Being fairly new to Digital I am a little confused over the following
statement that I found from a person who has handled a D800.

"The D800 will demand skilled handling technique, the 36 megapixels are extremely unforgiving in terms of motion blur even at 1/100 on an 85 mm lens. (or maybe I was shaking with excitement!!!)

Could someone elaborate on this so I can better understand what he is trying to say.
Thanks
Bob

So was the D7K but you can learn how to use it. I think what the guy was talking about is the higher MP cameras show more blur than the lower ones. So you may need to change your shooting style a little, slow down and try to be more steady. I was about to throw my D7000 in the river when I first got it. Now it's the sharpest camera I've ever owned. Looking forward to learning all over again with my new D800!!

Visit my Nikonians gallery.

Bob,

The staement the person is making is relative to the pixel size.

If you take a picture with the D800 and then say a D700 (or other FX camera) with the same amount of hand shaking (or subject movement) and then print both photos the same size, the amount of blur will be identical.
However, if you go pixel peeping, since the D800 has more (and thus smaller) pixels, the same amount of motion blur will effect more pixels.
Another way to think about it is if you go pixel peeping on a D800 you are looking at a much larger overall image than when pixel peeping on say a D700. Thus the blur looks worse. But again when compared between the same size images, it is the same.


Brian

An easy way to think about it is to think about a television screen.

Let's say that you don't have HiDef reception. In one room you have a small 12" tube tv. When you look at the picture you don't see much grain or detail. In the other room you have a new 55" LED Hidef TV receiving the same image. The image will look terrible on the new 55" tv because it expands every detail of the received image.

Well the 55" tv is like the 36mgpxls and the old 12" tv is like a 12 megapixel sensor. The larger one is "less forgiving" because it makes visible every imperfection in the signal received or the image projected onto the sensor by the lens.

I hope that makes sense.

Larry Miller, Annapolis, MD
D700/D800
http://www.pbase.com/lmmiller9
http://lmmillerphotography.smugmug.com/

Me too, and I've been at this a while. Pixels captures detail, I think what I'm learning is some feel smaller pixels can move less relative to details and that movement can be perceived as greater blur.

Here is an example..
If 4(2x2) large pixels are used to capture a feature (bugs eye, or speck of dust on a bugs eye), with the smaller pixels it takes 16 (4x4) for the same coverage, when the eye "moves" relative to camera shake, details can slide around the larger pixel 2x2 and show less blur than the same movement applied to 4x4? (really?)

My thoughts:
If the cameras moves there is blur, smaller pixels may show more of the movement (relative to 1 pixel doing the work of 4 smaller ones). Keep in mind the 4 are already sharper (capturing greater detail) than the 1, so it's a push.

Robland
Seattle Washington area
Insert "catch phrase" here

Visit my Nikonians gallery.

Kinda hard to follow, David.

I posted an experiment awhile back where I tapped the lens hood of a 300mm lens in the middle of a one second exposure. I used a 16mp camera and a 12mp camera for the test, both DX. I repeated the experiment 4 or 5 times with each body. The 16mp camera showed more smear than the 12mp. That's the only test I've ever seen on this issue. Everything else was conjecture.

Len

Visit my Nikonians gallery.

>The 16mp camera showed more smear than the 12mp.

On 100% crop - or printed on the same format without cropping?

This complains started far 2004, when 12MP DX D2X hit the market (equal to 27 MP FF). Now 16MP DX D7K (equal to 36 MP FX D800) is considered the sharest camera ever.

Vladimir
Russian Nikonian in Switzerland

Well this was all interesting, thanks for your input.
I guess we'll just have to wait and see when the d800 gets released and the images start appearing.

Bob

PS here's another question that has puzzled me since I got into "digital"
Could some tell me how many megapixel it would take to duplicate the
resolution of a modern 35 mm film camera ?

>how many megapixel it would take to duplicate the
>resolution of a modern 35 mm film camera?

Film camera has no resolution at all. Resolution depends on the lens and on the film. Since DSLRs use the same lenses, your question is in fact about comparative resolution of film and sensor. Films are different, and differently used.

