D3 & D300 sensor - bokeh destructors?

It's about time that someone discussed this. Let me know when you decide which sensor has the least bad bokeh. In situations where these problems show up, you can take a perfectly acceptable slide and scan it. Then you get sensor bokeh from your scanner. I doubt that this is unique to Nikon products which is all I have used.

Carey,

 

>Can Shun or Arthur comment whether Active D-Lighting was on or off for each of their tests?

 

My Active-Dlighting was on at normal.

 

 

Shun,

 

>Arthur, before you spend time on more tests, I wonder what is so objectionable to the background redention in your D3 test shot (the one with the National Geographics flashlight? and tennis ball) in the first place?

 

That was my son's NG insect collection box for his field trips.

 

Anyway, I agree that the tests have, indeed, failed to deliver the objectionable bokeh which I was looking to show. No harm done, just a good discussion anyway and attracted more visits to the website. :)

 

I will be sure to carry both the D2X and D3 together in the event I see such a phenomenon again.

Arthur, if you are going to run the tests again, may I suggest that you don't move the tripod. Use the same ISO and shutter speed on both cameras. I would capture at f2, f4, and f8 so that you have samples at different aperture openings. When I ran the test yesterday, I got samples from f2.8 to f8 in one-stop increments.

Normally, a larger sensor will blur the background smoother than a crop sensor at the same angle of view. Lenses do have bokeh properties built in by their manufactures. Even though I shoot with a Canon 5D, I am surprised by your D3 results. Also surprised by how nice the D2x images are. I believe Nikon changed something in the processing.

 

Bob Atkins did a full vs crop on bokeh study (look at normal lens section).

 

http://www.bobatkins.com/photography/digital/eos_5D_vs_eos_40D.html

Start here:

 

http://www.bobatkins.com/photography/technical/digitaldof.html

 

This is the simple stuff. It gets even more complicated after that.

 

I shoot mostly Canon, and so I am no apologist for Nikon. There is simply no substance to the claim that the D3 is somehow damaging bokeh. An adequate response would take a volume and still be unintelligible to those who are making this claim. I am sorry to sound eliltist, but some of you are in way over your heads on this one.

 

--Lannie

My remark was not directed at Arthur, who simply raised the question.

 

--Lannie

Whata moke who broke a bokeh!

 

Justa poke, you know.

 

A jokeh.

Here's another test, no shifting of tripod. D2X and D3 (DX-crop turned ON).All I can say is that I still cannot reproduce the same type of dirty bokeh I got in the soccer shots above. These have specular highlights in bright afternoon sunshine.

<br>

<a href="http://d6d2h4gfvy8t8.cloudfront.net/7388401-lg.jpg"><img src="http://d6d2h4gfvy8t8.cloudfront.net/7388401-md.jpg"></a>

<br>D2X @f2 (Above)

<br>

<br>

<a href="http://d6d2h4gfvy8t8.cloudfront.net/7388413-lg.jpg"><img src="http://d6d2h4gfvy8t8.cloudfront.net/7388413-md.jpg"></a>

<br>D3 @f2 (Above)

<br>

<br>

<a href="http://d6d2h4gfvy8t8.cloudfront.net/7388409-lg.jpg"><img src="http://d6d2h4gfvy8t8.cloudfront.net/7388409-md.jpg"></a>

<br>D2X @f4.5 (Above)

<br>

<br>

<a href="http://d6d2h4gfvy8t8.cloudfront.net/7388414-lg.jpg"><img src="http://d6d2h4gfvy8t8.cloudfront.net/7388414-md.jpg"></a>

<br>D3 @f4.5 (Above)

<br>

<br>

<a href="http://d6d2h4gfvy8t8.cloudfront.net/7388411-lg.jpg"><img src="http://d6d2h4gfvy8t8.cloudfront.net/7388411-md.jpg"></a>

<br>D2X @f8 (Above)

<br>

<br>

<a href="http://d6d2h4gfvy8t8.cloudfront.net/7388417-lg.jpg"><img src="http://d6d2h4gfvy8t8.cloudfront.net/7388417-md.jpg"></a>

<br>D3 @f8 (Above)

<br>

Just to throw in this idea: if the grain density of films can affect the bokeh (e.g. Fujifilm Astia vs the old Kodak Gold 100), why can't the pixel density do likewise?

