D810A one from left field

<p>I see <a href="http://www.nikonusa.com/en/Nikon-Products/Product/dslr-cameras/D810A.html">Nikon has announded a D810A </a> a specialist body for astrophotographers.</p>

<p>I was not expecting to see such a useage specific body to come from any of the major manufacturers....</p>

<p>Out of left field?</p>

<p>Canon introduced the EOS 60Da back around 2012 specifically for astrophotography:<br>

<a href="http://www.usa.canon.com/cusa/consumer/products/cameras/slr_cameras/eos_60da#Features">http://www.usa.canon.com/cusa/consumer/products/cameras/slr_cameras/eos_60da#Features</a></p>

<p>Presumably because there are enough astrophotographers modifying their cameras that major players thought it might be a worthy niche market. </p>

<p>See also the discussion on this camera in <a href="http://www.photo.net/nikon-camera-forum/00d7PI">this thread</a>. Given that the second thing I photographed with my new D810 was the night sky, it seems useful to me, although I don't personally want one. And Nikon did otherwise seem to turn up to CP+ without much to announce, so one could argue that this was low-hanging fruit.</p><div></div>

<p>I used to work with some guy in the Silicon Valley. He is really into astronomy and astrophotography. He bought some land in Oregon because that area has the least amount of light pollution (in the entire continental US??); i.e. the sky there has the least amount of artificial lighting that affects viewing the stars. He drives over there (about 10 hours) a few times a year. His annual vacation is to watch things such as the transit of Venus, some eclipse ... around the world.</p>

<p>Last time I saw him a year or two ago, he was using a D800E for astro work. The D810A would be perfect for him. I am not sure how big that market is, but there are people who are really into astrophotography. I find it a bit strange that Nikon is willing to produce niche products such as the Df and D810A but wouldn't continue to D200-D300-D300S product line.</p>

<p>At least in the case of the D810A, it is a fairly minor modification from the regular D810 so that Nikon doesn't need to spend a lot of extra R&D money to produce a separate camera.</p>

<p>Knowing Nikon's other optical products, notably microscopes/spotting scopes etc, I'm surprised they didn't go down the same route as the Fuji IS Pro ie. making it inherently sensitive from ~350 - 1000nm and applying <strong>external</strong> filters to limit spectral response where needed. </p>

<p>There's a large market in Forensic type photography and the ability to shoot UV-VIS-IR is very handy in a single body option. Of course you'd need a lens such as the 60mm Coastal Optics to make it a single body/single lens combo..the wonders of a truly modern APO lens..:-)</p>

<p>It would <em>seem</em> that this camera is fully IR shielded, although until they produce a Spectral Response graph, who knows?</p>

 

<blockquote>

<p> so one could argue that this was low-hanging fruit.</p>

 

</blockquote>

<p>Indeed, it's also a very minor amount of work to do to produce such a D810 variant. Maybe there's an IR variant in the works too?</p>

<p>I doubt Nikon are going to be making many more SKUs of the D810 - it's in their interests not to flood the market with minor variations.<br />

<br />

Arguably a more interesting variant would have been to leave off <i>all</i> filtering, and leave it to external filters. Particularly, removing the Bayer filter set. That would satisfy both the astronomical community (who often want their own colour filters anyway) and those of the "Leica Monochrom" persuasion. It would also increase (a little artificially) the effective resolution and dynamic range compared, and counter the 5Ds announcement a bit. <i>But</i> I suspect it would have involved a bigger change to the sensor construction (since the RGB filters are more integrated than the IR cut-filter, as I understand it), so it wouldn't have been the "low-hanging fruit" that I suggested. It would have been a bigger change than the difference between the D800 and D800e, for example, whereas the IR filter change is very much in that category.</p>

<blockquote>

<p>At least in the case of the D810A, it is a fairly minor modification from the regular D810 so that Nikon doesn't need to spend a lot of extra R&D money to produce a separate camera.</p>

</blockquote>

<p>Shun is right here. With a small move, Nikon got all the press that would normally be accorded a new camera design for free. The FX bodies are profitable, and astrophotographers, who I suspect are by-and-large well heeled, will have to buy lenses to use with it, which is where Nikon makes a good profit. It will continue to whet people's appetite for the D810. </p>

<p>DX users, by contrast, are usually either "kit lens only" or "kit lens plus one" customers. [i think it was Thom who came up with the 1.7 lenses per DX user metric.] Perhaps one day, but only if the numbers are there, they will come up with a D9000 or the equivalent. The fact that they haven't suggests to me that their ongoing customers are upwardly-mobile and aspire to FX, but I'm not certain.</p>

