• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Just Announced: Nikon D800 & D800E

Cem_Usakligil

Well-known member
The new DSLR from Nikon has just been announced. It is the D800 and it's sibling D800E. The camera has a full frame sensor with a 36.3 megapixel CMOS sensor and an EXPEED 3 imaging processor. Of course, lots of improvements for filming as usual nowadays. The D800E is identical, except for the lack of an anti aliasing filter. This will surely result in a lot of chatter across the Internet.

A couple of items worth mentioning:

4.88μm pixels: smaller pixels as a result of the higher resolution. I am curious what the resulting DR and IQ will be like. The top ISO is set at 6400 (without using the overdrive to 25600), which may be telling. Also, what about the diffraction limit? My initial calculation shows that it becomes diffraction limited just above f5.6 (CoC being 1.5 times that of the pixel pitch, that is).

51 focus points, just like in the D4.

Exposure compensation: -5 to +5 EV in increments of 1/3, 1/2 or 1 EV and exposure bracketing: 2 to 9 frames in steps of 1/3, 1/2, 2/3 or 1 EV. (NB: this is a pet peeve of mine with the 5DII, why won't Canon give us this? It is just a matter of firmware upgrade. I know that they are trying to differentiate the market between the 1d series and the 5d series, but this is simply not a reason for enthusiasts like me to buy the 1d anyway.)


Official Nikon link.

Here is a hands-on preview from dpreview.com.

Here are the sample images on Nikon's website.
 

Andrew Stannard

pro member
Certainly look like an interesting camera, and will be fun to see all the sample images and chatter spread across the internet.

I wondered about the diffraction limit, but hadn't calculated it - thanks for doing so.

I reckon on my 5D MkII I can start to see the effects at anything beyond f11, so will be interesting to see how easy it is to extract all of those 36MPix in a deep DOF situation without a tilt-shift lens.


Regards,
ANdrew.
 

Doug Kerr

Well-known member
Hi, Cem,

Thanks for the nice summary of these new machines.

My initial calculation shows that it becomes diffraction limited just above f5.6 (CoC being 1.5 times that of the pixel pitch, that is).
Indeed.

To be a bit more finicky about notation (no criticism of your report, Cem), assuming a wavelength of about 550 nm, at f/5.6 the diameter of the "central spot" of the Airy pattern produced by diffraction (that is, the pattern out to its first minimum, which we might think of as a circle of confusion) would be about 1.5 times the pixel pitch.

Sort of gives one pause, doesn't it!

The "1.5 times the pixel pitch criterion" can be thought of as based on equating Rayleigh's criterion for resolution in the face of diffraction to an inherent resolution of the sensor of about 1.5 times the pixel pitch, as would be suggested by a Kell factor of about 0.7.

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Cem,

The D800E is identical, except for the lack of an anti aliasing filter.

It is curious that the language of the Nikon announcement (as presented on opf) speaks of the "anti-aliasing properties of the OLPF [optical low-pass filter]" being "canceled" on the D800E.

This may just be a "manner of speaking", perhaps involving translation from the Japanese, or it may mean something a little different than just "eliminating" the filter.

Best regards,

Doug
 
Certainly look like an interesting camera, and will be fun to see all the sample images and chatter spread across the internet.

I wondered about the diffraction limit, but hadn't calculated it - thanks for doing so.

I reckon on my 5D MkII I can start to see the effects at anything beyond f11, so will be interesting to see how easy it is to extract all of those 36MPix in a deep DOF situation without a tilt-shift lens.

Hi Andrew,

It'll surely generate a lot of chatter. It's an interesting direction towards more pixels, where Canon sofar seem to stick to the current status, and focus more on video, for the professional range anyway.

The diffraction of course only plays a role at the per-pixel level, when one indeed uses the additional pixels for larger output or more cropping. The amount of diffraction doesn't change because of a different sensor, but it will be sub-sampled at a higher density.

The diffraction effect on the individual pixels will start to manifest itself when the radius of the diffraction pattern exceeds roughly 1.5x the sensel pitch. With a sensel pitch in the order of 4.88 micron, that would mean that at f/5.6 for green light the microdetail will begin to suffer at the pixel level. That might be somewhat welcome for the version without AA-filter, and it will be interesting to see how that works out in practice.

One useful aspect of the denser sampling of the diffraction 'limited' image is that it can be restored relatively well if one applies a deconvolution sharpening with a good model of the PSF.

What's also interesting is to see how the smaller sensels (and thus smaller well depth) affect the Dynamic Range. With sensels this small, I expect the storage capacity to be 42800 electrons per sensel at best. In that case, the read noise should be better than 11 - 21 electrons for a dynamic range (engineering definition) of 12 to 11 stops. We'll have to see what materializes when the camera actually becomes available and can be analysed by serious users and/or by DxOmark.com .

