Shared publicly  - 
D800 Article Posted

I've posted my D800 introduction article at Let's try to keep the discussion about this camera here and not spreading through my other streams ;~)
Nikon D800 Introduction comments by Thom Hogan
Diego Maicu's profile photoSilvia Ganora's profile photo_ Fynali _'s profile photoArgho Sen's profile photo
the question i have is will there be a D4X and will it have more resolution yet? i might get the D800 anyway but a D4X with more resolution matters more.
I guess at some point Nikon will release a D800 with a D4 sensor (maybe they want to wait until they sell a bunch of D4s first?), but... why kill the D700? They could offer a very very cheap FX camera for those that don't want the latest or the greatest. Oh, and it's sad to see the D3S's sensor going away... Maybe a D700S with that sensor could make some sense? Well, I know I would buy such a camera... But maybe it's just me.
+Herb Chong A D4x seems a bit doubtful at the moment, doesn't it? I see two likely paths: (1) the D800 sensor replicates into a D4x, the D4 sensor replicates into a D800s (hmm, that's a naming problem, isn't it?); or (2) we don't get a D4x or D4 sensor in a smaller body.
Did anybody else tell all their starving artist friends to start looking for used D700s today?
Want: FX body for less than 1500 Euros. Wouldn't care about pixel count or other features. Well, maybe used D700's will start dropping in price come March or April.
Apart from not being able to afford it, I like the D800, btw. Can't think of anything I would change.
+Thom Hogan a D4X seems pretty unlikely to me too given the 36MP but there are advantages for me that i could use a D4X with both the same sensor but preferably one with even higher resolution. i have been budgeting for both a D3 and D3X replacements for a year now and will continue to set money aside until the tea leaves are a little more clear on D4X. however, i am likely to get the D800E anyway. i've always planned to keep my D3X and replace the hot mirror filter for a high resolution IR camera.

have you heard about a supposed problem where the D3 series uses an IR LED to monitor shutter speed and that leaks enough IR to cause hot spots on long IR captures? i'd hate to convert my D3X only to find that a 20 second exposure is fogged on one side. i wasn't going to investigate until i actually had a D3x replacement to order but since i am discussing D3X replacement now, i am curious if you have heard one way or another. i want to replace my Fuji IS-Pro with a much higher resolution IR camera.
+Christoph Breitkopf I want a Ferrari for less than US$50,000, too. Problem is that at current sensor costs, it won't happen on the new market. Sony tried underpricing FX bodies and they lost their product margin on them and didn't gain market share. They're not likely to repeat that mistake. US$2000 is a very aggressive price, one that doesn't really make a company money. Let me illustrate.

Price to customer: US$2000
Price to dealer: US$1700
Price to US subsidiary: US$1500
Cost of sensor alone: US$500
Expected end-of-life price: US$1500
Expected end-of-life dealer cost: US$1275
Expected end-of-life US subsidiary price: US$1150
Expected end-of-life sensor cost: still very close to US$500

That leaves US$1000 at introduction for all the other parts of the camera, the manufacturing of it, the return on R&D, and profit. It can be done, but the ROI starts to get very low. Too low to justify.

Meanwhile, volume of sales is going down in the DSLRs. So you have to be very careful that you don't actually fail to recoup R&D costs.
Great article.

I think the anger (we should never get angry over gear in the first place, unless it fails on us or doesn't do what it promises to do) is more that for some people, they're at the point of enough megapixels with 18 or 24 or whatnot. I think a better analogy may be that they want the handling and features of an M3, but are happy with the power of and want the fuel efficiency of a 335i (which is why BMW brings out the m sport packages; to bring the base 3 series closer to, though not quite up to par with, the M3).

It would be curious to see what the take rate of the D800 versus a hypothetical D800S with the D4 sensor and which cost, say, $500 or so cheaper.
Ok, that 1500 Euro was maybe too aggressive. My point was more that the D700 was at a different price point than the D800, and the D700 price point ("FX prosumer"?) is now vacant. I'd like to see it being served, too. I have to admit that the D800 is less expensive than expected - higher prices were floating around the net for some time. Given the specs, I can't argue with the price.
EDIT: on second thought, if I take your numbers for granted, a lower end model would be at most 500$ less. Tough decision to make.
+Rohith Thumati As I've noted several times, I've surveyed over 20,000 Nikon users in the last few years. Results have been reasonably consistent on what people wanted out of a D700 replacement: exactly half wanted a D700s (e.g. D3s type sensor), while the other half wanted a D700x (e.g. D3x type sensor). We got a D700x. Thus, almost by definition, Nikon has this problem of convincing half the updaters that the D800 is really what they want. I have my doubts that they'll manage to do so. They might get some of those D700s wishers, but not all of them.
For me, it all comes down to this: If I could go back in time, which photo would I rather have of the Hetch Hetchy Valley before it was flooded: A 12mp or a 36mp?
+Christoph Breitkopf Actually, you're incorrect. The D800 is EXACTLY at the D700 price point, despite inflation and currency collapse against the yen. You're looking at latter-in-life D700 pricing, not the price it was introduced at. Simply put, you're not a "buy the latest" kind of person. You need to wait, probably at least a half year, likely more, to see a lower price. The lowest price I expect to see a D800 at is exactly the same as a D700: US$2250. But the initial demand will have to completely clear before we'll see any movement towards that.
+Richard Swearinger If you're talking about documenting the valley as best as possible, neither. I'd probably put a telephoto on my 12mp D3s and run a Gigapan of 100 or more images. The only thing that a D800 would do is allow me to shoot 33 images instead of 100.
Another excellent article +Thom Hogan. Always a pleasure to know we can count on you to write the best analyses. Only thing that worries me: you make it sound like I could be interested in the D800 ;) (no, I don't need it)
For the D700 owners looking for a D700s, I wonder how many will simply keep their D700 and look toward supplementing their system with an Fuji X-Pro1 or Olympus OM-D.
+Michael Shaw i am surprised you get reliable enough results with your K-5 for bird photography. i have never had reliable enough results on easier subjects to bother trying anymore with my Pentax bodies. it's why i bought into Nikon in the first place 8 years ago. i use my K-5 for some landscape but mostly for portrait work because i like the lens renderings better than Nikon's offerings but the AF has always forced me to take 2-3 times as many shots to end up with an acceptable number of in focus ones. in emergencies, i have had to use my K-5 for some wildlife shooting and the in-focus rate is always too low to choose to do it.
+Thom Hogan BTW, i saw that Nikon announced a 1 month delay on the D4 shipments last night. they are targeting March 15 now because pre-orders have overwhelmed them. i suspect that the D800 ship schedule has already accounted for latent demand but Nikon has underestimated for its higher end bodies several times now.
+Thomas Paris Oh come on Thomas, you know you want it and probably have already ordered it ;~).

But to your point: I think a lot of the early grumbling as the leaks started was that the D800 wasn't a worthy successor to the D700. Most of those folk were focusing on the sensor, though. Here's an experiment, take the sensor out of the equation (substitute the D700, D3s, or D4 sensor as your choice) and see if the camera is a worthy successor to the D700. My answer would be yes. There are a lot of improvements, pretty much everywhere you look, plus it's a darned good video camera where the D700 wasn't a video camera at all.

So that all brings you back to the sensor. 36mp is more than anyone asked for. 36mp without AA is more than anyone could imagine. Some will love that simply because more is better. A number will run into the shot discipline wall. Others will find out how different the corners of their lenses perform compared to center. Some will get stuck on diffraction or moire.

But the funny thing is, take out the FX bit. Now compare it to the D7000. Which is the better camera? Why, we're back to the D800 again. Hmm. Okay. VERRRRRY INTERESTING (but not dumb if you recognize the reference).
+Michael Shaw Really? The D800 has about the same pixel density as the D7000 and thus about the same number of pixels in DX mode (15mp+). The highest density APS is the Sony 24mp (and likely Nikon soon), but is that really enough more pixels to get excited about? The bigger issue I would guess is frame rate, not pixels.

You're not really hijacking the subject here, so no need to apologize. Everyone has to put the new camera announcement in the context of what they're using and what they could use, so as long as we keep that spine, everything's fair game.
Great insight. To tie into your previous article on DR, any comment on how the DR of the d800's sensor might look compared to that of the other FX bodies strictly on the numerical data?
I don't see how the D800 will kill the D700? The D700 will still be a good camera. Also, the introduction of a new camera has never effected the price of the previous version.
+Thom Hogan Nikon press conference at 04:24 GMT according to Bloomberg news release behind a paywall. i don't see a public release yet.
+Ikaika Arnado, +Antonio Rojilla I believe the new battery regulations in Japan forced the killing of the D700. I think there certainly will be a robust used market for the D700's. In fact, that could be a way to your 1,500 Euro FX camera, +Christoph Breitkopf :)

And thanks, Thom, for the info re: surveys. I've been reading your stuff for a while but hadn't realized that you've done that kind of research. I wonder if Nikon has even surveyed that many FX users :)
+Herb Chong Email me the URL behind the paywall. I don't see the article you refer to (I'm not sure I can get to all their stuff with my account, but I can get to most of it).
+Geoff Powell Given that it's apparently EXMOR based and only slightly smaller in photosite size than the D7000, you'd expect D7000 like DR.
+Ikaika Arnado D700 used quotes are going down as we speak. Where they'll end up, I don't know. Some of this seems to be panic selling.
+Rohith Thumati Well, they have surveyed that many FX users, because I've given them the results of most of my surveys ;~).
It's looking like my eventual D700 replacement will be a used D3s. I'm fine with that.
Shall I buy one D4 or two D800? If 10 frames a second is not important to me, is the D4 worth the investment? Or asked another way: What technological difference increases the price for the D4? If you, as a tech expert, could tell me what makes the D4 so much more expensive I might better know if I can live without that or not.
My thinking is the inverse of Herb's in that I can barely see where 100 lppm (36 MP on FX) is attainable compared to 80 lppm (24 MP on FX), so as far as resolution is concerned I don't see the D800 as a compelling upgrade to the sensors in the D3x, A900 and A850 sensors. Given the format, I can't imagine how many people will be able to take advantage of 48 MP, but Herb certainly could be exceptional in that regard. For me then, the price and features are the what is most compelling about the D800, although I welcome 36 MP over 24 MP.