<p><em>I'm surprised they didn't go down the same route as the Fuji IS Pro ie. making it inherently sensitive from ~350 - 1000nm and applying <strong>external</strong> filters to limit spectral response where needed.</em></p>

<p>The 14-24/2.8 is highly regarded for wide field astrophotography. As we know the 14-24 doesn't take any "normal" size filters and interference based filters might not even be possible to make work on this lens. Nikkors use such a wide array of filter sizes that a camera based solution (with modified spectral response on the sensor) may be preferable for users with existing collection of Nikon mount lenses who want to get better results in the imaging of nebulae.</p>

<p>I agree with Luke that these cameras (Df, D810A) are possible to make because their buyers are not as price sensitive as DX users so even if they are made in small numbers, a profit can be made. Also these cameras do not introduce technically demanding new features that would be expensive to develop. Some astrophotographers spend a lot of money on equipment and Nikon can expet some follow up lens sales to these customers if they're attracted to the D810A and D810. Finally the deep field astrophotographs can be extremely impressive and have some PR value that may help Nikon market their brand.</p>

<p>I believe the reasons for no "D400" are mainly technical. Autofocus tracking of an approaching subject at 10fps frame rate with each frame so well in focus that the 24MP DX resolution advantage is apparent in the resulting images is very challenging especially if a very wide aperture is used. Also it may be that the Nikon 24MP DX sensor achieves its dynamic range in part by utilizing slow readouts and a fast camera operated at high speed may turn out a bit noisier results. (Remember the D300 slowed down to 2.5fps for 14-bit NEFs.) I think this is a minor issue and the focusing at high speed is the real challenge. If the technical issues can be solved, a product is likely to be brought on the market.</p>

<p>I got a simple (not computer controlled) tracking mount to use with my cameras. On Monday there was some gusty wind and I noticed even with a 135mm lens that my image was vibrating a bit (all the way zoomed in in live view) so the results were not optimal. It made me wonder what kind of a mount would be needed for a proper telescope.</p>

<p>I noticed in Nikon's D810A promotional video that the Takahashi FSQ-85ED was one of the less expensive of the telescopes used to make the video, with weight 4kg. With tripod, equatorial mount, and counterweight (in a kit by the manufacturer) the total weight was about 23kg not including the camera or a computer. ;-) Now, it is not surprising if the results are impressive! However, it really dawned to me that if this is really equipment for amateur astrophotographers, they must be very serious about it indeed. ;-) The more expensive 5600mm telescope used in the video seemed to be made to order and I found a price of $160000 for this type of a telescope. I don't want to think about what the weight of the mount for that telescope would be or how it might be transported. I guess you need a permanent residence for it in a low light pollution area and then you make use of it when you can. I read that people can take months to capture the constituent images that are coregistered and averaged to make a final image. I'm not quite prepared to do that even though I've invested a lot of time in photography I can't see myself spending so much time to capture the data required for a single image. But I see the results very fascinating.</p>

<blockquote>

<p>I believe the reasons for no "D400" are mainly technical. Autofocus tracking of an approaching subject at 10fps frame rate with each frame so well in focus that the 24MP DX resolution advantage is apparent in the resulting images is very challenging especially if a very wide aperture is used.</p>

</blockquote>

<p>That means you believe Nikon is seriously lagging behind Canon technically, at least in that area. It took Canon over 5 years to introduce a 7D Mark II, but that can do 10 fps at 20MP with a $1800 price tag.</p>

<p>Back in 2007, the D300 was able to do 8 fps at 12MP. That was 7+ years ago. I would imagine that if Nikon wanted to, they could have easily updated the D300 to 16MP or 20MP while maintaining at least 8 fps some time in the last 4, 5 years.</p>

<p>I doubt photographers would be very hung up on whether it is 8, 9, or 10 fps or 16, 20, or 24MP. Those differences are rather small.</p>

<p>I believe the performance of the Exmor sensors is based in part on a slow A-D clock, which is sub mHz, at least in its earlier instances (D300/D3x). This might be a rate limiting factor.</p>

<p>The D4/Df sensor uses an on-sensor A-D to reduce pre-conversion noise, but only 24 channels, and a relatively fast clock, which makes the number of pixels a limiting factor. It is noise-limited at low ISO values, but benefits from a full well capacity of approx 120e- at ISO 100, which gives it a stop or more DR advantage over its predecessors and the Canon sensors. </p>

<p>Nikon and Canon have their own respective advantages over each other technologically; in general Nikon seems to have better sensors in terms of colour sensitivity and dynamic range but at the moment Canon has some advantage in fps and AF sensors which have more cross type points than Nikon (also they have PDAF on the main sensor in some models such as the 7D II for better video/live view AF). On the other hand, in conjunction of the 7D Mark II review, dpreview.com noted that the D750 performed better in their automatic subject recognition and tracking testing when the subject moved across the image area than the Canon 7D II. So it depends on what you're trying to do with the camera, which manufacturer has the lead.</p>

