I've about had it with Nikon's VR

But I have noticed that when I get something similar, the more pp specifically sharpening and contrast applied, the more prevalent it is. Maybe something different, but looks the same.

michaeladawson said:

Is that a reasonable summary of what you are saying?

Would it be reasonable to infer that nisen bokeh would be much less prevalent if aspherical lens elements were not used? I suppose theside effect of this would be less/imperfect correction of other types of distortions and aberrations. And lenses would be bigger and heavier. And perhaps quite a bit more expensive as a result.

Click to expand...

That is a reasonable summary.

If aspherical lenses were not uses nisen bokeh would not occur in the BG elements; it would occur in FG elements instead. IDT there is such a thing as a perfectly corrected lens; but if there were it would have "perfect bokeh" (even throughout) in both the FG and BG.

But aspherical elements are also used to correct for longitudinal chromatic aberrations where different wavelengths focus at different distances. And if the different wavelengths, or regions of the elements are not focused at the same distance, then the lens cannot be as sharp. This can often be seen as a colored ring to a white bokeh ball.

This example shows soap-bubble (bright edge), LoCA (magenta outline), and some onion skin characteristics; which all have different causes. This is actually an example from the FG of an undercorrected lens.



SA can be reduced by reducing how much of the lens periphery is used (stopping down); LoCA cannot. So SA is often slightly overcorrected in favor of CA. Especially since SA is most pronounced close to the focal plane; if the point is severely defocused SA will not be apparent, and if the point is sufficiently in focus it won't be apparent either.


Conversely, undercorrected SA is sometimes a good tradeoff for a portrait lens; as it leads to those "creamy BG's" photographers want.

PF lenses are quite a bit different; they correct for CA by cancellation; because the PF element reverses the distances at which the different wavelengths focus. But it still has to compensate for SA. And a PF lens can have that ring type of characteristic to the bokeh just due to the steps built into it. I believe both of those can lead to nisen bokeh... I have no idea if it could be classified as being more/less prone to it though.
--
https://www.flickr.com/skersting

Exactly. Critter is sharp as can be.

wcan said:

I would think if the subject is tack sharp, than the VR is doing it's job, end of story. If there is some type artifact in the OOF areas, what could the VR system do differently?...its job is to keep the subject "in one place" on the sensor.

Click to expand...

Boken quality is very much a function of lens design and VR implimentation.

Mako2011 said:

barkleytom said:

I am with you here. This has been a common occurence for many years and nothings was done about it!

Click to expand...

The manufacture can't because the issue shown here has nothing to do with VR. The issue is related to nisen bokeh which is more an inherent lens/atmospheric/scene dynamics issue rather than a VR one. Happens with or without stabilization and not limited to Nikon

Click to expand...

Optic designs/VR implimentation should mitigate this issue. Why mostly seen in Nikon VR lens?

well done Leonard!

barkleytom said:

Mako2011 said:

barkleytom said:

I am with you here. This has been a common occurence for many years and nothings was done about it!

Click to expand...

The manufacture can't because the issue shown here has nothing to do with VR. The issue is related to nisen bokeh which is more an inherent lens/atmospheric/scene dynamics issue rather than a VR one. Happens with or without stabilization and not limited to Nikon

Click to expand...

Optic designs/VR implimentation should mitigate this issue. Why mostly seen in Nikon VR lens?

Click to expand...

It isn't.

wcan said:

I would think if the subject is tack sharp, than the VR is doing it's job, end of story. If there is some type artifact in the OOF areas, what could the VR system do differently?...its job is to keep the subject "in one place" on the sensor.

Click to expand...

I suppose VR operating to keep the subject "in one place" on the sensor could perhaps result in artifacts in OOF areas perhaps, but I would think that if the VR is keeping the subject tack sharp, the tradeoff may be artifacts in OOF areas. So it comes down to perhaps a tradeoff.

wcan said:

I suppose VR operating to keep the subject "in one place" on the sensor could perhaps result in artifacts in OOF areas perhaps, but I would think that if the VR is keeping the subject tack sharp, the tradeoff may be artifacts in OOF areas. So it comes down to perhaps a tradeoff.

Click to expand...

It can be wise to recognise that sometimes there is more than a single issue to resolve.

Very occasionally I find a situation where VR is moving perhaps to the right to keep the main subject steady and something like wind is moving foliage to the left, that there can be a minor background issue.

If the intended subject is stationary and a tripod is used with a longer shutter speed instead of VR, background movement is more blurred.

It is up to each photographer to resolve competing issues, except that if a tripod is not available or the main subject is moving then VR may be the only likely solution.

wcan said:

wcan said:

I would think if the subject is tack sharp, than the VR is doing it's job, end of story. If there is some type artifact in the OOF areas, what could the VR system do differently?...its job is to keep the subject "in one place" on the sensor.

