Focusing behavior on Nikon d810 with 24-70/2.8 G ED

[John Di Leo]

I am using my d810 for artwork reproduction, paintings mostly of various sizes. I am using my much loved 24-70 Nikkor 2.8 G  ED.

I am trying to make sure that the sensor and the lean of the painting is coplanar.

I am using manual focus and have used both the M/A and the M on the lens and AF and M on the body in all combinations, but the result is the same, IOW that setting makes no difference on the issue.

I am on a tripod, using a timer shutter, shooting RAW. The problem is not shutter related; shutter speed does not make a difference.

I use live view to focus, zooming in to the max to critically adjust focus, and achieve it.

I trip the shutter and when I review the resulting "playback" image it is not the same sharpness as when I was focusing. It's slightly, but noticeably, off.

Some degree of pixel peeping is required, but it can be seen on the camera.

The images are still very usable, fortunately (shooting and processing art can be kinda hard!) but why does my focus seem to miss during capture when it was tack sharp in live view? Is the sensor better at showing a live view image because it is live, and capturing something digitally, then digitally displaying it, will always be second best. Or is something happening in my lens or body that causes it?

I am able to recover the sharpness I think (still awaiting the client's word on it) using wither Topaz Photo AI or Sharpness AI.

In my researching it I see there is a fine tuning for AF, but that I presume is something completely different.

Thanks

John

[ShunCheung]

The pre-VR (2007) version of the 24-70/2.8 AF-S is well known to be sharp in the center but less so into the edges. The AF-S VR version is more even from edge to edge but not as sharp in the center. Those are more party, group of people type lenses and don’t need to be sharp into the corners.

Is the subject a painting so something? Perhaps use a 60mm macro lens; it also depends on the size of the subject.

[joseph_smith3]

I would double check your technique by making sure that mirror lock up is set or that you use the self timer when you trigger the shutter to minimize any movement from the mirror or any other source. You can also vary the f stop used too and see if a smaller aperture improves sharpness. if you have access to the old manual focus Nikon 55mm macro lens, use it. It is a flat field design to eliminate any focus issues across the focus plane. I believe both the f3.5 and f2.8 versions of the 55mm are flat field. 

[ilkka_nissila]

That lens has a field curvature issue; essentially the plane of focus is curved in a complicated way. It would not be an ideal choice for photographing flat artwork. Use another lens if you can, such as a Micro-Nikkor or at least a prime lens.

 

Otherwise the 24-70/2.8G is an excellent lens.

[John Di Leo]

Interesting comments all.

8 hours ago, ShunCheung said:

Is the subject a painting so something?

Yes, all paintings and all flat; sometimes framed, sometimes just canvas. I got the lens new, probably around 2010, and I believe it is the first version. I see no Roman numeral II on it, and it is pre VR

 

7 hours ago, joseph_smith3 said:

I would double check your technique by making sure that mirror lock up is set or that you use the self timer when you trigger the shutter to minimize any movement from the mirror or any other source.

I shot it both from Live View and with the mirror down, same behavior. I am using a pretty heavy Bogen Tripod, without center extended. I shot 2 images with each shutter trip. When zoomed in in Lightroom I can sometimes see minimal variation in sharpness, but generally they look the same.

As I said it requires pixel peeping to see, like going to 1600% in Lightroom. I can't capture the LCD image of the focus to compare. I only see what I am focusing as I do and then the playback...and a noticeable difference in there.

 

7 hours ago, joseph_smith3 said:

You can also vary the f stop used too and see if a smaller aperture improves sharpness.

I have been using 5.6, 8, and 11, usually f8

 

7 hours ago, joseph_smith3 said:

if you have access to the old manual focus Nikon 55mm macro lens, use it. It is a flat field design to eliminate any focus issues across the focus plane. I believe both the f3.5 and f2.8 versions of the 55mm are flat field. 

Actually, I have the 55/3.5. I don't have the 105 (yet. Christmas is not far away!) I will play with the 55 at home to see if that changes things. The problem with a prime, and it is a physics problem is how we are shooting the artwork because of the artwork involved. Sometimes, rarely, it is on a wall, ie too intricately mounted or too heavy to move. Most often though we are moving them to a nicely lit area, shooting a grey card, then shooting the work. They vary in size a good bit. Friday we were shooting in an old plantation house and the largest work was 90" x 40" and the smallest about 12"x16", so the zoom is helpful for composition and as a time factor, IOW I don't have to keeping moving the tripod forwards and backwards. My knees are screaming from what we did do.

continued

[John Di Leo]

3 hours ago, ilkka_nissila said:

That lens has a field curvature issue; essentially the plane of focus is curved in a complicated way. It would not be an ideal choice for photographing flat artwork. Use another lens if you can, such as a Micro-Nikkor or at least a prime lens.