With that in mind, I find the $3000 compelling given the features of the D800. I expect though that I can get an A900 in like-new condition (maybe even brand new) this summer for well under $2000. A Canon user is probably going to be looking at 5DII cameras for $1500. With Sony and Canon due to upgrade their A900 and 5DII, Nikon has really put the screws on them with the impressively featured and aggressively priced D800. If I didn't stand to lose money selling off my Sony/Minolta lenses and a flash, I would probably just sell my A850 and buy a D800; I might still do that, but it would mostly be for Live View, better AF, more DR, and better build quality -- and not as much for the increased resolution.

I'm going to be reading the reviews, and especially Thom's, with some interest in the coming months. The D800 announcement comes at an opportune time as I'm just now saving up for Mirex T/S adapters and just bought a Pentax 67 55/4 for landscape photography (BTW, that lens is a real beast), and the difference between $3000 and $4000 is enough to make me consider using those lenses on the D800 instead of the A850.
+Anthony Beach i think 100lppmm is attainable but by far fewer lenses than most people think, with far better technique than most people think, and only at wider apertures (f5.6 or so). i would for certain continue to shoot at f11-f13 because sometimes DOF is more important but in macro work, i can see doing a lot more focus stacking than before. also, a fair portion of my work is with lenses shot f1.4-f2 and there is no problem with diffraction at those apertures. lots of other problems but not diffraction.
"what exactly would be missing from my DSLR? Oh, yeah: communicating, programmable, modular (CPM)."
Thom, have you looked at Magic Lantern software? That would solve "programmable" part. The only problem is to make it work on anything but Canon :)
I don't see any D700's on my local craigslist. I'll throw up my D700 when the D800 becomes availiable. If I get a bite, I'll hawk it and get a D800. If not, it's fine. I can't wait.

I've been wanting to get into video eventually, although, I'm not in a rush. I think the D800 would make a great camera.

Although, I'd actually prefer a camera in the 20mp range with decent ISO, that shot in DNG, good video capibilites, and put out 16-bit files. That'd really suit my wants and needs.
Thanks for the info +Thom Hogan it is always thought provoking reading your stuff even gear reviews. My first reaction to 36Mp is yikes that's a lot to store. Then I think - more pixels would be nice some times. But then I bring my self back to reality - I don't really need them. I regularly print 12x18s now (since I switched to digital I have been taking more pictures and printing more pictures). Just the other day I printed a heavily cropped picture from my D700 and was very happy with the results. Others were happy too since it did win 1st place at my photo club competition.

PS by heavy I mean the final version had about 2/3 the pixels.
+Mike Heller My thought would be D800E since most of the time Moire is not a problem and when you hit it a bit of post processing can fix it up nicely.

Just my opinion as a landscape photographer...
+Ian Leslie i've been reading many conflicting statements about the ease of cleaning up moire. in landscape work, i can see it being fairly easy. in product work, i can see it being very difficult.
My initial reaction to the thought of a 36MP DSLR was that I had no need for one, and if I did need that sort of resolution then it would be better to go the medium format route.
I dont see the need to upgrade my D3s to a D4-there is not really much to gain for the sizeable cost, but a nice, cheap by comparison high res camera in a much smaller package for macro and landscapes is starting to get very tempting. I suspect that I would go for the D800E but I will wait and see what the real world testers come up with before making a decision.
+Phil Bishop price out going the medium format route. i have been for a few years now.
Interesting +Herb Chong. Do you have a pointer to some of what you have been reading regarding moire and product work? I have not really looked into it much - but that never stopped me having an opinion:-)

I'll happily change my opinion if I learn more and I always want to do that.
+Ian Leslie there are several ongoing discussions on in the Nikon forum debating this issue over whether to get the 800 or the 800E. i have seen seen a few instances of moire from my D3 in the roughly 100K captures that i have put on it. i haven't had my M9 that long and i haven't yet run into the problem but then again, i have under 1K captures on the M9 and i haven't done any product shooting with it. the moire instances i saw on my D3 were pictures of fabric and were essentially uncorrectable without huge amounts of effort, effort that i was not willing to do. i understand the math behind moire and i have seen more realistic examples that were not worth correcting and they have all been of fabric.
The only reasons for not stepping up is that you need the extra fps or you don't want to spend the money. The D800 has everything the D700 has and more.
+Piotr Debek Yes, I have looked at Magic Lantern and a number of other hacked firmwares. The problem isn't that we can't program our cameras, it's just that it is a real pain in the butt to do so, and has a high incidence of being clobbered on firmware updates. We need camera makers that create and support APIs.
+Thom Hogan As usual, a concise, precise and excellent analysis. Now if only those upgraders started unloading their D700's soon... I need a new backup (well a used new one, that is).
+Younes Bounhar I'll let go of my D700 for $2300 upon release of the D800. If you are really interested.

I'm more interested in the video than the mp, although, the D800 looks like a nice hybrid. Let me know if you are seriously interested.

As soon as the D800 is released we can make the exchange.
Personally: Nice camera, but don't need all the megapixels, and still waiting for that D400.
Is it just my imagination, or does the buzz on the D800 seem similar to the D3?
+Anthony Beach no, it's not your imagination. people have short memories but then, context changes too. there are people who squawk about far more megapixels than anyone in their right mind needs, six being their optimum. there are those that think that high ISO noise is too much and that the D3s is just right. then there are those that think that the D800 ought to sell for half the price Nikon is asking on general principles because "its all profit anyway".
Right Herb, but what I mean is that the D3 was a bombshell because it did two things:
1.) It was Nikon's first FX ("full frame") DSLR.
2.) It took the lead in high ISO performance.
Likewise, the D800E does two similar things:
1.) It is Nikon's first smaller body, relatively lower priced, high MP DSLR; and Nikon's first non-AA filter DSLR.
2.) It takes the lead in resolution among all DSLRs.

What I'm saying is that it's not just that the camera is a first for Nikon (such as the lower priced D700 was when it came out), but it is also taking the lead in some easily identifiable way over its competition -- ergo, the buzz we are seeing today all over the Internet.
+Thom Hogan: You said "slightly smaller in photosite size than the D7000". Really? Calculating from sensor area and number of pixels I come to a slightly larger photosite (D800 has less MP in DX-crop than D7000).
+Robert Franklin removing the AA filter isn't a new concept. Medium format digital cameras have been doing it for years. Photoshop, phocus, capture one all have tools to correct it.

You'll be fine if you are editing your pictures. It will just be an extra step.
+Thomas Rubach I think Thom listed the pixel pitch. The D7000 is like 5 microns and the D800 is like 4.3...don't have the exact numbers on hand, but the cameras pixels are pretty close to the same size.

Eh, I don't think pixel pitch is as beneficial to ISO as people think. My hassy has 9 micron pixels and it doesn't come close to the ISO performance of the D700.
+Anthony Beach +Herb Chong Where are you guys getting that 100lppm figure for a D3X? Extinction resolution is around 76 and absolute resolution is around 56,5. Even with that, hitting those in hand-held shots would be extremely difficult in most non-strobe situations. Anyway, if either of you have links to tests, I would be happy to view them.
I will take Thom's comparison that pixel density of a D7000 is close to a D800, so that gives 83lppm absolute and 90lppm extinction resolution. Lenses, whether or not one has a strobe or tripod in use, and technique will play a great deal more into whether anyone buying a D800 gets near the limits of the sensor.
+Thomas Rubach +Ikaika Arnado D800 is 4.88 microns. D7000 is 4.78 microns. So yes, my comment was backwards. But the point still is the same, they're relatively close, so should be similar.
+Thom Hogan Your comment "So that all brings you back to the sensor. 36mp is more than anyone asked for. 36mp without AA is more than anyone could imagine. Some will love that simply because more is better. A number will run into the shot discipline wall. Others will find out how different the corners of their lenses perform compared to center. Some will get stuck on diffraction or moire." really says it all for me.
I recall a discussion with a Kodak sensor engineer that a 100MP roughly 6x4,5 sensor would be the final working limit to get high resolution, high fill factor, and avoid diffraction at usual working apertures. Seems the D800 is hitting on the edge of that now.
As for whether I would buy a D800, I would probably look more into the D4 before a D800, just for how it fits into what I shoot, but all things considered, and I'm not doing video, a D3S still solves my needs and generates profits. I think the D800 is more aimed at the fine art print crowd.
Huh? 36mp with the absence of an AA is old news in the medium format world.

Nikon is just playing off those cues because it wants to market the D800 as an alterntive to the H4D-40 which is the standard new gen Hasselblad.

36mp is a tad much...but it isn't that crazy. Phase One has been putting out 80mp backs for a while now.
Gordon asked, "Where are you guys getting that 100lppm figure for a D3X?"

Re-read my post where I said 36 MP FX (D800) is 100 lppm and 24 MP FX (D3x) is 80 lppm.
+Anthony Beach Okay, where are you getting 80lppm for a D3X? I've never seen a reliable source claim more than 60lppm. (Edit) Seems to me you are equating file size with resolution.
+Anthony Beach Thanks for sharing those. While I do not consider Imatest reliable, due to operator interpretation in the analysis, it is interesting to note that Photozone does mention the error potential in their MTF testing. A little background for you, I've worked through a subcontractor on MFDB development in the past. Every White Paper I have seen relates the pixel cell site size and spacing to true resolution capability. I suspect we will differ on this. My take on the newest Nikon sensors and cameras is that they are very close to a practical limit in sensor technology. Canon made some ground in reducing the area between cell sites, but I think Nikon just jumped ahead of them with the D800; on that we might agree.
You're welcome Gordon.

All I've been saying is that the maximum potential resolution of a D3x is about 80 lppm. Photozone claims that many top notch lenses are at that level, that may or may not be the case; their website acknowledges ~5% margin of error at wider apertures, and they even test at different focus distances for the edges of the lenses to compensate for field curvature and such (which I would consider not relevant to real world shooting).

