Nikon D800 is outed with pics and specs. 36 MPX

[Fine_Art]

#$%@^*&$ all those camera companies!

After eating ramen for 6 months and plunking down 8500 bucks on the Pentax 645d, and spending a few thousand more on lenses and filters
(I was saving for a Phase one back, thank god for small favors!) I was happily snapping away, only slightly perturbed when thoughts of the Pentax 645 25mm would slip into my head (digitally optimized read: really expensive!)  I had been convinced after waiting years that neither Canon nor Nikon would put a big sensor into one of their cameras, now this!

hmm, Nikon has that really cool 135mm f/2 DC lens... couple of primes...oh a 70-200mm f/2.8 you don't say...

Oh the horror, the horror....

lol

There will be a lot of medium format salespeople reading up on how to prepare ramen when D800 ships.

[John Camp]

The success of the NEX must not have remained un-noticed at Nikon. <snip>
- Develop yet another mount with a shorter distance for a rangefinder like design that would be very compact. This would require the development of 4 new fixed focal lenses (24, 35, 50 and 80mm). It would be able to accept F lenses with an adapter.
Considering the clear positioning of Nikon as a high end market leader and their ability to listen to market needs, option 3 sounds extremely likely. The only question is when and I would say 2013.

Cheers,
Bernard

Why a different mount? The F mount is pretty small, and the NEX lenses are pretty large. If they just produced a small mirrorless body, that matched the size of the NEX, they'd then have the same kind of awkward system the NEX is -- One that saves you almost no space when traveling with a system, that is virtually as heavy as a full-sized system (because the "small" part is such a minor piece of the system), but that already has all the required lenses in place.

Alternatively, they could produce a digital rangefinder that uses Nikon rangefinder lenses. Those would sell like hotcakes...for a while.

[BernardLanguillier]

Alternatively, they could produce a digital rangefinder that uses Nikon rangefinder lenses. Those would sell like hotcakes...for a while.

Which is the same thing, they would have to produce lenses for this system, right?

As far as i know it is not based on the F mount.

Cheers,
Bernard

[BJL]

Why a different mount? The F mount is pretty small ...

Depth, not diameter, is the main issue. This is the same reason that no 35mm format rangefinder has a mount anywhere near the roughly 46mm depth of F mount: to reduce body size and allow great flexibility in lens design by allowing the option of far lower back-focus distance, meaning rear lens elements far closer to the sensor. (Yes, this can be done with "near telecentric lens designs": back-focus distance and exit pupil height are quite different.) [Added: like the Nikon rangefinders, whose lenses would need a lens mount shallower the "F" to be usable.]

... the NEX lenses are pretty large.

On one hand, Sony has for some reasons done the worst job on lens downsizing amongst current mirrorless systems: look at micro Four Thirds for far better exploitation of the downsizing possibilities, and it is not only the 25% difference in format size. In particular, the collapsable and pancake lenses give a body depth than neither F mount nor Pentax's K mount (as in the K-01) can match: the front of some of those lenses sits closer to the sensor than the F or K bayonet does.

On the other hand, it could be that for many lenses, the front elements end up about the same distance from the focal plane anyway, so that a shallower lens mount then just means that the lens body is longer. So maybe the strategy of the Pentax K-01 can be made to work: for the lenses that do benefit from a low back-focus distance, have the rear elements protrude back into the empty space where the mirror used to be. If this can be done well enough, then having the same overall camera+lens depth might give some advantage to the "deeper body, shorter lens" approach, because you have more room in the body for stuff like external controls and battery space.


Added: If the system is designed to support existing lenses via an adaptor, the main argument against a new shallower lens mount comes to "I do not want to have to use an adaptor". This seems a rather weak reason to hamper future lens design options, since one could simply leave the adaptor permanently in place. Maybe I am just insufficiently concerned about backward compatibility, like my skepticism of the idea that the ideal format for interchangeable lens digital cameras is 36x24mm.

[Mulis Pictus]

In case you missed it on rumors site, there are samples at various ISO settings:

http://forum.xitek.com/forum-viewthread-tid-969280-extra-page%3D1-ordertype-2-t-1329013603.html

Higher ISO samples look good to me, the earlier samples on Russian site looked worse IMO.

[seanesopenko]

More 100% crops.  The D800 with the 70-200 at f/8, 1/15sec and iso 100.  I want to see some shots at F5.6.  There aren't too many out there.

[BernardLanguillier]

In case you missed it on rumors site, there are samples at various ISO settings:

http://forum.xitek.com/forum-viewthread-tid-969280-extra-page%3D1-ordertype-2-t-1329013603.html

Higher ISO samples look good to me, the earlier samples on Russian site looked worse IMO.