 

Otherwise the 24-70/2.8G is an excellent lens.

Hmm. When I focus I am focusing always in the center of the painting, and it shows tack sharp in the LCD. I do not recompose, but only trip the shutter, sometimes from Live View, sometimes to conserve battery, I turn live view off and shoot. When I check during playback I look again at the same spot on maximum zoomed in on the image, so it's not like I am focusing in the center then checking for focus at the periphery.

Yes, on the 24-70, it a favorite.

 

Thanks to all for the comments and suggestions.

I am going to put the 55 micro on and see how that does.

[John Di Leo]

I reproduced the issue just now with the 24-70. Then I tried it with the 55 at f8 and 1/4 sec. It handles it significantly better. There is still a drop off in sharpness—very slight, but not nearly as much.

So, I guess I will be sacrificing my knees for art.

Thanks for the guidance. Much appreciated.

 

[ShunCheung]

There is only one version of the F mount 24-70/2.8 AF-S (no VR), introduced on August 23, 2007 along with the D3 and D300. I have used 5, 6 different samples of that lens but have never owned one. I do have the later VR version introduced in late 2015 as an E lens. That is among one of the last 15 or so F mount lenses ever introduced as the Z era started less than 3 years later.

For document, painting reproduction, the F mount 50/55/60mm macros are better choices because they are flat field. Not sure that carries to the Z mount. 

[mike_halliwell]

I don't suppose the zoom lens is/or has moved during the exposure, ie zoom slip?

Sites like this can provide the best aperture to use for best edges

https://www.lenstip.com/298.4-Lens_review-Nikon_Nikkor_AF-S_24-70_mm_f_2.8G_ED_Image_resolution.html

https://www.lenstip.com/298.6-Lens_review-Nikon_Nikkor_AF-S_24-70_mm_f_2.8G_ED_Distortion.html

So, about 50mm @f8...😉

I suspect the sharp middle + soft edges is field curvature. If you Live View zoom into a corner and critically focus it and it comes out sharp, but the middle is not so great, that would prove it.

But, as Shun just noted, a fixed prime macro is designed for this type of work.

It might sound obvious, but the last time i did a big batch of flat copy, I went through them and sorted them in size order so i didn't have to move the set-up too much per different object.

 

Sometimes I wish Nikon would resurrect the 70-180mm macro....🙂

Having a modest zoom allows much more flexible framing of various sized art.

[jose_angel]

Ilkka i

10 hours ago, John Di Leo said:

When I focus I am focusing always in the center of the painting, and it shows tack sharp in the LCD. I do not recompose, but only trip the shutter, sometimes from Live View, sometimes to conserve battery, I turn live view off and shoot. When I check during playback I look again at the same spot on maximum zoomed in on the image, so it's not like I am focusing in the center then checking for focus at the periphery.

Are you focusing at the working aperture? Field curvature is a tricky characteristic, so I wonder if your plane of focus can vary when shooting at the working aperture.

There is also focus shift, where the plane of focus change with the aperture.

I have some very good lenses that show this annoying issues. When I just received the AFS 24-70 G back at the release time I checked focus as usual, and compared it to the lens I used to use before (AFS DX 17-35). The 24-70 was quite odd in comparison, so I remember to spend a lot of time trying to set the fine focus feature. At the end I left the lens at the factory settings, the most useful one. As Shun says, is not a lens for that task, or at least in the way you are using it. Despite of this I consider it an amazing lens for indoor people photography and portraiture.

For copy work I use the old 55/2.8 and 105/4 Ai-S Micro-Nikkors (size depending), with good results even on my D850. Another practical option (faster to use) is the AFS105VR.

 

 

 

Edited July 10, 2023 by jose_angel

[Dieter Schaefer]

1 hour ago, jose_angel said:

Are you focusing at the working aperture?

On the D810 using live view that should be a given - provided you don't press the DOF preview button as that will open the lens to full aperture. Though I don't know if there is an aperture value cutoff beyond which the lens is not stopped down further (like on some mirrorless which focus at working aperture down to f/5.6 or f/8 but then stay at that value even if the lens is stopped down further). Easy to check though by looking at the front of the lens when changing aperture. Of course, focusing manually at working aperture can be both a blessing and a curse. It certainly helps with field curvature and focus shift issues but because of the increased DOF is not optimum for achieving critical focus. I assume you use the longest shutter delay available on the D810 and EFCS to avoid vibration-induced loss of sharpness. Also, with f/11 you might be reducing sharpness slightly due to diffraction. In addition, looking at 1600% magnification is a bit of an overkill - but it could reveal that the tripod vibrations have not subsided completely. Are you close to a street with traffic? Or any motors, pumps etc. operating on the same floor. People walking while you are taking the shots?