Click to expand...

I suppose VR operating to keep the subject "in one place" on the sensor could perhaps result in artifacts in OOF areas perhaps, but I would think that if the VR is keeping the subject tack sharp, the tradeoff may be artifacts in OOF areas. So it comes down to perhaps a tradeoff.

Click to expand...

It seems to me that perfection is not possible. A lens can be corrected pretty well at a given focused distance but, as various elements move in relation to each other some of the corrections might change very slightly. The designer might have eliminated them in theory but engineering tolerances will inevitably mean each example of the finished lens is slightly different. Such differences will mean that the lens deviates slightly from the original design created by the CAD system.



Any stabilisation system that works by moving lenses elements is going to affect the path of light through the lens. Effectively undoing a small part of one or more corrections. Sensor shift image stabilisation doesn’t affect the relationship between lens elements and thus shouldn’t affect the image in the same way as lens based stabilisation.

There is, of course, another factor which neither lens design nor stabilisation can eliminate, obstructions between subject and lens. I think we have established that the “artefact” in the image posted at the start of this thread results from something in the foreground. Almost impossible to eliminate with ground level subjects in the natural environment.



It has been instructive to learn of the causes of some of the characteristics we see in out of focus areas of our photographs. In most cases these things are outside our control but sometimes a small change of viewpoint can make a difference, equally often such a change is impossible.



Ultimately what I will take away from this discussion is that we need to eliminate our own contribution to the “problem” before considering the possibility of an equipment issue. Something that isn’t easily done and sometimes not doing so can lead to greater understanding of what is going on.

SKersting66 said:

That's why the normal VR mode recenters the VR elements just before the image is taken.

Click to expand...

That's to increase the VR efficiency,

I noticed this "artifact" on the 50mm f1.8, VR was enabled and set to normal. Processed the raw and cropped with Photolab. No editing was done.

TOF guy said:

SKersting66 said:

That's why the normal VR mode recenters the VR elements just before the image is taken.

Click to expand...

That's to increase the VR efficiency,

Click to expand...

How does recentering increase "efficiency?" It only needs to recenter if it is reaching a travel limit; and even then it doesn't need to recenter for every image. Recentering and reestablishing VR just before the image is taken is the opposite of "efficiency."

If you mean effectivity, as in optimal image quality, then yes...

SKersting66 said:

TOF guy said:

SKersting66 said:

That's why the normal VR mode recenters the VR elements just before the image is taken.

Click to expand...

That's to increase the VR efficiency,

Click to expand...

How does recentering increase "efficiency?" It only needs to recenter if it is reaching a travel limit; and even then it doesn't need to recenter for every image. Recentering and reestablishing VR just before the image is taken is the opposite of "efficiency."

Click to expand...

Maybe not? ...as stating the exposure process when the VR element is already at or near a limit can result in a loss/degradation of the ability of the VR element to be able to mitigate camera shake in that shot.

Starting the exposure with the VR element re-centered, means it can maximize the ability to achieve the goal (mitigate camera shake...VR) with little to no wasted effort or energy (efficiency).

SKersting66 said:

If you mean effectivity, as in optimal image quality, then yes...

Click to expand...

Yes...with regards to "optimal image quality" in regards to minimizing the affects of camera shake

--
My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

MrKazador31 said:

I noticed this "artifact" on the 50mm f1.8, VR was enabled and set to normal.

Click to expand...

I presume you mean you did not notice the out of focus vegetation between your intended subject and the camera, to some extend negatively affecting the image on the right hand side.

Ways to avoid the problem include carefully checking what is between you and the intended subject, possibly in combination with using a camera button to stop down to f16 as then obstructions are brought into sharper focus.

As the 50 mm f1.8 does not have in lens VR, I presume you mean you were using IBIS

--
Leonard Shepherd
In lots of ways good photography is similar to learning to play a piano - it takes practice to develop skill in either activity.

Mako2011 said:

Starting the exposure with the VR element re-centered, means it can maximize the ability to achieve the goal (mitigate camera shake...VR) with little to no wasted effort or energy (efficiency).

Click to expand...

What effort/energy is wasted if the VR elements are off center? What penalty does that impose on the voice coil motors?
Normal mode recenters before the image is taken for optimal image quality (corrections/sharpness); the negative is that the viewfinder jumps significantly, composition shifts, and even focus can shift some. It also has to stop VR movements momentarily while the motors are recentering. It is best suited for more stationary situations where those factors are less significant.
If anything, the recentering is wasted effort/energy strictly in terms of stabilization.


Sport mode does not recenter before the image is taken; the negative is less optimal image quality; especially for things that are defocused some (where aberrations become more apparent anyway). It is best suited to situations where there is a lot of movement... which would result in large image shifts (viewfinder jumps/composition changes).