In line with what you wrote, DPR rates the D3x and A900 at about 76-77 lppm in absolute resolution and 56 lppm in extinction resolution. Having looked at DPR's comparison shots though, I have to wonder just how accurate their numbers are since they seem to miss focus an awful lot. I still think the D800E will come close to its theoretical maximum resolution of 102 lppm, but only under very isolated circumstances.
I have a generic question about AA filters. Given that they seem to split the image in four sub-images to slightly smear it out, replicating this effect in post processing ought to require only a simple convolution of the data with an appropriate point-spread function. Adobe will have to reverse-engineer it a bit and maybe use a "one size fits most" type of approach, but I see no reason why Nikon cannot implement something rigorously in NX that comes very close to an excellent replication of the function of the in-body AA filter. Am I completely out in left field? If that is the case, I see little reason to get a D800 instead of D800E (assuming that $300 is not an issue)
What could have been the driving force behind this lower then expected MSRP? lower R&D, manufacturing and sourcing costs? Or Nikon to compete in the market not only on specs but also price and trying to compensate the lower margin with bigger volume? The latter suspects that manufacturing probably will be balanced of against the volume needed, otherwise it would not make sense....

For the moment the D800 spec's and price look like trying to land a sucker-punch.....dangerous? maybe you get hammered back much harder......
+Huib Smeets They moved production to Thailand. Lower wages, better alignment with dollar conversion, more capacity, closer to some of the part sources, and more.
+Boyan Boyanov you assume you haven't lost anything that you are trying to reverse when you are doing AA after the fact. when aliasing occurs, you can't tell what the noise and what the signal is anymore. only from context can you tell that it has happened. you have lost information that is unrecoverable. that is why in manual repairs of aliasing, you take context and put in what you think is supposed to be there. it may look right but that doesn't mean that it is right.
When does upgrading make sense?

I use my D700 primarily for events, often with fast wide lenses. It is not photography where ultimate resolution is critical, Generally it is not even possible because you are shooting handheld, on the move, and not in great light. (I do not use a DX because I like the bigger finder and there are no fast wide DX lenses for Nikon.) Handling highlights is very important, shadows less so in event shooting. I am not sure I see any advantage to the D800 for this work, and the extra file size looks to be a real issue when you might shoot 1000 images for a event.
+Edward Richards for current D700/D3/D3s owners, the realistic upgrade from a performance POV is the D4. that's why i am getting one. the D800 is to replace my D3x for landscape and other fine art work. the D4 replaces my D3 for event and portrait work.
+Herb Chong I'm curious about the D800 replacing the D3x for landscape and fine art work - is that from the perspective of getting closer to a medium-format look, or is there a specific deficiency in the D3x that the D800 resolves?

I ask because I think, even at 36 megapixels with no AA filter, that 35mm digital is still going to look like 35mm digital regardless of how many pixels you throw at it. So I guess I'm at a loss for where the improvement is.

At the end of the article it shows some shots at different ISO, you can download JPEG at full resoltion to get an impression about noise. I've not let google translate the website, so we don't know what JPEG's these are: in camera or from post processing a RAW file and what happned during that PP.

++Update: in camera JPEG's.
+Herb Chong +Boyan Boyanov Once you have pixels, it's much as Herb says. However, the moire itself is generated by the conversion of raw data to pixels. Thus, yes, it is possible to have an alternate demosaic routine that tries to be moire aware and correct that during conversion. Certainly on the computer side we have enough horsepower to do so. That brings up a whole 'nother thing, though: today's raw converters aren't very well thought out. There are all kinds of things I want them to do that they aren't doing. Sometimes I wonder if the folk in tech are just following formulas or if they actually allow themselves to dream.
+Travis Pugh i have a friend who used to shoot 4x5 and then the original several Hasselblad medium format backs. he no longer uses them as well exposed D3X captures properly processed produce results that are every bit as good or better than the lower end medium format backs. the price of a 39MP medium format system is roughly 4x as much as the D800 for a small difference under certain situations and a lot of drawbacks in handling and portability. properly exposed D3X images are very hard to beat and the D800 will for all intents and purposes be about the same if not slightly better in terms of DR and noise. that's all that is needed to make most uses of medium format in that megapixel range unnecessary. those backs my friend has are older ones and technology has certainly advanced but all that means is that one has to buy a current back at its correspondingly high price.
+Thom Hogan i'll just have to disagree about the source of the moire. the use of discrete sensor wells alone is enough to cause moire. what it looks like in the output file is determined by the demosaicing algorithm. you know what the 2D DFT looks like and it's inherent in being discretely sampled.
+Herb Chong #2 message: no, "every bit as good or better" I don't believe. It's not actually possible assuming proper shooting and processing. It's the thing I was trying to get to on a previous discussion (how many pixels do you need?). Most people simply need to get above certain bars. Once they're above that bar, "more" doesn't generally gain them anything useful. What you're claiming is that if I shot with a D3x and a Leica S2 and the top Hasselblad, that I'd get "as good or better." What I'm contending is that, no I wouldn't. But I might get "good enough for my purposes."
Sure, a D4 would be nice, if my tennis elbow could stand it.:-) I am not doing pro shooting, so the incremental benefit of a D4 makes no sense from a $$ pov. I still shoot 4x5 black and white for things that do not move, so I have a high pixel option. I have found that the best way to beat 4x5 is with stitching. Once you make that step, then 36mp is a waste. Of course you can also stitch a couple of sheets of 4x5, which is even easier than digital stitching. Frankly, my pocketbook is relieved. 12/16/20 MP but with MF dynamic range would have been much more exciting for my work.That is what I envy about the MF backs.
+Thom Hogan i specifically referenced older model backs at comparable resolutions. my friend is not willing to buy the most current model of back at those resolutions.
+Herb Chong Ah, ok, so you were thinking of an acceptable capture quality to replace a medium format digital, output on something like a Lambda at 400 dpi using a paper that will resolve 100 lp/mm or thereabouts?

Medium format is still going to look like medium format, to me - there's more to the platform than sensor megapixels. Film size, working at higher stops, and the artistic qualities of different lenses all play into the look.
+Herb Chong Yes, Moire can in principle be generated even in a strictly monochrome image that requires no de-mosaicing when you have two superimposed periodic patterns with closely spaced periods. However, those rarely occur naturally, and most of the time the source of Moire is the fact that you are attempting to split a single periodic pattern into its RGB subcomponents, and the demosaicing process naturally ends up shifting them by 1 pixel relative to each other. +Thom Hogan I still don't understand why the need to have an alternate demosaicing routine. The effect of the AA filter is simply to spread out the image intensity over multiple pixels. As far as I can tell there is no phase information involved (unless there is cross-talk between the pixels and thus interference), and such amplitude spreading should be trivial to generate after the fact with an appropriate convolution before demosaicing. And yes, people do follow formulas, myself included. At least I am aware of it and willing to admit it :-)
+Herb Chong #3 message: I carefully used the word "generated" in my post. Maybe I should have used "instantiated" instead. The converter is the last place that we can do something about how aliasing/antialiasing gets into pixels. We most certainly can create a demosaic that doesn't produce visible moire when it is present in the underlying data (that's especially true of color moire). I won't go into the details, as there are patents pending. Indeed, as I noted, there are a lot of things that you can do at the demosaic that we aren't doing today. To some degree, that fits into what I call computational photography. Taken to the current extreme, we have Lytro and Light Field photography, where we actually use an extra set of microlenses to produce information that can be used in the demosaic to produce more information and better integrity for any given pixel. There's a lot going on in this realm right now, but it's mostly happening quietly, because everyone sees the income potential for their PhD work now (witness Lytro). Indeed, if I were working on a PhD in this area, I WOULDN'T actually finish my thesis (because it would become public). I'd take a year off and both patent and create a startup to monetize my work before going back and completing the final doctorate work.
+Thom Hogan I think you nailed the D800 wright up. How low have you seen the D700 go for used? (that might change my ending to this story :) but most likely not)

I look back at the DX side and see that we have vastly improved pattern noise (huge problem with the D1). Transfer speed has been improved past the point that is needed for still images (nikon 1/d4). Resolved problems with reflective CCD with flash, wireless flash and battery (D2h). From the D2h on I feel the DSLR has not vastly improved. Most jobs that could use 35mm in the past could get done with that or similar camera. The D2x (even D1x) and beyond started to truly replace 645 in a lot of studios. (not saying they were as good, but the job did get done with them instead of a Mamiya.)

Brings us to FX and the improved native resolution, it also gave back some lens that were "missing" and a better viewfinder, but set us back in pixel density. For me the job can be done with any modern camera, but the D800 (D3/D700/D3x/5D II/1D V) seem to be what is needed to slow the upgrade time frame. I would like to see nikon go to a 6 or even an 8 year pro update cycle (they wont do that, but nor will I be upgrading from the D800 until a major advisement or OS / modular the becomes norm (I want to see how a EP121 (with external LCD option) packed into a DSLR would change on-site photography (just need to wait 10 or so years)).

P.S. I just pre-ordered my 1st new camera (D800) have been using the D1x and D2h sense 04 and 07. I will soon see if my clients see the difference or if I need all new lens (I dont expect they will and I wont, but who knows.)
+Thom Hogan I think that to handle a 36Mp sensor you need a very well done lens... any thoughts on that?
+Thom Hogan +Ikaika Arnado +Christoph Breitkopf As to used D700 pricing, I bought one for $1700 on the 3rd before the announcement and have had serious inquiries for an offer of $1600 for a 2nd body. It may be panic selling, or not. But people asking $2200+ for a used D700 when a new D800 can be had for 1/3 more I think are being unrealistic. As was observed early in this thread, the real (inflation adjusted) cost of the D800 is lower than the D700 introductory price.

That kind of downward price pressure makes a lot of sense to me sense since we are talking about iterations of a computer with a flapping mirror. And not many expect computers to hold resale value very well. If I do decide to sell the D700s in the future, I expect to "lose" money. (Obviously I think the purchase is a gain ... financially compared to renting equipment and financially / artistically getting value out of the use of it ...).
That's fine, if I can't get it for $2,300 than I don't mind holding on to it. Just because the D800 hits the market doesn't make the D700 a bad camera by any means.

I just eventually want to get into video. I'm not in a rush though, because that's a whole nother medium to learn, and right now my focus is on design.

I was thinking about getting a D700 for video, but if I can sell my D700 for $23 hundred, then I'll consider upgrading.

The truth of the matter is that I'm actually pretty turned off by Nikon. They recently stopped selling parts to independent repair shops. Frankly, Nikon has always been pretty horrid with after market support, and it's getting worse.

Also, the 5D markii is a well balanced camera. 21mp, decent ISO, and video. Frankly, the D800 is just playing catch up to that. Canon has historically always been a step ahead of Nikon.

