D810 Critical Focus - PV button with OVF / Focus Indicator

MaxBerlin

Either my understanding of or the Nikon PV button function is convoluted.

Here's why -

The focus indication is pretty reliable in the D810 when a lens has a minimal amount of focus shift (e.g. an Otus) and a number of other compositional factors are in control -focal length, f stop and distance to subject and their interrelationships.

Without the PV button depressed you are focusing at wide open, in this case an Otus 55mm. With the f-stop set in the camera at f 2.0 the shift is minimal and sharpness is retained. As one goes up to f4 or higher they can rely increasingly on DOF to cover any critical focus errors as long as the subject is far enough away to allow deeper DOF to come into play. But this isn't always reliable. As subjects are nearer, critical focus is lost as the DOF contracts. In order to work around this, one has to use the PV button.

However, when one presses the PV button while looking through the OVF the focus indicator isn't shown.

Thus it's impossible to focus under the actual aperture the image will be captured and have the assistance/confirmation of the OVF internal focus indicator at the same time. (If anyone knows a setting that can change this, please advise)

Based upon this the only 'workaround' is to use the live view on the LCD in conjunction with a loupe, activate the PV button and then achieve critical focus through the loupe and LCD in live view.

Seems like Nikon (and Nikon seems to be a way more responsive company than Sony) could easily fix this.

Any reason Nikon can't make it possible to use the PV button and the focus indicator at the same time while using the OVF ?

danhowl

I could totally be wrong, but I always assumed that critical focus was the center of focus (whether it be an old split-image focusing screen or an OVF/EVF) and the the DOF was 1/3 to the foreground and 2/3 to the background due to the properties of focus and lens barrels. This shouldn't change if you are stopped down or wide open.

I don't know why the PV button interrupt the OVF, but it was my assumption that that the PV button was for DOF more than critical focus.

haroldp

@MaxBerlin

Thank you for calling attention to this.

Focus shift is not well understood by most.

When I autofocus fine tune, I usually take one series wide open and another down one stop, and their is often a slight difference. My practice has been to set the offset in between the two optimums si it is never off too far, but with modern assymetric designs and aspherical elements, focus shift when stopping down is real.

Leica had to redesign one of their best lenses (24mm f1.4 asph) because it had so much focus shift that customers avoided it. The new one actually has moving elements to compensate, as well as improve near field sharpness.

... H

Ironheart

The only other workaround that I know is to determine the fine tune setting for each aperture, or at least every other, write it down and plug it in as you change aperture. PITA? Yes, but not much other choice.
You know you can zoom live view too if you don't have a loupe.

spraynpray

I will own up by saying that I was aware of focus shift with zoom, but I was not aware that focus even could shift when stopping down. I have always thought that the increasing DoF with smaller apertures made that something I didn't have to worry about.

I learned something new today on NR.

Ironheart

Nasim covers this well:
https://photographylife.com/what-is-focus-shift

Ironheart

Oh, also, the reason this isn't too much of a problem with Nikon AF lenses is because Nikon factory lens calibration and camera AF systems compensate for this automatically. The issue is much more pronounced with non-Nikon MF lenses.

MaxBerlin

Ironheart - there may be some process that compensates during AF and it may operate at high percentage rate but a lot of this is due the baseline of the lenses (the trinity @ 2.8, or even the f4 lenses which only have less than 2 stops to cover before AF capabilities and diffraction erode IQ. For certain though they aren't capable of critical focus other than randomness.

Using live view without a loupe (aka sunshade) is difficult to impossible for me in daylight.

The nice compromise would be for Nikon to enable 'focus confirmation' while the PV button is activated.

PB_PM

Oh, also, the reason this isn't too much of a problem with Nikon AF lenses is because Nikon factory lens calibration and camera AF systems compensate for this automatically. The issue is much more pronounced with non-Nikon MF lenses.

As long as the elements are centred properly that could to be true, but does not seem to be the case. The AF-S 28mm F1.8G, among a few others, has been widely reported to have focus shift when stopped down, so clearly Nikon's AF system is not correcting for focus shift on the fly.

Ironheart

I didn't say it was perfect, I said it was better. I own the 28mm, so I understand what it's doing. No doubt we are pushing up against the limits of the PDAF system.

PB_PM

Camera makers, especially in the m4/3s camp, rely heavily on in camera, and post processing algorithms to make up for lazy optical designs. Maybe lazy is not the right term, so much as compromised, to keep the lenses lighter and smaller as users are demanding them to be.

I don't think the issue is the PDAF system, since mirrorless systems lenses also suffer from optical issues, and focus shift. The issue is the construction of the optics. Mirrorless systems still focus with the lens wide open for the same reason that PDAF systems do, the requirement for light to focus within the time frame users expect.

haroldp

I think the main reasons why aperture based focus shifting has become more of an issue lately are:

1- The greater resolution of digital sensors requires higher resolution optics which has resulted in more asymmetrical designs using more aspherical elments. CAD / CAM design and manufacturing enables mass production of such lenses. It is likely that classic symmetrical double gauss designs with spherical elements were less likely to focus shift. They also could not achieve the performance of modern lenses.