Cheers,
Bart
 
Hi, Cem,



It is curious that the language of the Nikon announcement (as presented on opf) speaks of the "anti-aliasing properties of the OLPF [optical low-pass filter]" being "canceled" on the D800E.

This may just be a "manner of speaking", perhaps involving translation from the Japanese, or it may mean something a little different than just "eliminating" the filter.

Hi Doug,

It's indeed a strange way of descibing it. I expect it means that the OLPF is eliminated from the usual IR-filter+OLPF combination. It's going to be interesting to see how they will cope with the color aliasing that will exceed the luminance aliasing due to the different sampling densities, and if that is already somewhat addressed in the data that finally gets stored in the Raw data. How Raw is Raw is not a new question with Nikon's NEFs.

Cheers,
Bart
 

Jerome Marot

Well-known member
"Diffraction limit", "dynamic range", "well size", etc...

I'd like to remind everyone that we have had sensels that small for some time now on smaller sensors, in particular the µ4/3 kind but also the D7000. While measurements show that indeed degradation of the MTF happens over f/5.6, in practical photographic usage f/11 looks very much the same as f/5.6.

Of course, compact or video cameras use even smaller sensels. They often use a neutral gray filter in place of a diaphragm because the influence of diffraction quickly shows. But we are not there yet.
 

Asher Kelman

OPF Owner/Editor-in-Chief
"Diffraction limit", "dynamic range", "well size", etc...

I'd like to remind everyone that we have had sensels that small for some time now on smaller sensors, in particular the µ4/3 kind but also the D7000. While measurements show that indeed degradation of the MTF happens over f/5.6, in practical photographic usage f/11 looks very much the same as f/5.6.

Of course, compact or video cameras use even smaller sensels. They often use a neutral gray filter in place of a diaphragm because the influence of diffraction quickly shows. But we are not there yet.

Jerome,

Indeed! Bart has repeatedly taught us this and even given us tables to make it easy for us to check where we are photographing with reference to resolving the blur caused by diffraction. We should consider that the degradation we see over the diffraction limits suggested merely reveal the waves each focus point is blurred by. Those same waves landing on a much larger sensel are simply read as on level of brightness. It's amusing to see cameras with 7 micron sensels used at f16 or f 22!

Asher
 

Asher Kelman

OPF Owner/Editor-in-Chief
If the D800 can support binning, the use of say 4 adjacent pixels to make giant ones, then photographers would be able to have a choice of extended dynamic range at the expense of resolution. So what difference is there between in-camera binning and downsizing resolution in post-processing RAW files?

Asher
 

Nigel Allan

Member
Wow, all this scientific analysis is way over my head. I just can't wait to play with it and see if I like taking pictures with it.

I have been waiting for a while to see what they would follow the D700 with as I was interested in a full frame camera so I could get their magic trio of FX lenses (starting with 24-70mm f/2.8 etc) but without the size and bulk of a D3/D4...

So the bottom line is this: Was it worth the wait? Is it demonstrably better than a D700? It certainly sounds cutting edge and the 36 mp sensor is intriguing. I'm just pleased that Nikon is showing some cahoonas (sp?) again.
 

Jerome Marot

Well-known member

Cojones.

Cajones means drawers. It is also a kind of wooden box to sit on and use as a drum.

To come back to the D800: 36 Mpix is not such a big deal as the Internet frenzy would make one believe. In actual use and print, the difference with 24 Mpix is tiny.

People do not realize that the difference in linear resolution (what counts to the eye) vary like the square root of the percentage increase. If one doubles the number of pixels, linear resolution increases 40% (square root of 2).To be make an obvious difference, one has to quadruple the number of pixels: the difference between 6 and 24 Mpix is indeed obvious, between 12 and 24 less so.

36 is 50% more pixels than 24, or an increase in linear resolution of 22%. Good to take, but nothing to be mad about.

But it may make a difference for people shooting for stock (e.g. Yuri Arcurs...), since resolution is asked for in their business (and they will probably get the version without AA filter).
 

Nigel Allan

Member
Cojones.

Cajones means drawers. It is also a kind of wooden box to sit on and use as a drum.

To come back to the D800: 36 Mpix is not such a big deal as the Internet frenzy would make one believe. In actual use and print, the difference with 24 Mpix is tiny.

People do not realize that the difference in linear resolution (what counts to the eye) vary like the square root of the percentage increase. If one doubles the number of pixels, linear resolution increases 40% (square root of 2).To be make an obvious difference, one has to quadruple the number of pixels: the difference between 6 and 24 Mpix is indeed obvious, between 12 and 24 less so.

