How to get critical sharpness - Nikon 500 PF?

[bjanes]

Forget the charts, forget the theories, look at your images. If the target is sharp, the corresponding area of sharpness in front of the target will appear greater than what is behind it. This is very easy to confirm.

You can only see EXIF info for this image if you are logged in.

Before we get all bent out of shape as to who is correct, we should check our definitions. I am interpreting front as proximal or nearer to the camera as shown here. At any rate, that should have been clear from the PhotoPills screen show I originally posted. Perhaps you are using the reverse definition. In any event, the correct situation is shown here.

Bill

 

[Capturingtheunseen.com]

View attachment 6595

Before we get all bent out of shape as to who is correct, we should check our definitions. I am interpreting front as proximal or nearer to the camera as shown here. At any rate, that should have been clear from the PhotoPills screen show I originally posted. Perhaps you are using the reverse definition. In any event, the correct situation is shown here.

Bill

I'm not bent out of shape, I'm just saying to look at your images instead of a chart or a graph. That's all. lol

 

[Woodpecker]

Before we get all bent out of shape as to who is correct, we should check our definitions. I am interpreting front as proximal or nearer to the camera as shown here. At any rate, that should have been clear from the PhotoPills screen show I originally posted. Perhaps you are using the reverse definition. In any event, the correct situation is shown here.

I'm not bent out of shape, I'm just saying to look at your images instead of a chart or a graph. That's all. lol

Well, as always the thruth is somewhere in between .

YES, it makes sense to look a chart or graph like the one above, because it just describes and explains the physical truth.
YES, it does makes even more sense to look at your pictures instead of charts or graphs only, because

1. you have to find out that something is wrong with focusing in the first place
   
2. you have to take your own photos of something looking like a chart or a graph -i.e. a AF fine tuning chart or tool - to correct the problem.

YES, it even makes sense to apply the 1/3 / 2/3 rule as a principle (not necessarily mathematically correct), because - if talking wildlife - in the vast majority of cases the most part of the animal whose eye you focus on is behind this eye, so at least I want to have the larger portion of the DOF behind the focus layer. There might be exception like be a chameleon walking away from you while still staring at you with one eye .

You never can tell exactly - at least not with the tool I use (Spyder LensCal) - but when I am doing AF fine tuning in the last cycle I end up having two, three photos where the desired point (focus layer) is eqqually sharp, but the distribution of the blur in front or behind the focus layer is distributed differently. From these photos I choose the one where for two points on the scale that are equally far away from the focus layer (one in front, one behind) the one behind the focus layer is less blurred than the point in front of the focus layer. This photo is then giving me the AF fine tune value I put in the camera. This way I can at least make sure that I have the focus layer exactly sharp and the DOF is distributed the right way meaning I have the bigger part behind the focus layer. It is possible to do this academically correct, but for most of what I do this is good enough.

Why did I start doing this ? When I got my 500 f4 I had shot some test photos wide open with short subject distance. As the shorter the subject distance gets the DOF gets smaller as well and can get pretty small. I ended up with photos with the AF point at the right place, sharp eye, a sharp beak ... and a sharp blackberry that was hanging slightly closer to me than the bird, but to the back already the second half of the head started to get blurry. The guy that introduced me to the existence and implications of AF fine tuning some years ago recommended to me the above method and it worked.

If you deal with narrow DOF differences of 2 digits in the AF fine tune are most likely visible. Now look at the current settings for my cameras e.g. for my 500 f4:

D4S --> -8
D750 (1) --> 0 (This camera has been to Nikon service together with the 500 f4 sue to a little tripod accident ...)
D750 (2) -->+6
D7200 --> +2

I recently was able to purchase a 500mm PF lens for my D850, but try as I might I can’t seem to get critical sharpness like a majority of the reviews suggest I should. I tried AF fine tuning it but still no luck. My 200-500 on my D500 is much sharper but shouldn’t be the case .