They look better than the D3x at pixel level.

Cheers,
Bernard

[DeeJay]

I find the samples really quite underwhelming. They just look like large 35mm dslr samples rather than something verging Medium Format.

They have a typical Nikon or Canon look It's OK but doesn't have the wow factor and still look a bit soft. My guess it's the lenses and I still think the files from the 18MP M9 look better.

As for the beginning of the end for MF? Not even close.

[uaiomex]

Nothing beats cubic inches for a quarter of a mile. For a GP circuit the story is quite different. 
Eduardo

[aaron]

I find the samples really quite underwhelming. They just look like large 35mm dslr samples rather than something verging Medium Format.

They have a typical Nikon or Canon look It's OK but doesn't have the wow factor and still look a bit soft. My guess it's the lenses and I still think the files from the 18MP M9 look better.

As for the beginning of the end for MF? Not even close.

Why does it surprise you that they look like large 35mm dslr samples? or have a typical Nikon look? 

They are.

[DeeJay]

Why does it surprise you that they look like large 35mm dslr samples? or have a typical Nikon look? 

They are.

I didn't say I was surprised. Infact I really was expecting that it would be the case when they announced the camera. I was hoping for more though, like the marketing did tend to suggest and I did think the 800E would be a bit better than the samples.

Some people are questioning wether this to be the end of medium format sales, however it leads me to hypothesise that current dSLR technology probably plateued a while back while medium format continues to leapfrog i.e. Phase One IQ180 compared to P65+.

[Christoph C. Feldhaim]

Lenses, lenses, lenses .....

[BJL]

... current dSLR technology probably plateued a while back while medium format continues to leapfrog i.e. Phase One IQ180 compared to P65+.

I must have missed something: in what sense does the 80MP IQ180 leap-frog the 60MP P65+, other than a roughly 15% increase in linear resolution due to the reduced pixel pitch? I do not recall any claims of improvements at the 100% viewing, "per pixel" level in observed IQ or in measurements of noise, well capacity, dynamic range, or such. Do IQ180 images look better in 100% on screen crop comparisons, which I imagine is how you are judging the D800? Because the obvious point has to be made again that even with equal "pixel quality" as judged in 100% on-screen comparisons, an increase in pixel count allows printing or displaying at equal size using higher PPI, or making a larger print at equal PPI, both of which should give an IQ improvement. In fact, my rule of thumb for sensor development aimed at high resolution needs is that the best path forward is to keep per-pixel quality about the same while reducing pixel size and increasing pixel count, thus improving resolution, shadow noise, subject brightness range handling, moiré, and so, in fair comparisons at equal viewing size.

My feeling is that DMF has recently reached a plateau equivalent to what Canon, Nikon and Sony reached some years earlier: the high end sensors are now at "full frame size" with respect to the lens systems. But to the extent that lens performance is likely to become the main plateau, one would expect 54x40mm MF to end up on a higher plateau than 36x24mm. (As Christoph just said above!)

[BernardLanguillier]

I didn't say I was surprised. Infact I really was expecting that it would be the case when they announced the camera. I was hoping for more though, like the marketing did tend to suggest and I did think the 800E would be a bit better than the samples.

Some people are questioning wether this to be the end of medium format sales, however it leads me to hypothesise that current dSLR technology probably plateued a while back while medium format continues to leapfrog i.e. Phase One IQ180 compared to P65+.

Hum... I would not underestimate the ability of Capture NX to mess up the sharpening of .nef conversions! It does a very good job colorwise but is very far behind the curve in terms of achievable sharpness due to Nikon's focus on the total lack of digital artifacts.

C1 Pro or Raw Developper may show things on a very different light.

I have in the end decided to get the D800 and should be available in Tokyo in April for a shoot out against any of the 40 mp class backs. :-) whatever comes out it should be good fun!

Cheers,
Bernard

[DeeJay]

I must have missed something: in what sense does the 80MP IQ180 leap-frog the 60MP P65+, other than a roughly 15% increase in linear resolution due to the reduced pixel pitch? I do not recall any claims of improvements at the 100% viewing, "per pixel" level in observed IQ or in measurements of noise, well capacity, dynamic range, or such. Do IQ180 images look better in 100% on screen crop comparisons, which I imagine is how you are judging the D800? Because the obvious point has to be made again that even with equal "pixel quality" as judged in 100% on-screen comparisons, an increase in pixel count allows printing or displaying at equal size using higher PPI, or making a larger print at equal PPI, both of which should give an IQ improvement. In fact, my rule of thumb for sensor development aimed at high resolution needs is that the best path forward is to keep per-pixel quality about the same while reducing pixel size and increasing pixel count, thus improving resolution, shadow noise, subject brightness range handling, moiré, and so, in fair comparisons at equal viewing size.