36 is 50% more pixels than 24, or an increase in linear resolution of 22%. Good to take, but nothing to be mad about.

But it may make a difference for people shooting for stock (e.g. Yuri Arcurs...), since resolution is asked for in their business (and they will probably get the version without AA filter).


I realize that the prorated difference between 24 and 36mp might not be huge and maybe Nikon have gone for overkill here as a kind of headline grabbing PR exercise so people start talking about them. I also realize that 36mp may mean storing and processing enormous files with all that entails and that since the majority of pictures these days never get printed and only viewed on digital devices this may actually be far more than anyone really needs or could realistically use.

My interest (nay curiosity) lies in the fact I read somewhere (I forget where right now but an authoritative photography book) that the digital equivalent of a frame of 35mm film is about 24/25 million pixels in density. If this is accurate (and I realize there are other factors) then anything that is full frame and over 24 mp would be indistinguishable from the analogue images stored on film since the eye would not be able to tell the difference at this density. Personally I am not a fan of most digital pictures but accept and appreciate their convenience. I much prefer the look and feel of film. Now someone is probably going to say that my eye can't tell the difference and I don't know what I am talking about, but actually there is a 'feel' I get from viewing analogue film images that I do not get from digital, and no I can't describe it :)

So if 24/25mp FF is the crossover point for the digital boundary 'blurring' with the analogue one, this would be interesting to me because then I have something which might give film like resolution/images with digital convenience. ...and then again I might say so what and opt for convenience and speed :) since the likelihood of printing these properly and framing them is very very small indeed.

I still think it is great that Nikon chose to do something to grab headlines again, even if they might have gone too far overboard
 

Cem_Usakligil

Well-known member
Hi Nigel,

....
My interest (nay curiosity) lies in the fact I read somewhere (I forget where right now but an authoritative photography book) that the digital equivalent of a frame of 35mm film is about 24/25 million pixels in density. If this is accurate (and I realize there are other factors) then anything that is full frame and over 24 mp would be indistinguishable from the analogue images stored on film since the eye would not be able to tell the difference at this density. ...
This is not correct. Whether or not your eyes will be able to tell the difference is dependent on how big is your output and the viewing distance; all other factors being equal.
 

Doug Kerr

Well-known member
Hi, Jerome,

Cojones.

Cajones means drawers. It is also a kind of wooden box to sit on and use as a drum.

My error. Thanks for the correction!

Not sure how that happened - I knew better. My orthographic skill seems to be in a little decline these days.

Best regards,

Doug
 

Jerome Marot

Well-known member
My interest (nay curiosity) lies in the fact I read somewhere (I forget where right now but an authoritative photography book) that the digital equivalent of a frame of 35mm film is about 24/25 million pixels in density.

Probably closer to 12 Mpix for most color films.

For what it is worth, you'll find here a comparison I made a few years ago between 24 mpix and fuji color film scanned on a Nikon Coolscan. The full resolution is available by checking "all sizes" on each picture page.
 

Asher Kelman

OPF Owner/Editor-in-Chief
Cojones.

Cajones means drawers. It is also a kind of wooden box to sit on and use as a drum.

To come back to the D800: 36 Mpix is not such a big deal as the Internet frenzy would make one believe. In actual use and print, the difference with 24 Mpix is tiny.

People do not realize that the difference in linear resolution (what counts to the eye) vary like the square root of the percentage increase. If one doubles the number of pixels, linear resolution increases 40% (square root of 2).To be make an obvious difference, one has to quadruple the number of pixels: the difference between 6 and 24 Mpix is indeed obvious, between 12 and 24 less so.

36 is 50% more pixels than 24, or an increase in linear resolution of 22%. Good to take, but nothing to be mad about.

But it may make a difference for people shooting for stock (e.g. Yuri Arcurs...), since resolution is asked for in their business (and they will probably get the version without AA filter).

What that 22% more pixels means in practice is that one can project an image 22% wider on a wall and still get ~ the same sense of sharpness viewed from the same distance. So it allows a larger version of that image viewed under the same conditions.

Asher
 

Asher Kelman

OPF Owner/Editor-in-Chief
There are other questions such as the nature of gradations of dynamic range, noise, color and shade. Are these also maintained with the same quality as the pixels get smaller? With the Sony NEX 5N and the Sony NEX 7, going from 16 MP to 24 MP results in a stunning picture, perhaps indistinguishable from that made with a Canon 5D Mark II. The level of noise is, unfortunately, increased one stop, compared to it's sibling the NEX 5N, so that above ISO 800 some corrections are welcome and above 1600 are certainly needed, whereas with the NEX 5N, the noise is one stop less obvious. Still, if we take the stellar performance of the APS-C sized sensor, Sony 24 MP NEX 7 ~ matching the full frame 5DII, then I see no reason why Nikon can't also show brilliance with these smaller pixels in a 36 MP full frame camera. Yes, there will be noise issues at higher ISO, but correctable to a considerable extent.