 

Additional, the onboard processing is different for every camera.

 

And, how aggressive AA filter is should also add to the blur soup, doesn't it?

Arthur, I was the first one who responded to this thread. The reason I asked you to provide examples from the very beginning was that I wanted you to find out for yourself that you are talking about a problem that doesn't exist.

 

You are getting there, but as long as you are still not convinced, just keep trying to reproduce/create the problem.

I'm certainly no expert, but I have a question/comment.

 

Looking the the football/soccer pictures... In the background the areas of similar colors are significantly broken up with different shades of color & brightness. In the lacrosse shots the background has large areas of similar color.

 

Could it be these differences in contrast and color are creating the poor bokeh effect?

 

In latest A/B pictures the OOF has large areas of similar contrast and color. In looking at all of the test shots the areas of OOF are quite uniform.

 

Shun's shots come closest to reproducing the football/soccer conditions, but appears to be generally uniform and darker except for a few minor highlighted areas.

 

Again, is the bad bokeh being caused by how the background elements are being broken up?

 

Thanks for the discussion, it's very informative.

In your latest round of tests, the plane of focus is not the same in the D2 and D3 shots--the focus on the D2 shots is a bit further away.

 

In any case, it doesn't appear that the bokeh in the D3 photos is any worse than in the D2 shots.

"Lenses do have bokeh properties built in by their manufactures."

 

Exsqueeze me?

"Just to throw in this idea: if the grain density of films can affect the bokeh (e.g. Fujifilm Astia vs the old Kodak Gold 100), why can't the pixel density do likewise?"

 

It can't. Where did you get the notion that film has any influence on bokeh?

Shun,

 

I must have been really lucky to not encounter any of the those bokeh-killing situations throughout these years spent on the D2X. Is that possible? I guess it is possible. I will continue to keep this in mind.

<i> Is that possible?</i><P>

It seems far more likely than the existence of some destructive property of the D3 sensor (which disappears when you control for other variables).

Try and look at the highlights of the lates testshots made by Arthur. All of the shots are showing some change in bokeh in that area.

 

And when you look at the original testshots, from the soccer match, there are no highlights in the background. While the "bad" shoots had highlight areas.

 

I'm sure that the patterns we see, is made of several factors. Highlight areas, with specific built in patterns (ex. highlight in between leafs - or light on small round rocks), combined with focus distance, f/stop, number of diaphragm blades and sensor crop factor (DX format only uses a cropped area of a full format lens) etc.

Among the latest 6 test shots, Arhtur's point of focus has shifted a bit and the compositions are not entirely the same. Pay attention to the trees in the foreground; the sharpest point has moved around between shots. That alone has some effects on the background blur. That was why I picked a flat street sign to focus on to prevent this type of problem.

 

But even though there are some small differences among the background blur, what is important is that no bokeh is "destructed."

 

Arthur's original lacrosse vs. D3 soccer images were captured under drastically different lighting conditions. Pay attention to the complete absence of shadows in the lacrosse image because it was an overcast day.

"Just to throw in this idea: if the grain density of films can affect the bokeh (e.g. Fujifilm Astia vs the old Kodak Gold 100), why can't the pixel density do likewise " Arthur Yeo

 

Nonsense, what has film got to do with it? Bokeh is an optical thing.

Arthur, regarding not being able to recreate the 'nasty' bokeh of your lacrosse images, your 'pine' bushes are not shiny enough. If you shoot some deciduous tress under the same lighting, you're bound to get that bad stuff again. The redo your test, and both cameras should appear bad.

 

Its not simply specular highlights, but the close proximity of those numerous highlights (like the attached photo of the pebbles) where the 'halos' are mixing that creates the weird nasty stuff.

 

Good luck!

Sampson

oops, bad bokeh in the soccer, not lacrosse. very nice pictures anyway. I love looking at the relative DOF in the grass, only because non of my lenses have large enough apertures to do this unless focused very near.