--
https://www.flickr.com/skersting

SKersting66 said:

Mako2011 said:

Starting the exposure with the VR element re-centered, means it can maximize the ability to achieve the goal (mitigate camera shake...VR) with little to no wasted effort or energy (efficiency).

Click to expand...

What effort/energy is wasted if the VR elements are off center? What penalty does that impose on the voice coil motors?

Click to expand...

If the exposure starts near/at a VR element limit...then the camera moves in a direction that exceeds that limit...then the element wil/is forced to re-center. During the re centering...VR is no longer able to mitigate camera shake. If the VR element starts from a re-centered position at the start of the exposure...then chances are good it can mitigate camera shake during the entire exposure and those voice coil motors have less work to do (unlikely to require a re-center during exposure)

SKersting66 said:

Normal mode recenters before the image is taken for optimal image quality (corrections/sharpness); the negative is that the viewfinder jumps significantly,

Click to expand...

Which has no impact on image quality with regards to camera shake...and may improve the ability of VR to mitigate camera shake during the exposure if limits aren't reached during that exposure

SKersting66 said:

composition shifts, and even focus can shift some

Click to expand...

That can happen whether or not VR is on or off. Note that focus lock is generally unaffected by a shift to re-center the VR element if the subject is still in view. Also rare even in the case of a DSLR (mirror raises) unless the photographer has a habit of leaning in during shutter press.

SKersting66 said:

If anything, the recentering is wasted effort/energy strictly in terms of stabilization.

Click to expand...

Not with regards to total ability to mitigate camera shake during the entire exposure

SKersting66 said:

Sport mode does not recenter before the image is taken; the negative is less optimal image quality;

Click to expand...

Yes...as was pointed out, re-centering can improve IQ with regards to camera shake blur.

michaeladawson said:

I don't examine many photos from other brands. So I don't know if they do the same thing. But I am coming close to throwing in the towel on Nikon's VR. This has been going on for many years. I'm talking about Nikon's nervous double and triple ghosting of background elements.

Here's my latest example of this. Shot with the Z8 and 800mm f/6.3. VR is in sport mode. The Z8 is on a monopod. This is a crop to turn this from the original landscape to a portrait orientation. Otherwise it's close to full size.

That long stripe of mangled VR work on the left size mostly relegates this photo to the garbage heap. There's other spots in the photo where Nikon's VR work is evident.

I am just getting tired of just about all telephoto shots that have grasses or reeds exhibiting these VR artifacts.

Click to expand...

I wonder whether you've taken several photos shortly after each other, with and without VR (lens VR and in-body VR on/off in 4 combinations), to confirm that the phenomenon is actually caused by the VR system?

The differenct opinions between the two of you are not completely wrong in either case. Hard to say who has the stronger case, although I personally thing SKersting66 does.

You wouldn't intentionally buy and then keep a de-centered lens. But that is essentially what is happening when VR has moved lens elements around for the sake of mitigating camera shake. But that is the trade-off you accept.

What is not mentioned by either of you is what mode the camera is in. AF-S vs AF-C, and single frame vs. multi-frame drive mode. I'm not sure if AF-S vs. AF-C enters into the debate, maybe not. But single vs. multi-frame drive mode certainly does.

In the case of AF-S and single frame shot mode their is no need to recenter the frame when you press the shutter button if the reason is that the lens elements may have reached the end of their travel. The camera has a stable image and should simply take the shot and then recenter. Anything else and you introduce delay between when you trip the shutter and when the shot is taken. To me, the only logical reason to recenter first in this case is to optimize image quality by re-centering the image so that the lens elements are less out of alignment.

In multi-frame drive mode I can see a case to be made for re-centering the lens elements if they have neared the end of their allowed travel. The shooter may be taking a long burst and it would probably be better if the camera recentered the image at the start of the burst rather than in the middle of it. Or worse, run out of travel.

michaeladawson said:

I don't examine many photos from other brands. So I don't know if they do the same thing. But I am coming close to throwing in the towel on Nikon's VR. This has been going on for many years. I'm talking about Nikon's nervous double and triple ghosting of background elements.

Here's my latest example of this. Shot with the Z8 and 800mm f/6.3. VR is in sport mode. The Z8 is on a monopod. This is a crop to turn this from the original landscape to a portrait orientation. Otherwise it's close to full size.

That long stripe of mangled VR work on the left size mostly relegates this photo to the garbage heap. There's other spots in the photo where Nikon's VR work is evident.

I am just getting tired of just about all telephoto shots that have grasses or reeds exhibiting these VR artifacts.

Click to expand...

all my VR problems went away when i switched to sport VR mode in camera.

--