Eh, I'm just reconsidering dropping Nikon all together.
+Philip Long +Marco Maroccolo Philip - Wow... interesting... I might never buy a new camera... No i should keep my pre order... I think :P

Marco - people asked the same thing when we got the D2x.
That is also when people relay started to update there hand held shutter speed needs (when in fact with the smaller than 35mm format that 1/mm rule should of always changed to 1/2 x mm with DX).
but in the end more pixels have not impaired any ability to get great IQ images with any lens or even having a harder time to get that image. (though it might be harder to get the best image possible and some lens might not be able to give you the best IQ, thus you would have no justification to use the higher MP camera )
Don't do it Ikaika.

Just a quick follow-up here regarding lppm. I did a quick test of my D300 and A850 and came up with about 60 lppm and 56 lppm respectively. Factoring in the AA filter, I would expect the D800E to hit about 90 lppm.
+Ikaika Arnado historical has been superseded by events. compare Nikon and Canon image quality using a number of criteria including noise and color accuracy and Nikon has comfortably passed Canon and is only continuing to increase their lead. Canon thinks video is the direction to head for future profits. Nikon acknowledges that video is important but that still photography is what matters.

i don't think video in DSLRs will matter much in a while. i think that instead, we'll see video cameras with sufficiently good still capture being the most important in the long run for those that care about shooting video and video in a DSLR always as an appendage mostly because of the different form factors needed for good ergonomics.
+Thom Hogan Interesting that you mentioned the Leica S2 as on a file size basis it is comparable to the Nikon D800E. While I don't think the potential customers would even consider a comparison, it seems that might be interesting. The larger chip on the S2 would seem to allow for better fill factor, and potentially better colour response, or has Nikon jumped so far in technology that they passed Leica/Kodak CCD sensors?
Speaking of glass... I must say every sample I've seen thus far definitely shows off soft corners. This camera is going to be brutal on glass. I am very glad at this point I've saved and purchased the best I could afford.
But the most important part of the camera was't the lens?
Objectives such as Leica or Zeiss (on MF camera, say Leica S2 or Hasselblad) could not provide better results than Nikon lenses designed for smaller sensors than the D800?
+Marco Maroccolo The lens is arguably the most important part of the camera system (I would say glass or strobes depending) but the glass is the one analog piece of equipment left. As for could leica or zeiss make better lens for a smaller format I think a lens like the 14-24mm provides the most clear answer (or even the PC lens that could be MF lens). As for MF sharpness its simple, the larger analog equates to a larger sample rate.
+Marco Maroccolo There's no "most important part" when it comes to resolution. Resolution is a combo play. Something like:

1 / Resolution = ( 1 / Lens_Resolution ) + ( 1 / Sensor_Resolution ) + ( 1 / AAFilter_Resolution ) + ( 1 / ImagingASIC_Resolution )

Up any one of those things and the overall resolution goes up. Maybe not much, but it does go up. Lower one of those things (bad lens) and yes, the overall resolution is hampered.

It's a tricky thing, and I think at 36mp we're going to find that a lot of people are tripped up by it. If you can get good results out of a MF camera, you probably will figure out how to get good results out of the D800. If you've never used more than a 12mp camera and you've never had someone really evaluate your technique and the resulting images, you may be in for some surprises and a lot of work.
+Thom Hogan this is why i take all comments about how amazing any particular lens is with a large grain of salt. i have a fair number of Zeiss lenses for a reason. i have tried to use many of the Nikkors in those focal length ranges and they just don't measure up, even when i try a few examples. older Nikkors fare even worse. i also think that many people who brag about how well they are able to manually focus those fast old Nikkors are just that, bragging.

even with the D3X, i keep my shutter speeds well above 1/FL for sharpness reasons when hand holding, more like faster than 1/250s with my 85/1.4D wide open. one thing about my D3X though and that is its AF for portrait work with the 85/1.4D really is exceptionally precise on what i want it to focus on. i am counting on that level of accuracy with the D800. DOF at head and shoulder framing with the 85/1.4 at f1.4 isn't much and i want the eye's iris to be sharp at all times. with that many more megapixels, misfocusing is going to be more obvious. i just can't believe the number of people who do focus and recompose with the 85/1.4 and say they get razor sharp images.
One thing I'm also seeing a lot of with regards to "12mp is enough" is the constant example of "I have made nice 24-inch prints with my 6mp D40" all the way up to "I make great 40-inch prints with my D700." I have absolutely no arguments with any of that. My mother has a 40-inch print in her living room that I made from a D700 and it's gorgeous and sharp and full of detail. No question.

But I'm not sure any of this is terribly relevant to the issue. While your 40-inch prints from the D700 are just peachy... why would we complain about having 3x the detail in said print? I just feel all this "12mp is enough for 40-inches" is just justification and resistance. Yes, it is enough... is enough want we want?
+Herb Chong You're doing serious portrait work at f/1.4?

I don't see why anyone would do consistently do studio work at those apertures. Most strobes pulse at around 1/2000...slowest is probably around 1/400...but the fastest reach speeds of 1/13000. So shutter speeds aren't much of an issue, unless the ambient light is drowning out the flashes.

In any case, serious ambient portrait work should always use a tripod or monopod. Even if you are only shooting on a 12mp camera.
+Sean Molin The relevant question is this "once you have enough, is what you gain above that worth the extra cost/effort/discipline/time/etc.?" There's no set answer to that, and different people will answer differently. That's one reason why I've been careful to use the phrase (or variations of it) "all else equal, I'll take the extra pixels." (There are other reasons I use that phrase, too.)

We are, after all, talking about a US$3000 expense, possibly more if you find you need better support or lenses or training. Plus some people will have to upgrade other things, like their computer's memory/CPU/storage. Costs are costs. So the thing about "enough" is usually someone saying that the costs don't justify the gains to them.

Let's fast forward this out 12 years. Those of you chasing pixels or every last thing you can get from the sensor will have bought a D800, D900, D1000, D1100. Where are we then? 96mp? Three-layer photosites? Quantum dots? Light Field focus? Some or all of those? The reason why we're there is that the camera companies--and in particular Nikon because it's the only large company left in Japan that relies on camera/lens sales for the majority of their business--have to keep selling you something to stay alive, so they'll keep iterating. But as we discovered with film SLRs, the number of folk who stay on the train goes down with each iteration. Eventually everyone says to themselves "this US$3000 camera is good enough to last me the next 10-20 years and I don't need any of that new stuff."
Hi Thom, I know everyone's debating whether a D800 or D800E will be better for their purposes, but until testing of camera occurs, no one can say for sure. However, if someone is mainly doing landscapes and macro wildlife photography (smaller apertures), it seems the D800E might be better. Then again, the difference might be so small, it might not matter either way. Any thoughts?
It still seems like such an abstract discussion to me, when it doesn't include consideration of whether or not you're capturing pixels that will show up in an appreciable way in your output medium.

Is most of the work that would cause an upgrade from a D700 to a D800 presented digitally? RA4? Digital negative to alt process? Inkjet?
+Jack Goldfarb Landscapes maybe, macro probably not. Too many things at the macro level have patterns in them that can trigger moire, in my experience.

Thing is, I think Nikon is probably panicking right now. They intended the E to be a choice a few people that knew what they were doing would pick, a very low percentage. But if my quick and dirty surveying is correct, it's 1/3 E, 2/3 non-E at a minimum right now, and a lot of the people pre-ordering E's don't know enough to make that decision. This is going to be a dealer nightmare if the E demand is that high, because it essentially means stocking two products instead of one.
Cool. Thanks, Thom. It definitely seems many people are pre-ordering the D800E without really knowing the differences or even caring about them. People read "maximum resolution unleashed," and stop there.
+Travis Pugh That's exactly the problem. I've been dealing with hundreds, maybe thousands, of people asking "resolution" type questions from the day the 36mp rumor first broke. The majority of those have no clue as to what is or isn't visible in their output and aren't thinking about output correctly. Moire can occur at lots of levels. If it's buried deep in 360 dpi, it's not going to show up in the print when viewed normally. But moire, as well as output, is a variable thing. So first you have to consider what you're shooting, how it might trigger moire, and at what levels. Then you have to consider your output and at what levels moire would be visible.

Thing is, we all pixel peep. I do it because I try to build images that are sound from the lowest foundations. The better the foundation is, the more solid (and sometimes outlandish) things I can build on top. But a lot of people are going to get a D800, crank up the volume (ISO 3200, Active D-Lighting, HDR, and maybe VIVID for a Picture Control), and they're going to run screaming from their screen when they see absolute crud in the pixels. When they come back to the screen, they're going to post some nasty thing about how bad the D800 is.

Here's my rule of thumb: if you have no idea what your current camera is doing in the pixel foundation and why, if you seem to have no control over that foundation, you shouldn't be thinking about upgrading to a new camera. A new camera isn't going to solve your problem, no matter how good it is. On the other hand, if you know exactly what your current camera produces at the foundation and know that you've maximized that, by all means, upgrade (and I should also point out you probably don't really need anyone else's help in making that decision).
+Thom Hogan Great article! Many thanks! I've been waiting for upgrading to D400 (US$2,000) for months, expecting a body construction similar to D300 and a sensor similar to D7000. I have a few FX lens and a few DX lens, including both DX and FX prime lenses. Now D800 (US$3,000) looks like a D400 if we put a DX lens on it. Assuming the extra US$1,000 and extra weight/size is no question, and 4 fps is enough for me, should I buy D800?
A few problems with the D800:
1) No support for the cheap but effective ML-L3 remote. Instead we have to get overpriced and outdated cable releases. But my D90 and even the D4 will work just fine with the remote.
2) The grip cost over $600.....that's just insane.
3) DX and other crop modes are useless for RAW shooters (JPG only). When are we going to get cropped RAW files? That especially needed with a 36mp camera. And the grip only increases jpg speeds so that is also useless for RAW shooters.
4) To E or not to E. That is a question that I just can't answer. I'm mostly a wildlife shooter but like to shoot many other things too. Just reading reviews doesn't give you enough info to make that call.

Still, I'm really temped to pre-order!!!!
+Thom Hogan That's a sound rule of thumb. It's also why I won't rush out for an 800. It looks like a fine camera on paper, but it doesn't fix any of my problems right now.

I'd still like to know what problems it does fix, for other folks.
Okay, I'll address that Travis. For Nikon the D800 and D800E are 5DII and 5DIII killers. Nikon has nothing to lose right now discontinuing the D700 and leaving that gap in their line-up; their competitors aren't filling it. Nikon has plenty to gain coming out with a camera that matches or exceeds what Canon or Sony have to offer.

