of Field at 300 feet. Shooting closer not only removes a lot of atmospheric effects, it also lets you work with very slim DoF to make critical focusing differences easier to see.
 
Last edited:
For what it's worth, FoCal was unusable for me outside. Results were very inconsistent, even minute-to-minute with a sturdy tripod, good light, short focal lengths, and minimal wind.

I switched to a 3D target (LensAlign) and manually dialed in my preferred AF Fine Tune values after visually analyzing the images at different values. I'm much happier with my results now.
 
I agree with DR - DOF makes your test irrelevant. That may reflect how you use the lens, but it's virtually impossible to see a focus error and direction.

Start with a nearer target to establish a baseline - maybe 100 feet. FoCal is quite variable when the target is too far away and essentially gives up or throws in bad data. Use Live View to establish you and the lens can produce a sharp image. Live View uses Contrast Detect AF - meaning it measures focus adjustments based on image contrast. Be sure your technique is excellent with all that entails to prevent shutter shock and any kind of mirror slap. After you have confirmed you can get a sharp image with LiveView, then start testing with 0, +5, and -5 adjustments or something similar. You should have a clear best among those tests and it should be highly repeatable. Then refine as needed. Shoot at least 10 frames with each setting defocusing in between.
 
Shooting a USAF chart at about 100' distance (51 x 600mm) the AFS II resolved 88 lp/mm centrally with this lens. When I use to own the 600E FL, resolution was about the same but contrast was higher. Looking at the on axis Nikon's MTF curve frequency @ 30 l/mm, the 600E FL was at least 98%, while the AFS II was around 96%.
 
It's quit simple to do micro focus adjustment on the Nikon D850, it will do it almost by itself.
https://nps.nikonimaging.com/technical_solutions/d850_tips/useful/auto_af_fine-tuning/
There is no need for a fancy target, you should be able to use the labels shown as they have enough contrast. I have used largest aperture available on the lens and set distance to the target 20-50 times focal length, with 600 mm that gives 12,000-30,000 mm=> 12-30 meters (47'-118'). Has given rather fine results.
 
When I used the D850 Live View and engaged focus peaking in manual focus, it shows sharpest possible focus. I then adjusted the AF fine tune so that it settled on the exact same focus setting and also confirmed resolution/sharpness in LR. The nice thing about a USAF resolution chart is that you can quantitativley see how well the lens resolved. If soft, resolution will be less. Not the same as MTF which may be more qualitative but it does show degree of sharpness. The question I have is does this linear measure also translate to best focus when shooting animals or birds?
 
The question I have is does this linear measure also translate to best focus when shooting animals or birds?
Yes, best focus is best focus.

Though it's nice and an interesting exercise to quantify lp/mm respolution the real key in focus fine tuning is that you're not really improving the lens focus or resolution abilities, you're just getting the focus sensor's sense of best focus to align with what's actually best focus. IOW, focus fine tuning doesn't make a lens any sharper, it just aligns the distance of best focus between the AF sensor and reality to account for slight optical path length differences between the main sensor and the AF sensor.
 
In photography light, light, light, the target looks very dull to me.
Try using a remote shutter release and a tripod, no hand holding when doing any auto fine tuning.
Nikon inbuilt AF tuning has not been any use to me as I cannot hold the two buttons and camera steady.
+1 for Dot Tune for something that has worked for me.
 
Thanks DRwyoming. Makes perfect sense that "best focus is best focus". As you say, the whole point of AF fine tuning is to make the AF sensor focus in agreement with best focus on the digital sensor, not to compensate for a lens' optical shortcomings . Live View does the same thing since you are looking at the image on the digital sensor which is how I confirmed AF fine tuning was correct. Below are 100% center crops at around 100' for both the 600E FL (left) and 600D AFS II (right). The 600E FL resolves group 0,5 (80 lp/mm) and the 600D AFS II resolves group 0,6 (88 lp/mm) however contrast appears a little higher on the 600E FL. The images were taken on different dates and environmental settings. Taking into account minor differences in shooting distances, the 600E FL measures closer to 85/p/mm.
_DSC1030, 600 f4E FL, center 85 lppmm, @100'.jpg
You can only see EXIF info for this image if you are logged in.
DSC_0440, 600 f4D AFS II, center 90 lppmm @103.jpg
You can only see EXIF info for this image if you are logged in.
 
