D3 & D300 sensor - bokeh destructors?

I thot it was only the D3 but looks like the D300 is also showing this

characteristic even though the pixel density if a lot higher than the D3. Is

this the sensor that's causing it or something else (like postprocessing issues) ?

 

Look at how the bokeh from the Nikkor 200/2VR is totally destroyed ...

 

http://www.nikoncafe.com/vforums/showthread.php?t=171497

 

(these are not my images, BTW).

If the bokeh were not "destroyed," how would that look like?

 

Does the D2X have the same problem? If not, can you post some examples with the 200mm/f2 on the D3 vs. D2X, for example.

Could you at least post some samples with a camera that doesn't have this problem?

I don't get it. Looks fine to me. Maybe you need to throw out your D3 and buy a leica? (you can mail the D3 to me, i'll pay shipping)

Bokeh is totally different on digi compared to film. Agree with Arthur, those look horrible for such a $$$ lens.

Was this posted on April 1, but somehow got delayed before appearing on the forum? I can't see a thing that's wrong in the bokeh in the shots. Could you explain what exactly you would like to be different?

You are looking at images that have been compressed and reduced for the web. With that type of re-sampling you can loose quite a bit of subtlety from an image. If the photographer has a habit of over-sharpening their images that can really mess with the blurry areas. There are many post processing techniques which might leave the blurry bits less attractive. I don't feel this is the fault of the sensor but rather a difference in artistic opinion between the photographer and the viewer.

Judging from the amount of posterization in the backgrounds, I'd go with John, post processing, JPEG conversion, and all the other cruft involved in "webification" mangled the images.

 

There's nothing in the bokeh that's an artifact of the D3 or D300, per se. Garrison's claims aside, the only thing digital cameras can do to bokeh is cause the OOF highlights to assume a slightly rectangular shape, and I've never seen that with any camera on an f2 lens, only occasionally with f1.4 lenses wide open, and most frequently with f1.2 lenses, or the f1.0 Canon 50mm that I borrowed with a 1Ds once. Anything else is post processing mangling, and you can do that to scanned film just as easily as to a DSLR shot.

Interesting discussion. I have only recently noticed the destructive aspect of the sharpening tool and consequently view my images at a much higher magnification

while using it to prevent toggling too far.

When I first raise your attention to this issue, some of you swept under the rug and said it was due to the lower pixel density of the D3 that was causing the tone and bokeh to lose its quality.

<br><br>

<a href="http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00PN4O">REF link</a>

<br><br>

Shun,

<p>

I cannot reproduce the lower quality bokeh at will: it is not consistent and I cannot place my finger on the exact situation that can cause it.

<br><br>

> Does the D2X have the same problem?

<p>

All I can say is that out in the field where sports are played in the the outdoors, the D2X consistently produce a smoother bokeh with this lens.

I am profoundly disturbed that on the D300 (with almost the same pixel density as the D2X), the bokeh can come out this way. It appears to be something Nikon did for the D3 & D300, not pixel density related.

<br><br>

> If the bokeh were not "destroyed," how would that look like?

<br><br>

Here are D2X examples. Click on the image itself to see larger versions.

<br><br>

<a href="http://farm4.static.flickr.com/3094/2463002960_bb3b5ea77c_o.jpg"><img src="http://farm4.static.flickr.com/3094/2463002960_20846ee74e.jpg"></a>

<br><br>

<a href="http://farm1.static.flickr.com/165/394809368_9d33773bf8_o.jpg"><img src="http://farm1.static.flickr.com/165/394809368_9d33773bf8.jpg"></a>

 

<br><br>

Here's one from the D3 when it is behaving well:

<br>

<a href="http://farm3.static.flickr.com/2084/2217215705_1f9ca9953c_o.jpg"><img src="http://farm3.static.flickr.com/2084/2217215705_f36efdc176.jpg"></a>

 

<br><br>

The bokeh in these 2 looked like a $400 kit zoom lens. Here are a couple from the D3 when it NOT behaving well:

<br>

<a href="http://farm4.static.flickr.com/3105/2444380986_8c5637f340_o.jpg"><img src="http://farm4.static.flickr.com/3105/2444380986_06f285e1f4.jpg"></a>

<br>

<a href="http://farm3.static.flickr.com/2152/2443575167_6ff912cbe0_o.jpg"><img src="http://farm3.static.flickr.com/2152/2443575167_c0e4d2cde4.jpg"></a>

I find, occasionally, the VR on the 70 - 200 can mess a little with the bokeh.

Arthur, if such a problem indeed exists, I don't see why you have difficulty to reproduce the problem.

 

Your examples represent a very unfair comparison. The D2X images were captured in an overcast day without a lot of very bright areas in the background. The D3 images were captured near noon time in a sunny day as the shadows are quite short.

 

An A/B test with the same lens and same aperture setting taken within minutes under identical conditions should illustrate those problems. I don't have a D3 nor the 200mm/f2; I am not arguing one way or another, but I'd like to see some demonstration of this problem.

Are the issues related to the aperture chosen?

"Your examples represent a very unfair comparison."

 

Shun, not sure what the lighting has to do with it. The images in question looks like, in the good ole days, the enlarger was bumped. The others have a smoother gradation between the shapes.