The D800/D800E addresses the issue for a fair number of Nikon users for whom (justified or not) 12 or 16 MP isn't enough, and $8000 for a D3x is too much. If the response at DPR is any indication, Nikon has scored a hit with the D800/D800E. Geez, it's almost discouraging to me because the huge demand for D800E makes it that much more difficult for me to transition back to Nikon this year.

For myself, I have an A850 (which is how I got 24 MP without paying $8000). It's a nice camera and I'm happy with it. What appeals to me about the D800E is Live View (with actual MLU), better AF, access to Nikon lenses, dual memory card slots that are actually useful to me, non-CPU lens data (I hate that about my A850), more resolution, more dynamic range, and a wider range of useful ISO. Given all of that, I'm still not sure I will be getting a D800E this year, even though the price is very attractive to me.
Hey Thom, I am contemplating picking up a D800 and would like to know if my Nikkor lenses i.e: 14-24, 24-70, 70-200VR and the 80-400VR would be good enough to go with the 36MP resolution? Would be obliged if you would help me out with this:)
+Bryan Czar 1. The D4 doesn't support the ML-L3 remote. The highest camera that does is the D7000.
2. True. But I'm not a believer in the grip, and most people that are headed that way would be better served by a D4. Better grip controls and structural integrity to start with, and better fps.
3. Not sure where you got that from. DX and 1.2x crops create smaller raw files. FX=74.4MB, DX=32.5MB.
4. See my comment further up. Generally, if you can't answer the question, you probably aren't a candidate for the E.
+Pavan Kaul Well, if those aren't up to the job, which ones would be? Can't say about the 80-400mm VR, but that's a lens that's due for an update, anyway.
+Ikaika Arnado yes, i do serious portrait work at f1.4. it's my style and why people seek me out for my portrait work. it's not the same as other people's and that matters to them. it has the look that you expect from shooting with an 85/1.4, 100/2 and 105/2 wide open. i can count on one hand the number of times i have shot my 85/1.4D stopped down past f2. if i need to do that, i am not using the 85/1.4D. that's what my Zeiss 85/1.4 or 90 Summicron ASPH are for.
+Herb Chong If that is your style, maybe you should be looking at 6x9cm MF ... :)
Finally I can upgrade my d80.
+Anthony Beach Thanks Anthony! I can definitely see the marketing advantage that Nikon will be shouting from the rooftops, although the Canon shooters I've spoken to personally are pretty of laid back about the whole thing. Once you've got lenses, a system change is painful, so painful :) The Nikon would have to make coffee in bed for them, and even then a butler might be cheaper.

Live view, and the other Nikon handling features, I absolutely agree. I love, love, love focusing the 24mm PC-E via live view, although it's a monstrous drain on my already-reduced battery life at high altitude. Nikon handling alone is worth the price of admission. So I can definitely see why you'd want to upgrade from an A850.

I don't get the megapixel thing, but that's probably just me.

Further questions: Is the 800E version a trial balloon? There's no such option on the D4, and Nikon must have been doing the R&D on those two platforms concurrently. What about the D3x? Did they just cannibalize its sales?
I'm not sure what Nikon's thinking is behind the "E" version, but if it pans out then expect it to show up in other cameras.

As for the D3x, I think it's toast -- just like what Nikon did to the D2xs when they came out with the D300.
+Dennis Linden I think you need to study with the right wildlife pro. It's rarely about frame rate. In fact, as my assistant will tell you, I'm almost always on single frame shooting wildlife, and when I'm on continuous it's Low and I rarely fire more than a couple of frames. If it were about frame rate, just get a V1 ;~). It'll crank 60 fps at 10mp. Using fast frame rates is a bit like hunting with a shotgun. But the photos you see and admire are often made by the sniper shooters--single shot, perfectly executed.

I suspect we'll see the V1 phase detect sensor in a low-end DSLR body within two generations, if not next generation. But that still won't solve all your wildlife problems ;~)

BTW, lots of people curse me, including myself. It's why I've grown such thick skin. Oh, wait, that's fat.
+Thom Hogan starting to see D90 and the like users talking about upgrading to the D800 and what they expect to do with the camera. from what i see, most of them haven't tried using a D3x to see how well their hand holding technique holds up. some of them, in fact, avow that they never use a tripod with their D90 and get razor sharp captures. i think there are a lot of people who are going to discover just how shaky their technique is or that the D800 is wasted on them. i try hard to keep my 85/1.4 wide open shots at 1/250s and over when hand holding to prevent motion blur because of my unsteadiness and i still manage to get a few now and then where that shutter speed is not adequate. having shoulder problems for a few months just made it worse.
@Thom. Thanks for the info. FYI, The Nikon website says the ML-L3 is compatible with D4.

I'm very glad to be wrong about crop raw files. I've never read any literature that said you could do that. My bad. : ) My (poorly made) point about E or no is that one really needs to shoot with both to decide which is right. I don't think reading specs and opinions is going to be enough. Unfortunately, amateurs like me won't get a real chance to do that.
+Dennis Linden I shoot a lot of wildlife with the D90. I'll second Thom that the frame rate is not an issue. I use continuous low at 3fps. If I go to 4fps I'll often get frames I didn't intend. (I must be slow off the trigger finger.) I don't shoot sports but the D90's weak point is low light. You just have to deal with it. Even so, I still don't think frame rate is the problem. Slow shutter speeds and slower lenses would be bigger problems.
+Bryan Czar Well I'll have to double check again, but I think someone got that wrong. AFAIK it doesn't have the IR receiver.
+Dennis Linden At 1/500 shutter speed and 10 fps, you miss 98/100 of the second. Most of us eventually learn we need to time a jump, not shoot a burst on it.

Of course, I'm just guessing here based upon basketball, baseball, football, and other sport experience. If anyone here can get me press access to a major ice skating event, I can verify that for you ;~)
+Thom Hogan Can you elaborate a little bit for those like me who are not tecky as you... When you say D800 will loose resolution above f/8.. Actually, is that really what you're saying? When I shoot a photobooth at F14 to get everybody in Focus, will my images be as neat as on my D300 ? Tks in advance.
I'll add to the frame rate issue. I mostly shoot sports, typically hockey and running. (Odd mix, I know). I use a D300 w/ grip so I have 8fps available to me (yes a D3s would be better. No I can't afford it). I almost never fire more than 1 shot at a time it and when I do it is in a moment of desperation. While a high fps can be combined in post to show a nice progression, and I can occasionally see a use for it, I found that whenever I relied on a high frame rate to capture the one true "moment" I would always miss it.

So I had to actually learn how to time the shot. For running, I focus on the runners arm swing and can usually nail them every time with a single shot. Hockey can be a bit more difficult, but once I figured out how to see the game my "keep" rate went up as the number of shots I've taken during a game goes down. So I leave the system in CL at 6fps just in case and then get my finger off of that shutter as fast as possible.

I have not shot too many figure skaters, but the way they telegraph their jumps it was almost trivial for me to nail them the few times when some would be out on the ice before the hockey games. Really good and fast spins would take a bit more practice.

Now what is critical is a very low shutter latency and focus acquisition speed, but that is a different spec than frame rate.
Ghislain, in your photobooth scenario f/14 will have as much resolution within the DX portion of the D800 as it will over the entire frame of the D300. However, if you are at the same AOV using both cameras and you are getting everyone in focus at f/14 with the D300, using a longer lens you might not be getting everyone in focus with the D800, so you may have to stop the lens down a little over a stop to make up for that. Will that reduce your overall resolution with the D800 compared to the D300? No, because you have a lot more resolution to lose in the D800.
Thom: The last 2 posts (D800 How Good is the Q? and How Good is the Q, II?) on your webpage are confusing to me and I am having a hard time following the math and what is being said. Perhaps you can clarify the posts. Thank you
Well Reza, as a layman who shares your aversion to math, I would say the upshot of those comments is that at f/8.9 you are hitting the absolute limits of the resolution you can get from the D800E sensor, and that correlates well with where many lenses peak in performance and is as good as the imaging sensors used on "earth imaging satellite systems!"
+Reza Gorji Just keep watching the site. I seem to be channeling engineers. But each iteration gets a little more interesting ;~) (and understandable).
To Q or not to Q: Galls/Guys, Please raise your hand if you mainly print 20x24" or bigger, often strongly cropped images, your print viewing distance is at nose-lenghth, (or less), shot from a tripod with one of the better nikkor lenses. How many hands do I count? And I suspect ya'll prefer the NON AA version? :-)

Those that still have their hand in the air might want to take notice of Q and, all others probably not so much. I lowered may hand too.

There are so many factors/issues where you will loose resolution like lens QA/QC issues (the lemons), air vibration on a hot day (that's why astronomers hate twinkling stars), lens corner-center sharpness as function of the aperture, lens-sensor misalignment, AF issues, curvature of the plane of focus, vibrations (like with hand-held shooting), etc etc. You probably gave it away even before diffraction kicked in.

I've ordered a D800 too but not for the 36mpx. Same camera, same price, same specs but with a 18, 20 or 24mpx sensor and I would still have bought it.

But it is good to know that when I follow some "rules" I will be able to squeeze out some more detail when I want it.
Maybe I've misunderstood completely, but as the D7000 and D800 has about the same pixel sizes, won't all these things about non-handholdable and diffraction issues apply equally much for that camera?
+Magnar Myrtvelt : Yes,

I think people are triggered by pixel amount and not pixel densitity so at 16mpx nobody realy took notice.

But to archive the same DoF on DX as on FX you stop down less on DX (shorter focal length for same angle of view) so you experience less diffraction.

To have the same scene (group of people in a 3 by 2 feet 'frame'), composition:with same DoF

16mpx DX: f=35mm, A=f5.6 (not calculated, it's for the example)
36mpx FX: f=50mm, A=f11

FX will resolve the 2by3 feet in more detail then the DX as it has more pixels to sample. But it will not be the full pixel resolutuion as the diffraction at f11 is playing a role.

DX will resoleve the 2by3 feet in less detail than FX as it has less pixel to sample the same dimension. But it is less diffraction limited then FX.