24 MP and 14 bit color of D3X is more or less the same as professional color print film if used with Nikkor professional zooms (if film images are printed directly, with an enlarger).

To bypass quality of slides shot with Leica and scanned with a VERY expensive drum scanner at 20'000 dpi (resulting in 500 MP images), you need about 60 MP sensor with full 16 bit color and very high dynamic range. Why you need 10 times more MP scanning film than shooting directly on digital sensor? Because film image consists of irregular grains, eye does not see each of them, details are reconstructed in the brain from a mosaic of grains, so we have to produce an image of every grain on the secondary (scanned) image to deliver the same details, while on a primary (DSLR) image we can record the details themselves.

Middle format slides may be even better: no MF glass has linear resolving power like Leiz glass in lines per millimeter, but total amount of details captured with a bigger frame is higher, and these details are bigger and consist of bigger group of grains without need to enlarge this bigger frame that much as a smaller one, so you need a more affordable technology to scan MF slides (like Imacon). Big format slides can be even better, and can be scanned adequately even flatbed (something like Epson 750 is enough).

Vladimir
Russian Nikonian in Switzerland

>>how many megapixel it would take to duplicate the
>>resolution of a modern 35 mm film camera?
>
>Film camera has no resolution at all. Resolution depends on
>the lens and on the film. Since DSLRs use the same lenses,
>your question is in fact about comparative resolution of film
>and sensor. Films are different, and differently used.
>
True, but there are some analogies between digital pixels and film grains. Both are tiny fragments of a space that get exposed to light during taking a picture. One of the things that makes the comparison difficult is the irregular (by definition) nature of the film grain.

>24 MP and 14 bit color of D3X is more or less the same as
>professional color print film if used with Nikkor professional
>zooms (if film images are printed directly, with an
>enlarger).
According to Wikipedia the number of MP is slightly lower - 14MP (http://en.wikipedia.org/wiki/Digital_versus_film_photography), but again there are estimations.

Visit my Nikonians gallery.

My favourite Fuji Superia Reala delivers 125 lpm (not 125 black lines on every millimeter of white field, but 125 passes from white to black and from black to white, or 60 real lines). Good Nikkor lenses have about the same resolving power (60 real lpm). To record all this information even if lines are not parallel to photosites rows on the sensor, one need pixel density multiplied by square root of two - or 180 pixels per millimeter, for 24*36 frame it is 4320*6480 photosites, or 28 megapixels. Some Leica glass resolves better, and some special-purpose films (for example "KGB" MZ3L - extremely contrast B&W slides, ISO 3 - for microfilms) are capable to record this difference - so 60MP is real, if you can afford a half-ton, half-megabuck drum scanner.

Vladimir
Russian Nikonian in Switzerland

>>The 16mp camera showed more smear than the 12mp.
>
>On 100% crop - or printed on the same format without
>cropping?
>
>This complains started far 2004, when 12MP DX D2X hit the
>market (equal to 27 MP FF). Now 16MP DX D7K (equal to 36 MP FX
>D800) is considered the sharest camera ever.

It was easy to see at 100%.

I have photos in this thread. Look for the Red M&M's eye only photos near the end of the thread..

http://www.nikonians.org/forums/dcboard.php?az=show_topic&forum=329&topic_id=14940&mesg_id=14940&page=5

Len

Visit my Nikonians gallery.

>It was easy to see at 100%.

So, nothing to worry about - if with 36MP you make a mistake visible only on 100% magnification, you still can use the shot to print the same formats you normally print from 12MP (up to 60% magnification) without any visible loss. Q.E.D.

Vladimir
Russian Nikonian in Switzerland

This argument is all about magnification. The more you enlarge an image, the more visible things like motion blur will be. This only relates to pixel pitch in as much as people are buying higher pixel density cameras because they're hoping the extra pixels will allow them to enlarge (magnify) more.

Given identical technique, all other things (dynamic range, noise, etc.) being equal, the same size print from e.g. D800 & D700, the D800 print may show more detail due to the finer pixel pitch, but if poor technique (either camera movement, incorrect DOF, incorrect focus, excessive diffraction, etc.) is obscuring that detail, then the D800 could be reduced to being equal to D700. It won't be worse, because the same technique flaws are subject to the D700 as well. People who say the D800 will show worse have failed to account for another variable in their experiment. Most often this variable is magnification, by looking at 100% pixels on screen or print, they enlarge the higher pixel image considerably.