[John Di Leo]

great comments...I'll answer later, but generally I am doing what is suggested...It's just the 24-70, I'm asking it to do something it is not designed for. As many have said I should use a prime and will in the future of this job. Also, I have the 55/3.5 and shoot manual anyway.

Aside: There are three settings for manual focus vs AF on the lens/d810. There is the MA/M on the lens, there is the AF/M on the body, and there is the mode M.

If one is shooting manual focus and exposure, do the settings on the lens and the body  make a difference if the mode is set to M? I do not see any difference in outcome.

[mike_halliwell]

4 minutes ago, John Di Leo said:

mode M

AFAIK, Mode M refers to Manual aperture, Manual shutter speed and/or Manual ISO,

not Auto Focus or Manual Focus.

[mike_halliwell]

On 7/9/2023 at 3:10 AM, John Di Leo said:

I am trying to make sure that the sensor and the lean of the painting is coplanar.

Using the grid screen helps ensure 'squareness' on modern machine made frames, but on old 'manual' frames or stretched canvases it's possible to make them square or rectangular even if they are not in reality.

 

[John Di Leo]

1 hour ago, mike_halliwell said:

Using the grid screen helps ensure 'squareness' on modern machine made frames, but on old 'manual' frames or stretched canvases it's possible to make them square or rectangular even if they are not in reality.

 

Right, and don't get me started on irregular fenestrated frames.

[John Di Leo]

6 hours ago, Dieter Schaefer said:

7 hours ago, jose_angel said:

Are you focusing at the working aperture?

On the D810 using live view that should be a given - provided you don't press the DOF preview button as that will open the lens to full aperture.

When in Live View, changing the aperture does not change  the brightness of the image. I see a flicker when I change the f stop, but the image does not change brightness.

6 hours ago, Dieter Schaefer said:

Easy to check though by looking at the front of the lens when changing aperture

Yes, the aperture changes, but the display stays the same, and double checking, yes, I am in manual mode

 

6 hours ago, Dieter Schaefer said:

I assume you use the longest shutter delay available on the D810 and EFCS to avoid vibration-induced loss of sharpness. Also, with f/11 you might be reducing sharpness slightly due to diffraction.

yes, a 10 second delay. I did not have ECFS enabled...it is now. I was shooting mostly at 5.6 and 8

 

6 hours ago, Dieter Schaefer said:

looking at 1600% magnification is a bit of an overkill - but it could reveal that the tripod vibrations have not subsided completely. Are you close to a street with traffic? Or any motors, pumps etc. operating on the same floor. People walking while you are taking the shots?

Yes, definitely overkill. Is 1600% the amount of mag seen on the LCD when completely zoomed in during Live View? We were way off the beaten path at a 200 year old plantation down a backroad. No traffic, no pumps or anything vibrating nearby. I could feel no vibration from my feet. We shot on a concrete floor as well as a VERY sturdy 2nd floor veranda. I do not think here were ambient vibrations and I was tuned to that. I think it's a function of the zoom lens because I was able to reproduce it at home, and correct it, 90-95%, when I changed to my 55 micro.

But thanks for the interesting comments. Appreciated.

[John Di Leo]

8 hours ago, jose_angel said:

There is also focus shift, where the plane of focus change with the aperture.

I am using manual focus, so that should not be an issue, right?

 

8 hours ago, mike_halliwell said:

I don't suppose the zoom lens is/or has moved during the exposure, ie zoom slip?

Correct. The zoom did not move after focusing.

 

8 hours ago, mike_halliwell said:

So, about 50mm @f8...😉

I went back to see my details and it looks like most were shot between 32 and 60mm and at f8

8 hours ago, mike_halliwell said:

I suspect the sharp middle + soft edges is field curvature. If you Live View zoom into a corner and critically focus it and it comes out sharp, but the middle is not so great, that would prove it.

I just tried that, maybe I did it wrong or not critically enough? I Live Viewed critical focus near the edge of a frame, shot the picture without recomposing and the center is sharper than the edge—where I focused. Sounds like the opposite of what you describe.

8 hours ago, mike_halliwell said:

But, as Shun just noted, a fixed prime macro is designed for this type of work.