So, 36mpx will yield more detail but not (36/16) 2.25 more (or 1.5 in liniear fashion). as diffraction will take a % away from this.
+Magnar Myrtveit one additional factor. People use 1.5x wider lenses on average on DX. That covers up some technique issues. 
+Huib Smeets Technically, most of the factors you mention would also relate to Q. Let's put it a different way: if you're going to pixel peep or use all those pixels to print wide, Q is of interest to you. If you're going to shoot JPEG Small, maybe not.

The DOF calculations are indeed something that people need to consider. We'll get to that soon with my continued rolling out of Q.
+Magnar Myrtveit You are correct. However, I sure see a lot of people who moved from 6 or 12mp to 16mp that are struggling to get "more" out of their cameras.
+Thom Hogan i can't believe the number of people who are saying that since the D7000 has the same pixel density as the D800 that the D800 doesn't require any more technique than the D7000 to obtain equally sharp images.
Edwin L
I don't really think I would get much use out of a 36 MP images of people but I would like it for Landscapes.A Gigapan Epic is about $900 in my area. D800 is about $3200. Would a Gigapan head be a better choice for landscape shooters? The samples on the Gigapan website looks like it has tons of details.
+Edwin L time - it takes time to set up and capture a stitched panorama. weight - the GigaPan Epic is in addition to all your other equipment. if you shoot really high res stitched panoramas, you'll need fewer captures with the D800 because you can use a wider lens, assuming that the lens is good enough to provide the same net angular resolution.
+Edwin L What +Herb Chong said, for the most part. But stand by, Q is going to come up ;~) So what he wrote about using a wider lens may be false hope. Problem will be that you can't go beyond f/9 and retain the detail on a D800. Moreover, a wider lens is going to have corner issues. It's tricky to find the balance point where you're optimizing the data AND the number of captures. DOF and diffraction and lens limitations all come into play.
+Huib Smeets Side note: Monet looks horrible when I press my nose to the print.

Calculating diffraction limitations should take into account viewing distance, at some point.
Basically what +Thom Hogan writes is that you have the dilemma that for deep DoF (toe to horizon in focus) you need to stop down, but that kills the resolution due to diffraction and you will have effectively less resolution then 36mpx could deliver. But if you open up the aperture the less then ideal lens will be soft in the corners and you loose resolution again. Bummer.

By the way, the smaller you print the lesser the issues will be ;-) So if you plan to print D800 files not much bigger then 16x20 you can stop reading.

In certain situations, a Tilt and Shift lens can save the day: deep DOF with relative wide aperture as you can make use of the scheimplug rule. I used it on my 4x5" technical camera always, it was the 'only' way to get deep DoF with moderate apertures and retain resolution. Look at wikipedia what Scheimpflug is. But be aware it's not a panacea in all situation as the tilted plain of focus is still a plane and what you photograph is 3D, specially with nearby object a challenge.

As such a T/S lens is on my shopping list for those landscape shots I want toe to horizon sharpness without loosing resolution because of diffraction. Sadly however the Schneider-Kreuznach T/S one's need me to save some money year after I get my performance bonus of the company again (i hope ;-) )

With high Mpx camera we come back into the realm of old days with 4x5 and 8x10 negatives and technical camera's. That's were I started in photography and seem to be returning too . Whoahahahaha :-)
I'll admit that I don't understand the reported demand for the E model. When I first investigated the possibility of having my AA filter removed (e.g., I saw the difference, and yes the detail and edge acuity is just a bit better. But overall I understood why the camera makers all put an AA filter on. Moire jumps out and is highly annoying, and the detail increase is slight. So my preorder was for the base model.

Now, if I could press a button in lightroom and have it apply the AA filter in software with the same results as I could get in hardware for the images that happen to need it, the E model starts to make sense. But it really does seem like we're spending a whole lot of energy for a small gain. I suppose that we'll find out soon enough when the interweb reviewers get their hands on the base and E model side by side.
+Philip Long the D800E is only slightly more expensive and is covered by warrantee. $300 more is worth it. It's when people start running into aliasing and moire problems whether they will judge it a wise choice or not. With my M9, I have run into the problem less than I feared and have had other problems that I had hoped would not be.
+Thom Hogan I just wanted to say thank you for posting your writeup on this. I've been following your reviews online for years, ever since I started with my first DSLR (D70s) years ago, I guess I wanted to finally stop lurking and looked to you for the D800 review. I have a D700, absolutely love it, but am very interested in this camera. I don't see it as a D700 replacement really, but something that will help alongside. I focus on family portraits, but not quite the formal posed photos. I'm usually running around, jumping on playgrounds, rolling on the floor (We'll see how long that lasts in a few years..) but will also set up the group shots as well.

I think this camera has a nice balance for someone like me. I feel I could use the 36MP for the more posed portraits, where I know I typically have families getting 16x20-30x40", and for the quicker more candid fun ones with the children, I can shift to the 16MP if I wanted, or just use my D700 as a backup. I always hated the idea of blowing up a 12MP photo, even taking into account viewing distance when I help families select walls in their homes, and this could fill that gap.

I think that knowing that the images are 36MP should be a nice reminder to folks to not just click away all the time. I only focus on bursts where I know there is an active child and with as much anticipation as possible, you still have to hope for that cracked smile that lasted a fraction of a second.

I think the E version isn't necessary for me, but this could at least cover all my options of larger format portraits I was hoping for. Not everyone will be happy, but for $3k, it is a lot of power. Again, thank you for your post. Big fan.
+Thom Hogan Why not just a fisheye and D800 for landscape images? It would seem to get around diffraction and edge to edge exposure questions. Note: this is hypothetical and I'm curious if this leads to a different explanation than what I expect.
Amit RJ
Hi Thom,

Where do I comment on your article regarding Kodak and its hopes in the printing business?

Thom, been wondering for a while why Nikon does not launch the same camera with different sensors. For example why not :
- D800 : this 36mp sensor
- D800fx : the D4 full frame 16mp sensor
- D800dx : the 16mp dx sensor from the D7000

Identical bodies, controls and menu options.
The only difference would be the sensor and imaging pipeline.

Would this not reduce engineering and manufacturing costs for Nikon, and make life easier for photographers, since they would not need to remember the subtle differences in controls and menu options between different models ?

In fact why not the same with the D4 body - D4, D4x, D4dx ?

Why don't they do it ?

What are your thoughts ?
+Argho Sen well i'm not Thom but there are many reasons not only technology but also from a marketing and financial perspective to do like nikon does. These guys are not idiots :-). To take your question further: you just could also sell 1 body with 1 sensor and unlock functionality via software/license keys: fps, sensor resolution, fx or dx, af and ae performance etc etc. Yes it could be done, but will it be positive for the bottom line in the short and long term? I don't know and as we have not seen yet such strategy i assume it's not positive for the bottom line.
+Argho Sen well i'm not Thom but there are many reasons not only technology but also from a marketing and financial perspective to do like nikon does. These guys are not idiots :-). To take your question further: you just could also sell 1 body with 1 sensor and unlock functionality via software/license keys: fps, sensor resolution, fx or dx, af and ae performance etc etc. Yes it could be done, but will it be positive for the bottom line in the short and long term? I don't know and as we have not seen yet such strategy i assume it's not positive for the bottom line.
+Argho Sen +Thom Hogan I don't know if it's possible, but I'd really like a D800s (or L, or whatever) with the D4 sensor. I'm a freelance photojournalist and 36 mpx are way too much for me and my workflow. I know i'll have to update all my tech (computers, hard drives, memory cards) but not right now, not while Spanish economy go through a neverending hell... I wouldn't mind to wait for an year to have that camera, like i waited for the D700 since Nikon released the D3. But... if they aren't planning to do so, maybe I should buy another D700 or (in a better scenario) a D3s while they're still available. So, what do I do? I guess there are lots of freelance news photographers having the same thoughts.
+Philip Long The D800E has a Q of 1 at ~f/9. The D800 has a Q of...oh oh, I've got engineers fighting over that. Either 2 or 1.4 or something not 1, which changes the aperture crossover point. Like I said, I'll have more to say on Q as soon as I can get an article peer reviewed so that I don't simplify it too much ;~)
+Gordon Moat Are you going to correct linearity out of the fisheye? If so, we no longer have a 1:1 correspondence with our image data at the pixel level. Most fishes have some clear edge issues, too. I'm sure some people will try this, mainly because most of us tried even back in the 6mp days. There's something to be said for the simplicity.

If you're referring to stitching fished images, no, I won't go there. The math is possible, but in my experience it's not the way to generate a large pano print with clean detail.
+Argho Sen This is the car industry solution, to a large degree. But even there, they don't offer all options in all models.

My take on this problem is simple: just invent a modular camera and the problem solves itself, AND the economics work to the favor of both maker and customer.

If you analyze the cost in today's DSLRs, you find that it is decidedly one-sided. The expensive things are really only two: the sensor and the digital board it connects to (in Nikon's world, they've now pretty much internally modularized those; in the early days of DSLRs there were daughter boards and a lot of distributed "digital" cost throughout the camera). The problem is that when you trade in your D700 and get a D800, you're essentially replacing a lot of things that aren't worn out and don't really need changing: the viewfinder, the LCD, etc. You'll note that Nikon tends to make small changes to these to cover up the fact that you're essentially replacing something that really doesn't need replacing (100% viewfinder, 3.2" LCD, etc.). This is not the Apple way. It's not the way I'd do it. It's a traditional Japanese consumer electronics iteration model.

But there's a subtext, too. What you're suggesting means a dealer would have to stock 15 or more Nikon DSLRs. It would mean that the already confused customer would be confronted with even more options to confuse them. It would mean that both Nikon and the dealer would have to get really good at predicting sales, and as I've noted, I think Nikon has already goofed on their own internal predictions on the D800 versus D800E balance.

Modular systems basically remove a lot of that. You buy body on capability and you stuff it with sensor/digitizer suited for what you shoot. If either of those things changes, you buy the body or the sensor/digitizer that matches your new need.

I actually started to build a simple financial case for this in my presentation to Nikon in early 2010. I've since built a more elaborate case. It really does make sense to both maker and consumer, I think. It's another system lock-in, like lenses. It also encourages more buying (as I've noted, I'd want a high-res sensor, a high-ISO sensor, a monochrome sensor, and an IR sensor for my modular D5, and that's just for starters).
+Edgard Marques The money is betting against photojournalism. There is more money in videojournalism already, and the trend is clear in the shift. "Journalism" has a time component to it. We call it "news," after all. The media of choice for primary news consumption is shifting, as it has many times in the past two centuries. It's now TV and Internet. Coupled with the ubiquitous cell phone cameras and amateurs willing to give content to news outlets, the number of paid practitioners of traditional still photojournalism is declining, and somewhat rapidly.