With the 500 PF it is a little bit less critical critical because of f5.6 giving more DOF shooting wide open, but if you happen to have a camera that is a bit at the edge of the tolerance band you will still recognize it. At least I did not expereince any particular drawbacks with the 500PF on any of the cameras I currently have.
Of course compared with the D850 I have much less resolution on all of them but If you have a 500PF that is outperformed in sharpness by the 200-500 and you can't cure that by AF fine tune I would suggest that something might be wrong with your lens and this is not as uncommon as you might think. I needed three loops to get a 300 PF that was o.k.

 

[bjanes]

I'm not bent out of shape, I'm just saying to look at your images instead of a chart or a graph. That's all. lol

Mia Culpa!

I took a picture of a soup can placed at 10 feet with a 35 mm lens at f/2.8 and observed the sharpness in front and behind the can with the results shown. The arrows indicate the approximate depth of field. There is more depth of field in front of the focusing distance (black line) then behind.

You can only see EXIF info for this image if you are logged in.

I don't know how to reconcile this with the PhotoPills DOF representation shown below:

You can only see EXIF info for this image if you are logged in.

Perhaps someone with more technical expertise can explain.

Bill

 

[Capturingtheunseen.com]

Mia Culpa!

I took a picture of a soup can placed at 10 feet with a 35 mm lens at f/2.8 and observed the sharpness in front and behind the can with the results shown. The arrows indicate the approximate depth of field. There is more depth of field in front of the focusing distance (black line) then behind.

View attachment 6603

I don't know how to reconcile this with the PhotoPills DOF representation shown below:

View attachment 6604

Perhaps someone with more technical expertise can explain.

Bill

Thanks for your honesty, Bill. I admire that.

This is why I don't put much stock in charts, formulas, etc.

I used to teach land navigation in the military. I always told my people "The map is not the territory." Meaning what is on paper is not always reality.

 

[bjanes]

Thanks for your honesty, Bill. I admire that.

This is why I don't put much stock in charts, formulas, etc.

I used to teach land navigation in the military. I always told my people "The map is not the territory." Meaning what is on paper is not always reality.

I started a thread on the DPReview Science and Technology forum whose subject was the apparent paradox between depth of field calculators and my field observations concerning the depth of field proximal to and distal to the focused distance. Many technically savvy users participate in that forum. It turns out that my methodology of looking at blades of grass was not sufficiently sensitive to resolve the problem.

Here is the most helpful post that resolves the problem. With a 35 mm lens at f/4 focused on a slanted ruler at 10 feet, the depth of field is approximately 1/3 in front (proximal to the camera) and 2/3 distal, in agreement with the depth of field calculator. There is no paradox or disagreement between theory and practice.

Bill

 

[fcotterill]

If AFFT is erratic or you need to check a lens relatively quickly. Rig lens on a rigid tripod. In Liveview focus midway [ @ ~90mm mark] along a 150mm steel engineer's rule, which is lying flat at a shallow elevation on a strip of fabric, light coloured cloth/furred hide etc. Position to minimize any reflection off the rule.

The plane of focus shows up well on a substrate with fine fibres or hairs. The marked rule is a standard target and scale to compare your test photos.

This needs decent light conditions. A problem with front or back focus will be obvious and can be corrected iteratively with the AFFT menu setting.

 

[Viathelens]

I've read these reponses and viewed Steve's video (I'll also read the section on this subject in his nikon book) but I have a concern about shooting distance. In the Nikon support information (web site reference provided) they say:

"Do not select a focus distance which you do not regularly use as this may lead to unsatisfactory results later."

Nikon Web Site:

https://www.nikonimgsupport.com/eu/BV_article?articleNo=000006300&configured=1&lang=en_GB

For me there is no "regular distance." I could be shooting at 15 feet away (porch for hummers) or at infinity (wildlife in the distance). Has anyone found this distance issue to be a problem after fine tuning? I have not fine tuned, at this point, any of my lenses. Would I have to re-fine tune for each trip? Or would Steve's method of multiple settings and then average do this already for the most part?

 

[Capturingtheunseen.com]

It turns out that my methodology of looking at blades of grass was not sufficiently sensitive to resolve the problem.