My feeling is that DMF has recently reached a plateau equivalent to what Canon, Nikon and Sony reached some years earlier: the high end sensors are now at "full frame size" with respect to the lens systems. But to the extent that lens performance is likely to become the main plateau, one would expect 54x40mm MF to end up on a higher plateau than 36x24mm. (As Christoph just said above!)

All accounts, from what i read, including here on Luminous Landscape is that the IQ 180 is better than the P65.

I've personally found the same improvement between a P65 and P45 as I've owned them. The difference I've found between the 65 and 45 is a very noticible improvement.

I'm not seeing that kind of change in 135 generations at all.

[DeeJay]

Hum... I would not underestimate the ability of Capture NX to mess up the sharpening of .nef conversions! It does a very good job colorwise but is very far behind the curve in terms of achievable sharpness due to Nikon's focus on the total lack of digital artifacts.

C1 Pro or Raw Developper may show things on a very different light.

I have in the end decided to get the D800 and should be available in Tokyo in April for a shoot out against any of the 40 mp class backs. :-) whatever comes out it should be good fun!

Cheers,
Bernard

Interesting. Well I do hold hope for dSLRs as I would much rather use them. I use my M9 more than my MF at the moment. I look forward to seeing your and anyones tests with the new generation of dSLRs and 40mp backs. But I don't believe the Nikon and Canon lenses are up to the task though.

[BJL]

All accounts, from what i read, including here on Luminous Landscape is that the IQ 180 is better than the P65.

I am sure that IQ180 IQ is better, in the only sort of comparison that I care about: at equal viewing size (say equal print size and viewing distance.) That is why I specifically made the distinction between this and the goofy practice of comparing digital cameras of different pixel counts on the basis on 100% crops, which is, I think, all that you have to go on with the D800.

So may I repeat my question about the meaning of "better":

I do not recall any claims of improvements at the 100% viewing, "per pixel" level in observed IQ or in measurements of noise, well capacity, dynamic range, or such.
Do IQ180 images look better than P65+ images in 100% on screen crop comparisons, or only in fair, equal size comparisons?


Edit: on your comparison of P45 to P65, that was of course an improvement in both sensor size and resolution , so no surprise that IQ improved there. On the other hand, the spec's and discussions I have read indicate that there was no improvement in per-pixel specs between the P45 and P65 sensors, so 100% crop viewing would probably show no improvement, and possibly even an IQ decline, due to greater enlargement of lens imperfections, focusing errors, and such.

[DeeJay]

apparently, yes.

The LL review is interesting and if memory serves. The richness of the file and the brilliance in highlight detail in particular is worth the step up. It was , I believe, compared to the difference between 4x5 and 10x8 film.

[DeeJay]

Edit: on your comparison of P45 to P65, that was of course an improvement in both sensor size and resolution , so no surprise that IQ improved there. On the other hand, the spec's and discussions I have read indicate that there was no improvement in per-pixel specs between the P45 and P65 sensors, so 100% crop viewing would probably show no improvement, and possibly even an IQ decline, due to greater enlargement of lens imperfections, focusing errors, and such.

This is contrary to what I have found. I left the P45 for the P65 because I found it was a lot better. It's the tonality, the colour, the resolution. The downside was the extra computing power needed and that it was starting to out resolve my V series Blad so I needed to change to the H which is what I'm using now.

Even when you see a jpg of a scanned Large Format Film transparency, small on the internet, a comparison between medium format or dslr small on the internet, you still can clearly see you are looking at something different.

Don't get me wrong I'm not bashing 35mm for the sake of it. I use my Leica most of the time these days but there is no denying what a 60MP back can provide.

[BJL]

I think we are agreeing that the relevant comparisons are of final images under equal viewing conditions, not 100% crops entailing different degrees of enlargement, but your responses are not clear. If so, I await such comparisons (of equal sized prints?) of the D800 to previous 36x24mm format efforts.

This certainly sounds like a same size comparison, not a 100% crop comparison:

Even when you see a jpg of a scanned Large Format Film transparency, small on the internet, a comparison between medium format or dslr small on the internet, you still can clearly see you are looking at something different.

This is contrary to what I have found. I left the P45 for the P65 because I found it was a lot better. It's the tonality, the colour, the resolution.

That does not contradict anything of what I said. In particular, I expect things like tonality and resolution to improve in equal viewing size comparisons if one increases the pixel count while roughly maintaining per per-pixel performance (or even letting it drop slightly, as has happened with some or all MF CCDs as the pixel pitch decreases). This is what Kodak and Dalsa have been doing for many years now with their CCDs, in particular in the sensor progression from P45 -> P65 -> IQ180.