One thing that many amateurs don't understand is why so many still photographers have been moving over to the video side. Laforet, McNally, etc. Follow the money. When a still photograph is no longer worth much but a video shoot is commanding big bucks, guess where the pros go? But this transition from stills to video is happening all around imaging, and as bandwidths increase it's only going to get worse.

If I were a photojournalist, I believe I would now want a DIFFERENT kind of camera because I'd want to change my shooting style. I'd want to shoot primarily video but be able to snap stills at key moments during the video stream I'm capturing. Indeed, those stills may have very different parameters (faster shutter speeds, for example). Curiously, the only camera that comes close to doing what I want right now is the V1, which can take a full resolution still while recording HD video. But what I'd want is coming. Moreover, if we transition to 4K video, I might be able to pull stills from the video stream after the fact, so I'd really only have to "mark" the moments I want pulled. Basically put a marker in the video stream, then in post processing have my software go to that spot and show me the images at and on either side of that point for a still extraction. (Should I put a patent pending on that? ;~).
Tks for the responses guys.

Yes modularity is the best solution, but as you pointed out, Nikon does not want to offer modularity, as they want to sell more bodies. But even so, what I am suggesting will reduce complexity for Nikon.

Not sure if it will increase the number of models much, if we take DX out of the equation.

So 2 sensors (FX high ISO and FX high pixel) * 2 body types = 4 cameras
- D4
- D4x
- D800
- D800s

This would replace the 3 camera lineup we have right now - D4, D3x, D800.

Even if we add back DX to the pro bodies, it would still only give us 6 cameras, and there would be no need of a D300 replacement.
Then the D800E does have more resolution than the D800 up to f/9, and not just "false detail?" Also, I presume then that the D800 is at optimal resolution ("Q"?) at perhaps f/11? If this is correct, then a landscape photographer who spends a lot of time at f/11 and less so at f/9 wouldn't really be gaining much with the D800E? My primary concern is when I go to shoot the Golden Gate Bridge with a D800E I might run into aliasing (happened once with my D300), and that might be more avoidable with a D800.
+Argho Sen They may want to sell more "bodies," but that's a fool's gold. Simply put, we're at the nadir of DSLRs. Five years from now DSLR sales will be far lower than they are today. Why? Because we're well above the "good enough" bar now and the smaller stuff is getting over the bar, too. That, coupled with rise of cell phone cameras, just squeezes the market. DSLR unit volume went down a bit in 2011, and it's not going to recover overall, I think. We're headed right smack dab where we were in the 90's with film SLR bodies: once someone has a decent one, why upgrade?

Modular actually addresses the lower demand issue, and directly: you still have something to sell users. Just as Nikon/Canon/ started relying upon more lens sales in the 90's, you could rely upon more module sales. It's the only compelling answer other than come up with something that makes everyone want to dump their DSLRs for the new thing.
+Anthony Beach Technically, yes, removing the AA will resolve slightly more than AA. Filters aren't brick walls, they have slopes, so an AA filter doesn't let everything below Nyquist through and cancel everything above Nyquist. It starts stopping some things before we get to Nyquist and still lets a bit in after Nyquist. Put another way, filters are not binary (1 or 0, yes or no). Technically, if you look at any digital system close enough you find analog deep down in the muck. Digital systems, like filters, transition from one state to another, and not instantly.

The Q question will have to wait a bit for my attempt at a more definitive answer. Like I said, I'm trying to get a bit of peer review into it before publishing it, so that I don't make any silly mistake in trying to explain and simplify it. One thing that's being argued about is the value of Q for Bayer systems (it isn't a static 1, it's a variable because we're recording an octave of wavelengths). But I promise you I'll publish something on this in the not too distant future that tries to put things into more layman terms.

As for the Golden Gate Bridge, well, you're describing another reason why we have antialiasing filters (and a reason why Fujifilm's non-Bayer sensor on the ProX-1 is not the ultimate answer). For those not up to speed, aliasing is usually seen as those stairsteps (jagged edges) on diagonals. Which the GGB has not only multiples of, it has almost every angle of diagonal so it's prone to trigger stairsteps at some point on the cable if not multiple points. Antialiasing basically smooths diagonals and those stairsteps disappear at the expense of edge acuity. Which we try to put back in by using sharpening (contrast) at the smoothed edge.

The D800E will definitely stairstep somewhere on the bridge's cables. The Fujifilm ProX-1 will do something else: ANTIALIASING will be partially triggered by color, and unfortunately, since the cables on the GGB are not white but a color that is captured at a lower density than on a Bayer camera, the cables will do the opposite of what the D800E does with them. On the D800E you'll see blurred edge, blurred edge, stairstep, blurred edge, blurred edge. On the ProX-1 you'll see strairstep, stairstep, blurred edge, strairstep, stairstep. Neither is a great thing to have in your data if it's going to rise to the level of visible. It's one reason why I prefer AA filters over no AA filters. But, pixel count gets in the way, too. If you get enough pixels, all this gets buried deep in the pixels where it ISN'T normally visible. I'm not sure if 36mp is the point where I change my mind (and go to no AA), but 24mp was not that point.
+Thom Hogan Thanks for the reply. I was considering correcting to linear, though with that thought that there is enough information that there might finally be an advantage to doing that. However, with your current math based path on the D800 perhaps there is a mathematical explanation of why, or how many pixels would be needed for it to be effective?
Ricoh has a sort-of modular camera, but I don't think it has been that great a seller. Horseman briefly had the DigiFlex models. Some medium format cameras have been modular in regards to digital backs. There was also the earlier Leica R model digital module. Where you thinking along those lines somehow, or a different idea for a modular DSLR?
I use to do QTVR in the late 1990s using a Nikon fisheye on a 35mm film body. It worked well for re-display on a computer, but I never considered printing one of those at the time. My question on the fisheye and D800 came about because some pano rigs basically create a semi-spherical capture the way the lens is rotated in them. The only multi-shot panos I have printed have been two or three shot from a shift movement, which is linear to the scene.
Thanks for that detailed reply Thom. Resolution was the at the bottom of the "pros" in my "pro/con" list for the D800/D800E, and availability of the D800/D800E are at least as important to me. Factor in a host of other "pros" (mostly listed in an earlier reply of mine in this discussion) and it looks like the D800 (not D800E) is a slam dunk for me. Next up for me, put my D300 up for sale before the "D400" comes out, and that is part of the "cons" inasmuch as it's a hassle as well as an expense to buy and sell gear.

The only "pro" I know I'm going to miss about the A850 will be the Minolta 200/2.8 and 1.4x TC -- that is an optically excellent and delightfully compact combo for getting some reach into my bag. I might also miss the colors I get from my A850, I just read yesterday that the D800/D800E may not be as good in that regard. It appears that there's an inevitable trade-off between color separation and higher ISO performance, but I give the edge here to higher ISO performance. The biggest "con" for me sticking with Sony is that if the "A99" is anything like the A77, then I'm not interested in "upgrading" to that.

I can see the D800 being the last camera I ever buy, I think it's that good (for me, YMMV). I think that's a good segue into modular cameras though because another camera I can see getting would be a monochromatic version of a D800E. I think such a camera would be a good way for Nikon to keep selling some cameras to D800/D800E buyers, it would have more DR and resolution, it would address moire, and it would be a faster camera (I would think taking away the CFA would make it about a stop faster). If Nikon were to come out with a modular "D5/D500" system with multiple choices for the sensors, I think that would be very enticing.
+Thom Hogan +Edgard Marques Canon introduced a line of mostly video with adequate still capability cameras recently. they are following the money too. Nikon is unwilling to move in that direction yet though there are plenty of prognosticators that are in each camp.
+Herb Chong Not sure the C300 is "following the money." With these high-end video beasts we're talking about unit volumes that barely approach 10k units a year (the high-end pro DSLRs are an order of magnitude higher than that). Which do you want to be, a big fish in a small pond or a big fish in a bigger pond?

This is part of the dilemma that all the camera makers are facing, but which particularly is critical to Nikon, who relies on cameras for the majority of their sales and profit (unique amongst the Japanese camera makers). They've all got addicted to consumer electronics volumes. Nikon this year will sell 20m+ cameras/lenses (Canon more than that). Adding 10k units, even highly profitably, doesn't make a dent in the bottom line.

The reason to get into these new small markets is (1) prestige; (2) hope that the market below it will grow and you can move down into that; (3) taking any kind of growth anywhere you can get it, even if it's very small.
+Thom Hogan I just want to make sure I'm reading the chart correctly, because no one has mentioned it aside from me, but on the D800, won't lossless compressed 14-bit only be 41.3MB (average) instead of the 70+ that keeps getting thrown around? I'm trying to find the chart on Nikon's site right now but it's escaping me.

That's still a big file, but I think considerably more manageable... and that's the quality setting most of us RAW shooters will be at anyway.
+Sean Molin Yes, 70Mb is for the uncompressed 14b file and 41Mb for the lossless compressed 14b NEF. For storage space reasons the lossless compressed seems to be the sensible choice.
+Huib Smeets What possible reason is there for uncompressed NEF when lossless compression is available? Is there some raw converter out there that can't decompress it? The processing speed has to be faster for compressed NEF too (unzipping in ram is fast, moving a big file from card to disk then to ram is slow).
+Philip Long Nikon's lower end cameras don't offer uncompressed as an option. i don't know why they continue to offer uncompressed when lossless compressed is available.
+Philip Long A thing that comes to my mind is that compression requires more processing means more cpu cycles (or hardware logic) of the Expeed engine. More cpu means more energy consumed before the Expeed engine can go into sleep or low power mode. So in certain use cases you might get more shots from the batteries when the files are not compressed. But if and how big the difference will be......I don't now.

Many compressing/decompressing is a-symmetric (by design and by nature), meaning it is more costly to compress then to decompress, MP3 is such scheme, it requires relatively high cpu power to compress but the decompression is relative simple, that's the reason why your PC goes in 100% cpu utilization while you encode but your MP3 player runs for hours and hours on a very small battery. But I do not see the benefit here for a camera as the PC will be more powerfull in cpu and energy resources.