2 - Digital sensors have higher resolution, and flatter and thinner focal planes than film emulsions and are both more sensitive to slight focus errors and more likely to display it. Magnifications are also much higher than in the film era.

Focus shift with zoom is a different problem and common in lenses designed for stills which are presumed to be re-focused after zooming.

Zoom lenses not having such shift are called para-focal, and are common in movie / video designs which is why the best ones are so expensive.

... H

Vipmediastar_JZ

@maxberlin
I have a the 50mm 1.2 ais that I got for the DF. I ended up getting the K3 focusing screen and This helps greatly. At 1.2 it is rather sharp and focused in comparison to the 50mm 1.8g wide open. With the D810 at 1.2 it is rather soft or doesn't seem like its focused until f2.

Before the focusing screen with the DF I was zooming in via lcd or relying on the green dot. I almost never use the PV function but I understand how beneficial this would be to you.

Perhaps a focusing screen can benefit you?

MaxBerlin

I could be wrong but 99.9% sure that focus shift is strictly a property of the lens.

The issue is how Nikon doesn't allow one to use the OVF and the focus indicator while the PV button is depressed. The focus indicator isn't viewable through the OVF once the PV button is activated and held down.

I don't believe that a focusing screen can mitigate or remedy this. Live view can (in fact the PV button locks in live view).

Focus shift is a completely different subject than the quality of parfocal for a zoom lens.

BTW- Nikon users have 3 choices for a parfocal lens - but these lenses may also have focus shift despite being parfocal. http://www.lensrentals.com/blog/2011/02/photo-lenses-for-video/4

Ironheart

FYI, those lenses are not truly parfocal, they just minimize it under most conditions. If you really need parfocal and no focus shift, for $20K this will deliver the goods:



http://www.bhphotovideo.com/c/product/895001-REG/zeiss_2008_990_28_80mm_t2_9_compact_zoom.html

Ironheart

Back on topic, Zeiss is well aware of focus shift, and address it directly in this article:
http://www.zeiss.com/camera-lenses/en_us/website/photography/what_makes_the_difference/manual_focusing.html
Some of the things they recommend are, replacing the focus screen (thanks @Vipmediastar_JZ), using live view, and NOT relying on the AF confirmation.

@MaxBerlin, if you think about it, even if Nikon were to enable the focus confirmation dot with DoF preview engaged you would only be able to use it to f/5.6 , maybe f/8 on the center point only.

MaxBerlin

Ironheart - can you please indicate the text where Zeiss recommends that a different focusing screen will remedy focus shift? This passage on the same page indicates it does not. An earlier passage suggests that certain screens may help in overall manual use but again this would not address focus shift. PS Ming Thein might be the source for starting replacement focus screen edification. I;m not pursuing it because live view and pv button under 3x loupe assures critical focus. But would like to have higher hit probabilities via the EVF when not critical.

In order to take into account the effect of focus shift on precise focusing, the user should – as far as possible – carry out focusing at the aperture at which the shot is subsequently to be taken. Nevertheless, there are still a number of limitations to be considered: AF systems in current camera models do not take into account the focus shift of the lens. This means that the reliability of the AF indicator varies depending on the type of lens, f-stops and shooting distances.

Due to their surface structure, focusing screens in fast lenses are not capable of factoring in all the incoming rays. This can easily be seen from the fact that – depending on the design – the brightness of the focusing screen no longer changes with lenses such as those that are faster than 1:2.8. This means that the change in the optimum focus position can no longer be evaluated on the focusing screen at wide apertures when using very fast lenses.

Ironheart

This seems pretty straightforward:
"By replacing the focusing screen with a variant that has been optimized for manual focus, it is possible to achieve some major improvements. "

The passage you cite brings up the issue that the screen may not improve visual focusing faster than 2.8, but this isn't where you are having the issue, no? At 2.8 or faster the focus confirm will be sufficent, and at slower than 2.8 the AF becomes less and less useful anyway, and that's where the focus screen will help. Also, it's relatively easy to swap the screen out and back again, so no blood, no foul :-)

Ironheart

@MaxBerlin, I hope you are watching this thread with as much interest as I am:

http://forum.nikonrumors.com/discussion/4386/split-imagemicroprism-focusing-screen-for-d600

MaxBerlin

Thanks Ironheart. I just did a new post on Focus shift with PV button active and non-active with the Otus. It's on my blog but I am not supposed to post it.

Ironheart

You can put the top-level URL to your blog in your sig. Lots of folks do. I for one would like the link as I am fascinated by your deep technical dive into this stuff.

MaxBerlin

Looks like it's working in the sig now. Thanks for the help.

Ironheart

Question, why do you need to use a loupe on the rear screen when you can zoom? I would think the zoom would be much more effective, at least it is for me when achieving critical focus on starts and planets.

MaxBerlin

I am using both, the zoom is configured to 1 less than maximum by depressing center button under live view, the loupe helps to get fine detail as well as block incoming light (especially in the daytime) In the studio I prefer to use HDMI out to a large monitor.

Ming and Lloyd also use a 3x loupe except I thought the Zacuto was the ultimate rip off and felt the Hoodman was the proper amount of KY for what was surely to be my greatest shame.

This was compounded by the reality that none of them allow an RRS L -bracket so it really hurts.