Asher
 

Jerome Marot

Well-known member
What that 22% more pixels means in practice is that one can project an image 22% wider on a wall and still get ~ the same sense of sharpness viewed from the same distance. So it allows a larger version of that image viewed under the same conditions.

Obviously. But what I meant is that if you print the two images at the same size, you won't see much difference.

Still, if we take the stellar performance of the APS-C sized sensor, Sony 24 MP NEX 7 ~ matching the full frame 5DII, then I see no reason why Nikon can't also show brilliance with these smaller pixels in a 36 MP full frame camera.

Especially when one considers that the D800 pixels are of the same size as the NEX 5n pixels. The NEX 7 pixels are smaller.
 

Wendy Thurman

New member
I've been following the discussions about the D800/D800E with interest. I'm a D3x owner and it's a bit painful! It's clear that Nikon is aggressively attempting to reclaim the "king of the SLR" title they held during the film days and that's a good thing but the fact that I could buy two of these cameras with money left over for what I paid for the D3x makes me grumble a little. Well, that's technology- it advances. I'm not going to buy the new camera- shooting underwater, I'm locked in to the big camera for the time being as the housings represent considerable investments. Smaller is better underwater and the camera is getting serious looks on the wetpixel site- it will certainly be a success with the landscape, nature, and studio shooters as well.

I can't help but wonder what's next?!
 

Wendy Thurman

New member
I'm not going to buy the new camera

Never say never- I sold a Leica 50/1.4 that's been unused then promptly placed a pre-order for the E model at what amounts to a couple hundred dollars out of pocket. Should receive it in late April/early May, depending on where I am in the queue.
 
Many are complaining about the high MP count. As I see it, nikon has always chosen lower MP for a reason, so there must be reasoning behind the hike in MP.

Whether it be advancements in technology or marketing, I'll let you decide.

I'm putting my bet in the advancements in technology bin.


Speaking of the evolution of technology, has any one viewed or read Ray Kurzweil's views on the future? Interesting

http://transcendentman.com/
 

Asher Kelman

OPF Owner/Editor-in-Chief
Many are complaining about the high MP count. As I see it, nikon has always chosen lower MP for a reason, so there must be reasoning behind the hike in MP.

Whether it be advancements in technology or marketing, I'll let you decide.

I'm putting my bet in the advancements in technology bin.


Speaking of the evolution of technology, has any one viewed or read Ray Kurzweil's views on the future? Interesting

http://transcendentman.com/

Jake,

Companies go for sales and market share! This jump in MP is helpful in group pictures for weddings or in detail rich landscapes or copying art. Pixels are very useful id they are good. Given good lighting, the new camera will be well received.


520635.jpg


Frankly, this would mate perfectly with the Nikon's state of the art leader in superwide zoom, the 14-24mm. This would make arguably the finest ultrawide angle digital combination out, at least under $10,000!

Asher
 
Jake,

Companies go for sales and market share! This jump in MP is helpful in group pictures for weddings or in detail rich landscapes or copying art. Pixels are very useful id they are good. Given good lighting, the new camera will be well received.


520635.jpg


Frankly, this would mate perfectly with the Nikon's state of the art leader in superwide zoom, the 14-24mm. This would make arguably the finest ultrawide angle digital combination out, at least under $10,000!

Asher



As well as their wonderful tilt/shift lens! Someday.... Someday...
 

Greg Norrell

New member
Good analysis above. I would not have expected diffraction issues until smaller apertures. As primarily a landscape shooter, I have pre-ordered the D800E. I'm curious about differences between the two models when shooting video. I don't shoot video often, but may at times. Do you folks expect the E model to have have harder to correct video problems than the one with a traditional AA filter?
 

Nigel Allan

Member
If the D800 can support binning, the use of say 4 adjacent pixels to make giant ones, then photographers would be able to have a choice of extended dynamic range at the expense of resolution. So what difference is there between in-camera binning and downsizing resolution in post-processing RAW files?

Asher

for me this sounds like the most exciting concept, and I guess it is nothing more than a firmware issue to address several pixels instead of one.

If you could 'bin' and have low light capability, or 'not bin' and have high resolution, and then switch to a DX lens and have the focal range and equivalence of a 16mp APS c sensor, then you have a category killer. It satisfies all needs in one camera.

Brilliant idea. Is this just hypothetical or is this a reality?
 
Top