So, if uncompressed gives you no extra shots from the battery, then yes it seems that the uncompressed NEF does not make much sense. The spec's of the D800 in terms of fps and battery life do not differentiate, it sticks with 4 fps and x many shots.
You have touched on shot discipline a couple of times, but I'd like to hear more of your thoughts. Are you saying that to use a 36 (or a 24) mp camera we'll need to wait longer between mirror lockup and shutter release? Beef up tripods and/or heads? Hang something even heavier from the tripod? Or are you saying that a D800 owner will need to check off every box on your shooting checklist.
Thom. What the heck is Q? I tried various photography acronym lists and didn't find anything. This discussion thread refers to some earlier posts (on, I assume) - most recently on Feb 9, but your 2012 archive begins (ends?) at Jan 25 and your current front page begins on Feb 12, so any posts between Jan 26 and Feb 11 seem to have gone missing. Is this a conspiracy to keep me in the dark about Q (at least until the peer-reviewed article comes out)? :-(
+Herb Chong Thanks. I found some info about quantum efficiency online. I still think there are posts missing from, though, but I'm now willing to concede it's not a conspiracy aimed specifically at me!
+Richard Swearinger +Thom Hogan

Yes, of course one needs to apply proper techniques to extract the last bit of performance from camera's like the D800.

But does one always need to do so? Reading left and right on the web one could start thinking that it is the only way, the only appropriate way, all others dismissed. Now that would make the D800 a pretty useless camera if one only is "allowed" to shoot at f8, with MLU and tripod.

I'm not a member of the '12Mpx is enough" church and welcome the extra Mpx the D800 has on offer, however, I will not always utilize them as I don't need or want them always. I accept that at the pixel-peep-level motion blur or some focus issues might be visible as result of the shooting style. The best remedy is: do not go to the pixel-peep-show. Will it be visible in a print? maybe in murals, but I don't expect it to be a problem for 16x20 viewed from 2 feet away.

Let's face it, with this D800 and the like, we have surpassed what 35mm film was capable of, ever looked at stuff like Ilford Delta 3200 blown up to 20x24" prints? Who's complaining about DSLR's

One should not limit themselves to 36Mpx only :) That would be similar as driving a car on top speed you? I don't! :) but it is good to know that 36Mpx are there when I want/need them.

ow damn, yet another long winded response......
+David Meyers i don't follow via the web site so i didn't notice anything. i'm trying to figure out just how visible the effects are going to be when i print at greater than 100% if i shoot at f11 or f13, my usual landscape apertures.
+Herb Chong +David Meyers No, Quantum Efficiency refers to how well a photo diode matches electrons against the photons that hit it. If 100 photons hit the diode, how many actually get converted into electrons? The higher that number, the better, especially for low light. We refer to this a QE.

The notion of Q we're dealing with is a different measurement. I've been trying to find the original definition as originally stated back in the early days of reconnaissance imaging, but I haven't found it yet. As I'm finding, even amongst current sensor engineers, there's some disagreement over some of the aspects that Q tries to capture.
+Huib Smeets there are those who insist that since the pixel density is the same between the D7000 and the D800, any technique that works on the D7000 will work just as well on the D800. if you don't print any larger than before, and i don't see why anyone who doesn't print very large will benefit more than marginally from a D800, you might not see any difference. if that's really the case then i think the person who bought the D800 just wasted a lot of money for tiny benefit and doubly so for those who get the D800E without realizing the potential pitfalls of not having an AA filter.
Why has Nikon chosen to include a 1.2x crop mode in the D800? In what situations would I want to use that mode?
+Roger Wiley i imagine any situation when you would use a crop mode at all, because you know you are going to crop and you are confident you can frame in camera correctly and not require cropping after the fact. those are the situations when i use crop modes on my D3x.
+Herb Chong Correct: 36mpx is in my case often overkill as I wrote already. I'm fully aware of that. Isn't that funny? The only answer I have to the question: "Then why on heaven's earth did you order this camera and spend that amount of money" is simple: #1 (and most important one) The D800 is one of the first DSLR's I see myself still using 10 years from now (sorry Nikon for the bad news), it has enough of everything. #2 it is not cheap but affordable when you set it off against 10 or more years of usage. #3 I have the money to spare. Well you have it, #4: I like nice stuff, big toys for big boys.
+Herb Chong Basically true. It's why I suggested looking at the D7000 versus D700 in the first place, many moons before the D800 appeared ;~). But that was in regards to pixel level integrity at high ISO values. Other variables are important too. I think the one that's going to catch most is that, at pixel levels, they're going to see a big difference between center and corners. DX lenses for the most part have a larger than needed image circle. FX lenses (other than the PC-E ones, usually don't). But we're all waiting to be able to test to see what we really see.
+D-r D-r "Nikon doesn't have any serious presence in video." Well, they have a D800 and D4 that can record near state-of-the-art 1080P/30. This is the age old problem with systems: you need to move them forward simultaneously on all levels. Introducing the Nikon 1 without any true wide angle or fast prime puts it at a disadvantage to the other mirrorless systems. Introducing DSLRs that can drop uncompressed 1080P/30 to relatively inexpensive external recorders without having lenses that a video user would want is another. If you don't quickly follow up capability with options, you simply lose those potential users. Put another way, Nikon continues to make moves that indicate they want to play the video game, but they seem to think that we're going to do that with the existing lens set. Not without modification we won't. Friction. Friction. Friction.
Well, its the D800E plain and simple, am excited to get my hands on one soon, hopefully mid April to late April-put a deposit on one at the Camera Store in Calgary AB so time will tell when they finally ship- am intrigued to shoot landscapes with the Camera. I thought the D3X was it for landscapes and read many a photographers review of that fine camera's ability to capture your imagination,image wise.
I'm thinking the D800E will have the same or better ability at a fraction of the cost.
Thanks for your article “D3, D3s, D3x, D4, D700, D800, D800E, or Wait?” on your web page. It completely answered my previous question. I have pre-ordered one D800 (without E) and maybe, for the time I have learned how to squeeze almost everything out of it (I am an optimistic) a new D800 with the high ISO sensor of the D4 comes out and I can buy my second body and have two, for me more than enough good bodies, for the price of one D4.
Interesting article, what's in a name? +Thom Hogan

This is my best guess so I'm going to call the D4's little brother a D710 FX (D4 sensor&tech in a D800 body). I mean "why wouldn't they make it", the D700 pricing brought FX to photographers who could not afford a D3 so if they don't make it someone else will eat their lunch. The D800 is a technological ground breaker that's redefined the niche wedding/portrait category that Canon created with the 5D Mk I. If Nikon wants to increase market share, revenue and profit they need to create camera's that no one else makes or thought of. Using a boxing analogy, this would give them 1-2 punch combo that will leave their competitors on the ropes or mat.

I'm going to wait for the D710 (my estimated price around $2300-$2400) and hope it comes out in April or July/August. BTW I’m upgrading to FX from a D300 which I plan to keep because I’m tired of waiting (8 yrs) for Nikon DX primes for studio portraits, etc. I already have the afs 24-70 f2.8, afs 50 f1.4 and af 85 f1.4 lenses.

If the D710 does not pan out by then, my plan B is to buy a D800 (what's $600-$700 more if you consider it can replace my D300 in DX crop mode 80% of the time and after all the reviews are in to form my conclusion.) or a new D700 or new afs 70-200 f2.8 VRII.

On the pc side I'm ready either way with a new Quad core CPU w/16 GB RAM, 120 Gb SSD for W7 64 and 3 Tb of RAID 5 storage that I'm building by hand.
Nikon Shakes Up D800 Worshipers AKA Its the Tripod Stupid! What can i say don't like carrying around a tripod let me repeat Its the Tripod Stupid!
+Thom Hogan re: Q. have you read the paper by Catrysse and Wandell on CMOS sensor design? i imagine your friends who are peer reviewing your writeup must have if you haven't. "Roadmap for CMOS image sensors: Moore meets Plank and Sommerfeld". it's a few years old but pretty relevant, i think.
I understand the moire issues relating to still photography. But I've read that if video is important, we should get the D800 as the D800E will be moire hell. This sounds melodramatic. But is it true? Is there really a difference in the frequency of occurrence of moire on video vs stills? Assuming of course, that we're shooting the same subject matter. Not peacock feathers or fashion fabrics but rather landscape, urban enviro on travels....etc, any help would be appreciated.
+olaf willoughby I get conflicting information from different Nikon sources on whether video would be impacted by the change of filtration. Given that we're talking about subsampled sensors here, the issue is whether there is a difference in the artifacts produced, as even the AA-ed D800 will produce artifacts. From a theoretical standpoint, I can't see how moire would change, however, I can see how stairstepping would become more visible. But what's unknown is what the EXPEED3 is doing with the subsampled data. I've seen ASICs that try to mask the subsampling artifacts and others than don't. I have no idea what the two D800's do until I can test them or find someone I trust who's tested them.
I'm leaning heavily toward the 800e. I have an older Kodak DCS Pro 14nx (no AA filter) I still shoot on occasion and have never experienced a problem with moire. Love the detail and look that sensor produces. Won't do anything until I see what some prints look like. Also very interested to see what they do w/ the D400. Assume it will be 24mp DX w/ D300 like features.
I've been lucky in that I got my D200 for "free" using Amex points, and my D300 for $200 by selling my D200 and some old equipment that I was no longer using. Amex is running a "sale" of sorts on their website through April 2, in that all stuff that you can buy with points is now 40% off (or 1.2 points/dollar). I was GOING to get a D800 with my points, but after some reflection, I think I'm going to get a couple of lenses instead. That way, I can wait and see what the reviewers say about the D800. If I choose to upgrade, I'd need to get new glass anyway. And if I don't upgrade, well, a 14-24 f/2.9 plus a 24-120 f/4 or 24-70 f/2.8 will work JUST FINE on my D300.
What i find amusing is that you read all of the Nikon users complaining about the D800 having too many megapixels and now all of the Canon users are crying not enough megapixels in the new 5d MKIII. It's seems like the Nikon users wanted the MkIII and the Canon users wanted the D800!
hi Thom, you posted that "it looks like a D800 is both a D700s and D700x all in one". The implication is that after downscaling the D800 36mp image to 12mp the resulting image would be as good as a D3S image. Is the D800 noise handling really that good? Not trying to challenge you (as I don't have the camera), but to understand if that is indeed your view (as a prospective buyer). Thanks
Add a comment...