That's exactly my point, Bill. Who cares what the formulas and calculators say if you can't see it in the image? You looked at that image and saw that the depth of field was greater in front of the target than behind. So would everyone else who looked at the image.

To me, that is ALL that matters. Reality, not theory.

 

[Capturingtheunseen.com]

I've read these reponses and viewed Steve's video (I'll also read the section on this subject in his nikon book) but I have a concern about shooting distance. In the Nikon support information (web site reference provided) they say:

"Do not select a focus distance which you do not regularly use as this may lead to unsatisfactory results later."

Nikon Web Site:

https://www.nikonimgsupport.com/eu/BV_article?articleNo=000006300&configured=1&lang=en_GB

For me there is no "regular distance." I could be shooting at 15 feet away (porch for hummers) or at infinity (wildlife in the distance). Has anyone found this distance issue to be a problem after fine tuning? I have not fine tuned, at this point, any of my lenses. Would I have to re-fine tune for each trip? Or would Steve's method of multiple settings and then average do this already for the most part?

Great questions. I will wait to hear how it is answered.

 

[DRwyoming]

For me there is no "regular distance." I could be shooting at 15 feet away (porch for hummers) or at infinity (wildlife in the distance). Has anyone found this distance issue to be a problem after fine tuning? I have not fine tuned, at this point, any of my lenses. Would I have to re-fine tune for each trip?

I'm much more concerned about how AF Fine Tuning works on closer distances that I'll typically use than far away distances. The way I look at it, DoF is very limited at or near the close focus distance of the lens but when out near infinity focus for the same aperture and same lens there's a lot more DoF which covers small AF errors.

For example, take one of the common 500mm f/5.6 lenses shot wide open at f/5.6. With 12 feet of distance to the subject the DoF is 0.62 inches on a FF camera which is pretty critical for keeping the eyes sharp if there's even a bit of front or back focus with the camera and lens combo. Take the same lens at the same aperture and focus on a subject 100 feet away and the DoF increases to roughly 49 inches which is a lot of tolerance for small amounts of focus error. If the subject is say 300 feet away like a Grizzly at the National Park stand back distance the same lens and aperture gives you a DoF of 446 inches (37 feet) and tiny focusing errors won't have a big impact on focusing on your subject.

[edit] BTW, I realize this is a simplistic view and that in addition to DoF changing with subject distance the magnitude of front or back focus errors also likely scales with subject distance but I still think it's less of an issue for far subjects where you tend to have excessive DoF compared to close subjects where focusing is a lot more critical.

So if I have to pick one distance to place the target during AF Fine Tune adjustments I'll choose a distance that's pretty close but more or less representative of how I plan to use the lens most of the time as critical focus becomes less of an issue as distance to subject increases.

DoF estimates based on this calculator: https://www.dofmaster.com/dofjs.html

 

[Viathelens]

I'm much more concerned about how AF Fine Tuning works on closer distances that I'll typically use than far away distances. The way I look at it, DoF is very limited at or near the close focus distance of the lens but when out near infinity focus for the same aperture and same lens there's a lot more DoF which covers small AF errors.

For example, take one of the common 500mm f/5.6 lenses shot wide open at f/5.6. With 12 feet of distance to the subject the DoF is 0.62 inches which is pretty critical for keeping the eyes sharp if there's even a bit of front or back focus with the camera and lens combo. Take the same lens at the same aperture and focus on a subject 100 feet away and the DoF increases to roughly 49 inches which is a lot of tolerance for small amounts of focus error. If the subject is say 300 feet away like a Grizzly at the National Park stand back distance the same lens and aperture gives you a DoF of 446 inches (37 feet) and tiny focusing errors won't have a big impact on focusing on your subject.

So if I have to pick one distance to place the target during AF Fine Tune adjustments I'll choose a distance that's pretty close but more or less representative of how I plan to use the lens most of the time as critical focus becomes less of an issue as distance to subject increases.

DoF estimates based on this calculator: https://www.dofmaster.com/dofjs.html

Thanks!