The smaller pixels may allow you to enlarge more by capturing more detail, but given that you will also be magnifying flaws in technique, this ultimately means higher pixel density will ultimately yield diminishing returns.

Are we there yet?

Given that f/8 is generally considered most lens' sweet spot, and also widely recognized as the de-facto general purpose aperture setting, and that's it's been reported (with math that goes over my head) that D7000 & D800 will be diffraction limited smaller than f/8.9, I'd say with this latest generation we're about at the ideal pixel density now.

Larry - a Bay Area Nikonian
My Nikonians gallery

www.tempered-light.com

Ok WOW, lets put it another way.
I have some B&W prints that I took in the early 70's with what at that time was the top dog Nikon. Nikon F2 with a 50mm 1.2 lens.
These photos show detail so amazing that I've never seen it beat by any mainstream digital camera.
I know a Hassablad or something like that with a digital back may do the same thing but I'm would like to limit this discussion to normal affordable camera's not 50K machines.
These pictures that I speak of were taken with Kodak Tri-X.

Bob

I bet D800e with the same 50/1.2 Nikkir will record all the details F2 has captured on Tri-X. If any of these cameras will move during the exposure, the same amount of information will be lost - only with D800 you are more likely to discover these imperfections, because it is much easier to see a fragment on screen with 100% magnification than to enlarge a negative to the same scale.

Vladimir
Russian Nikonian in Switzerland

Well I guess time will tell, With all the confusion that has happened
overseas I hope that the D800's will hit the shelves in a timely manner.
Judging by the number of people on this forum that have said they pre-ordered I hope to see some example of this marvelous camera soon.
Then the REAL discussions can begin

Bob

>Given that f/8 is generally considered most lens' sweet spot,
>and also widely recognized as the de-facto general purpose
>aperture setting, and that's it's been reported (with math
>that goes over my head) that D7000 & D800 will be
>diffraction limited smaller than f/8.9, I'd say with this
>latest generation we're about at the ideal pixel density now.

Taking into consideration that: (1) f/5.6 is as sweet as f/8; (2) modern 12MP 4/3" sensors (1/4 area, 1/2 linear) are perfectly usable at up to f/11 and next generation with 16MP is coming (Olympus OM-D); (3) the very best 35mm lenses may record up to 60MP equivalent details; and (4) demand for more and more megapixels from the market is strong even if not rational - we still can wait another one step up to 48, 54 or even 72 MP before diffraction at f/8 becomes intolerable. But I am ready to bet - no 100MP 24*36mm camera will ever be produced.

Vladimir
Russian Nikonian in Switzerland

>Taking into consideration that: (1) f/5.6 is as sweet as f/8;
>(2) modern 12MP 4/3" sensors (1/4 area, 1/2 linear) are
>perfectly usable at up to f/11 and next generation with 16MP
>is coming (Olympus OM-D); (3) the very best 35mm lenses may
>record up to 60MP equivalent details; and (4) demand for more
>and more megapixels from the market is strong even if not
>rational - we still can wait another one step up to 48, 54 or
>even 72 MP before diffraction at f/8 becomes intolerable. But
>I am ready to bet - no 100MP 24*36mm camera will ever be
>produced.

Ultimately Vlad you're spot-on. I did these tests years ago with my D300, and found that although diffraction was mathematically limited below f/11, the type of detail found in my real-world subjects didn't really fall apart until f/22, which meant I'd go to f/16-f/19 when necessary without worry, which covered the vast majority of landscape subjects anyway. I strongly believe you're better off with a diffracted in-focus subject than a theoretically sharp, out-of-focus subject (which doesn't exist in photography!) ...and if a tilt-shift lens isn't an option, focus stacking always is.

But yes I agree on the surface, diffraction limited below f/8.9 isn't really a limitation at all, and anyway the real-world limit is probably 2 stops further.

Larry - a Bay Area Nikonian
My Nikonians gallery

www.tempered-light.com