 

"Garrison's claims aside, the only thing digital cameras can do to bokeh is cause the OOF highlights to assume a slightly rectangular shape... Anything else is post processing mangling"

 

Surprised this went unchallenged, here.

"Shun, not sure what the lighting has to do with it"

 

To me, saying lighting has nothing to do with it is like saying light has nothing to do with photography. Whether good or bad bokeh, it's bound to be affected (i.e. more or less apparent) depending on the subject and the lighting when the image was taken. I agree with Shun, there's no way you can make a truly meaningful comparison unless the images are of the same subject and captured under the same, or very similar, lighting conditions.

Agree with Shun, the comparison is poor and the argument is basically "my feeling is that the bokeh is worse". In order to analyse this, a proper a/b test is needed. Surely if bad bokeh is such a problem then it can easily be reproduced in this kind of test, no?

So does the D2x never exhibit this phenomenon with the same lens?

"Bokeh" or the quality of bokeh is a term without clear definition. Adding to this the complexity of many factors we get unlimited numbers of interpretation and as many opinions.

For a given lens and camera (film or sensor type) the major factors are

 

1) light conditions and colors both of the main subject in focus and the background. (Why on earth would anybody think light conditions are not important here? Just shoot the same lens in harsh light with bright spots in the BG and in soft light with soft BG.))

 

2) The focal length in case of zoom lenses and the

 

3) aperture.

 

4) focus distance

 

5) DOF

 

Just to name the most important ones. There is a bag of many more I can think of like use of lens hood, RAW processor or in camera jpg processing and presentation of shots for the web.

 

You may be able to name more.

 

So unless we see a dedicated well thought out test setup or the experience of many thousand shots I doubt any conclusions.

I agree with some above that the difference in pictures above can be explained more easily with lighting than with anything else. All the factors Walt and others mention should be controlled for to perform a real test, but the lighting in the different examples is dramatically different, and probably enough on its own to explain the current question. I've seen that same effect with a variety of lenses on my humble d80. Very-high-contrast lighting in the background can potentially make a background very unpleasant.

Garrison, in Arthur's D3 samples, there are a lot more very bright, out-of-focus areas in the background. Under strong sunlight, those bright areas naturally look more disturbing, at least to me. That is why I think this is a highly unfair comparison.

 

I simply would like to see some proofs that this is indeed an issue and, if so, how serious this issue is.

Another vote for side-by-side test.

I think most here would agree that ideal bokeh is complete blur of the background, with

little to no edge definition on OoF speculars, ie, we dinnae wanna see th'aperture

shape!

 

Oh, and OP, the bokeh on that first shot (the one of the lacrosse players) makes me

cry a little. It's like the trees were made out of butter and velvet.

If this were truly the sensors of the D# and D300 wouldn't the bokeh problem be apparent on all lenses the exhibit good bokeh on other cameras?

 

To my eyes the brightly lit backgrounds are destroying the effect you are wanting.

"Under strong sunlight, those bright areas naturally look more disturbing, at least to me. "

 

Certainly, Shun. It's indeed horrible. And, I'm not saying you did, but don't assume I feel that lighting has little to do with bokehy wokehy. Judging overall bokeh, lighting and specular highlights and how it is managed, is as important as the f stop used. I'm simply stating here, even if in shade, on an over cast day, the background would still be unpleasant in these samples of Arthur's.

 

 

Discussions about bokeh often go anywhere. I often just skip bokehy conversations as there seems to be no base point or established standard for all to agree on and then deviate from with meaningful constructive discussions.

Arthur's subject line clearly associates the bad bokeh to the D3 and D300's sensor. Therefore, several of us are asking for a similar shot with a D2X and/or a film body to demonstrate that it is not a problem with those sesnors (including film), or at least the problem is not as severe.

 

We may or may not agree that the bokeh is "bad" with those combinations, but all of us need to look at the same A/B test samples.

Without even that, this is a totally meaningless discussion.

 

I wouldn't mind shooting some test samples myself, but I don't have a 200mm/f2.

A proper test for out of focus effects involves controls; one has a controlled settup and just changes one item; typically the lens.<BR><BR> This is how movie makers in the 1960's ran tests for out of focus effects. With an uncontrolled test typically the lighting swamps subtle variables. Photo.net is full of falsehoods about out of focus effects. Its not a new thing; its old; pre wws2. A lens that had a perfectly round iris can have a bad out of focus look; this is because these effects are an optical design issue. A poor diaphram shape may add weirdness; but its not a always key feature like preached by many. Out of focus effects are typically confusing to amateurs since they typically chase camera gear; and its not in most folks DNA to appreciate lighting; the KEY factor.<BR><BR> If the sensor is a variable; its a weak one at best. You might as well add moon phase; your girl friends toenail coloring of the day; what exposure number; or what type of card used.

"...but all of us need to look at the same A/B test samples. Without even that, this is a totally meaningless discussion."

 

Yep.

 

"If the sensor is a variable; its a weak one at best."

 

Many here have used the same lenses on film, 4 mp, 6mp, 12mp, and cmos and ccd cameras. This is also a huge problem and discussion for DOP's that have wanted to use their favorite lenses once on an Arri with 35mm but now have a tough time with HD.

 

The first bokey wokey distinction I came across when I went to digital was double edged lines instead of one line, like on a telephone pole wire.