It might sound obvious, but the last time i did a big batch of flat copy, I went through them and sorted them in size order so i didn't have to move the set-up too much per different object.

Yes, on Shun's comment and I will use my 55/3.5 as much as possible from here. And regarding logistics, that is what we did as much as we could.

 

Thnaks again

[robert_bouknight1]

Late jumping in...When I read the OP I thought of recommending using a 55/3.5 MF for this project, I see that others are on the same page.  I use one of my copies frequently for copying building plans (cheap scanner).

There is some internet noise that the first version with compensating aperture might be slightly better at close ranges than later versions, while the later versions might be better at distance.  I have example(s) of both, I think.  Someday, I plan to test this rumour, wonder if 45MP will be enough resolution to tell.

For a while they were too cheap on Ebay for me to pass up, but they seem to have appreciated a little.  Still a bargain, IMO. 

[John Di Leo]

Let me add a red herring to this discussion.

2 days ago the 24-70 became very reluctant to AF.

I usually use back button focusing and pressing the AF-ON button >90% of the time now did nothing. Changing focusing modes and other parameters changed nothing, no AF.

When it did focus on those rare button pushes, it snapped to focus, like something wasn't making contact. All settings  are appropriate, ie on MA and AF, and the focusing indicators in the VF light up. Other lenses behave properly. I checked the contacts for spring...fine; I cleaned the contacts both on the lens and the camera...no joy.

I can still manually focus the lens fine. So, it is on its way to Jericho NY for service...ugh.

I do not think this has ANY bearing on the issue discussed here as I was using manual focus for the shooting.

Question though: If the lens is set to MA or the body is set to AF, and fine tuning of focus is done using the focusing ring, does that have any adverse effect on the focusing mechanism? I notice that when the camera and lens is set to AF and MA and I turn the focus ring, it can be very stiff. When on M and M, the ring moves more smoothing and with little effort. So, is that a bad thing to do, stressing the internal focusing mechanisms of the lens?

[ShunCheung]

That is an AF-S lens. You can turn the focus ring on the lens while the body is set to AF. A screwdriver AF lens would have been a different story.

[John Di Leo]

4 hours ago, ShunCheung said:

That is an AF-S lens. You can turn the focus ring on the lens while the body is set to AF. A screwdriver AF lens would have been a different story.

Meaning, "Don't do that!" ?

[ShunCheung]

For the early screwdriver AF lenses, commonly known as AF and AF-D lenses, before AF-I and AF-S, the AF motor is inside the body and there is a mechanical connection from the motor through a “screwdriver” shaft to the focusing mechanism inside the lens. Both manual focus and AF use the same mechanism. If the camera is set to AF and you turn the focus ring on the lens, your turning will be transferred through that shaft to turn the AF motor and gear inside the body. You can potentially do a lot of damage if you force that to turn.

When you switch the body to MF (manual focus, not medium format 😅), the body disengages that whole mechanism.

[John Di Leo]

Thanks. I have the 85 1.8 D.

Is that the difference between the D and G lenses

I will probably trade the 85 in. I don't trust myself to remember that 100% of the time.

[Rick Helmke]

I’m late getting to this as well but the simplest answer seems to me to just use manual focus and see if you can focus a sharp image by yourself. If you can, that narrows the possible problems.

 

Rick H.

[John Di Leo]

Some follow up

1) I started using the 55/3.5 and a 35/1.8 to good effect

2) I still notice, with both lenses above, using MF that the playback image is not as sharp as the focused image. Put another way, when I am focusing manually, I zoom in to the max using the + button on the camera body and pull focus. I move my hand, let the camera settle and I take the picture, sturdy tripod, no apparent vibration, timer, and when I review the image using the same degree of magnification there is a very slight difference in sharpness, less than there was with the 24-70, and VERY acceptable, but a difference.

Is this just the nature of the beast, that a sensor when recording an image may not be as sharp as when focusing manually? It seems it should be the same.  I notice this on both the d810 AND the z8 which I've finally been brave enough to take to a shoot. And TBC, this is max pixel peeping in camera and Topaz Sharpen AI makes short work to bring it even tackER sharp. It's a curiosity. It would not be noticeable if shooting portrait or street or general photos, but I am shooting paintings for an exhibit's catalog and the paintings have small geometric designs

I will take test pictures of the lcd when focused and then on playback and post over the weekend.

3) As I posted on 7/14, on 7/12 the AF of my 24-70 started being reluctant to focus. Sent to Nikon and $588 later I have a new silent wave motor.

 

Edited August 12, 2023 by John Di Leo