Last edited:
Focus fine tune is always an interesting endeavor. I have tried and tested FoCal, Dot Tune, and making and shooting a variety of different test targets. These are a few of my thoughts keeping in mind there are many on this site that know tons more than I do about focus fine tuning.

I have never had much success shooting flat targets and trying to compare focus accuracy. I like to have a target that lets me see that I am moving the focus point forward or backward. I have found that shooting an appropriate test scene or target that will clearly identify the focus accuracy is the best starting point to identifying that the lens needs fine tuning and if it is front of back focusing. Aperture and focus distances are critical to maximize the focus fine tune accuracy and FoCal and other programs offer online calculators that identify the appropriate distance. A sports photographer I trust shoots an image of a ball on the grass to regularly test focus accuracy. I have also focused on a specific spot on a brick wall or fence running diagonally to check the focus.

I love FoCal (probably because the mathematician in me loves the curve fitting process) but it requires very consistent testing environments to get consistent results. I believe the basis of the program is that it measures the contrast across a series of images, projects a focus fine tune value and then takes addition images to verify and recalculate the value. I have had no luck with it outside because changes in lighting alter the calculations for each image. I have even found it to be inconsistent inside if there are windows that might alter the lighting intensity or direction during the testing process. I have also found that my 70-200 F2.8 E FL focuses much more accurately and is much easier to focus fine tune that my 70-300 AF P or my 200-500. I run multiple tests on a lens and look for consistency in the results if I identify problems.

My process has been to first identify front or back focus issues, focus fine tune using several different methods, then verify improvement in practical applications.
 
@MartyD Agree that just a flat target is not as useful to me.
I use a target that has a flat surface and a ruler attached see here where I discussed trying focus peaking for AF fine tune.
 
-7 AF fine tune with LensAlign Mk II (10 sample shots) appears best which is very close to -6 AF fine tune with the USAF resolution chart. With these two methods in very close agreement, I'll go LensAlign Mk II. The FoCal has a lot of nice features but it's AF Fine tuning method of phase detection with this lens is less accurate than desired.

DSC_0599.jpg
You can only see EXIF info for this image if you are logged in.
 
Last edited:
-7 AF fine tune with LensAlign Mk II (10 sample shots) appears best which is very close to -6 AF fine tune with the USAF resolution chart. With these two methods in very close agreement, I'll go LensAlign Mk II. The FoCal has a lot of nice features but it's AF Fine tuning method of phase detection with this lens is less accurate than desired.

View attachment 25710
I'd seriously have that lens checked - none of that is acceptably sharp. I can see that focus is in the right place, but the acuity is very low.
 
Hi Steve, I bought it from a seller on FM who provided documentation from Nikon APS (Morton Grove, IL) that shows the lens had a complete service 11/2020 including motor replacement, element cleaning, bayonet mount replacement and general CLA. Do you suggest sending it back to them or Nikon in NY? If so what should I ask them to do?
 
Hi Steve, I bought it from a seller on FM who provided documentation from Nikon APS (Morton Grove, IL) that shows the lens had a complete service 11/2020 including motor replacement, element cleaning, bayonet mount replacement and general CLA. Do you suggest sending it back to them or Nikon in NY? If so what should I ask them to do?
I'd send it to Nikon in NY and just tell them you're not getting sharp results. I'm not an optics expert by any stretch, but it looks like an element is out of alignment or something.
 
I agree - unless there is something in your technique that is allowing vibration to cause the softness. That lens should be much sharper.

Working with Nikon for service, be sure you send them your test images including the LensAlign samples. Let them see that it's not sharp so they don't think it's a calibration issue. You might send your camera body as well. I know it's a nuisance, but it gives them everything needed to assess.

I'm assuming you have your gear insured. Check to verify, but most policies cover losses if gear is being shipped. There is no need to buy extra insurance for your gear to be shipped - and on that lens it could save you $100.
 
Back
Top