<p>Nikon clearly made a major shift in strategy around 2010-2 from high fps, moderate resolution cameras towards high resolution, moderate fps models across the lineup (with the exception of the D4s which remains a high speed model). Maybe the sensor designs used and the processing architectures of Nikon's current cameras are not easily adopted to high fps (without developing a new sensor just for the high speed DX camera). I think prior to the 7D Mark II, Nikon probably planned on continuing the D7100 series and not the D300s, but the 7D Mark II changes things and Nikon must have put the D300s successor project back as a priority project. Maybe an annoucement can be expected at around the same time as the D5 is planned to be launched.</p>

<p>A lot of people seem to think that there has been no D400 because Nikon is greedy and wants users to pay $6000 for a D4s and longer lenses. I think it is more likely that Nikon felt it was necessary (in response to customer feedback) to make a major effort on high resolution to compete with Canon (in which I think they succeeded) and the high speed prosumer models were left as low priority projects without the funding and resources needed to succeed. Note that in some Canon forums there are reports of 7D Mark II users who had to go through a number of samples before they got a camera that focused properly. So even with Canon's resources there can be difficulties in manufacturing a consistent, high precision, high speed camera at relatively low cost (compared to the cost of D4s/1DX). Still, whatever Canon does, they seem to sell a lot of cameras, no doubt because of their comprehensive lens system and established user base.</p>

<p>Nikon has some in-house resources for sensor development and they collaborate with a number of manufactures of sensors but I suspect they need to use the same sensor in multiple cameras to increase economy of scale in manufacturing. If they must (co-)develop all of 1) image quality optimized DX, 2) image quality optimized FX, 3) high speed optimized DX, and 4) high speed optimized FX sensors (three out of four they have) and corresponding processing and control circuitry for the cameras then surely they must get stretched on resources. I think Nikon had saturated the market with their high fps enthusiast models around 2008-10 and the customer reaction was overwhelmingly in favour of more resolution. Thus the shift in strategy and neglect of high speed enthusiast cameras. Whichever category of camera Nikon neglects, there will always be complaints. Personally I think FX works better for action than DX but I live in a country where half of the year is very dark and I'm much more likely to be at ISO 2000 to 6400 than ISO 400-800 for photographing people at events. And I happen to like shallow depth of field a lot, so FX works well for me. I think perhaps instead of the D800 Nikon should have developed the D750 first, and a D400 with 16MP sensor (from the D7000) and 8fps capability. This would have been the logical succession at the time. Instead they went all out on the resolution, getting the impression from customer feedback and sales information that high fps doesn't sell in this class of camera. Right or wrong; I don't know. I would have preferred the gradual development approach. But I'm fairly sure that there is a lot of people who think there should be another breakthrough camera instead of refinements to existing ones. The D8x0(E) was clearly such a camera for some people but it is not the right camera for everybody. I find it is quite ironic actually that there is such a push for high resolution in a time when most images never get printed at any size but are viewed on low-resolution monitors and 99% of the information is thrown away for most applications. I do make prints, in fact thousands of them every year, and I do the printing myself. But rarely would I see any difference between 36MP and 24MP FX, but I notice it quite severely in my editing time and storage requirements. In fact I'm moving to a JPG+NEF based workflow where I try to get the in camera settings as close as possible for usable JPG images that I can edit quickly and then do the difficult cases from the NEF files. This is a result of Nikon's abandonment of the Capture NX2 software which I liked to work with. The improved monochrome settings with filter options in camera do make it easier to get good in camera JPGs from events that I shoot. Still I would prefer that Nikon either make a deal with Google regarding further development of NX2 or put some additional resources on the development of NX-D in order to make it into a viable alternative. With 36MP there is so much excess resolution in the images that even the JPG images are quite flexible in how they can be used. With 12MP in camera JPGs I felt there was a significant detail loss not to mention the editability of colours and tones. For a long time I shot in colour picture control active even when I knew I was going to go for a black and white image because the filtration of the in camera monochrome settings was not what I preferred and Capture NX2 provided excellent black and white conversion which I haven't been able to match the quality of and ease of use using Adobe software. Some progress, some things go backwards.</p>

<p>There is one important feature which Canon has developed and I would urge Nikon to put in their next cameras. I shoot a lot in available artificial light (under fluorescent and compact fluorescent lights) and the anti flicker feature of the 7D II and 5Ds is something I would genuinely appreciate. Without this feature, the exposures and colours vary from shot to shot leading to a lot of colour correcting work in post-processing and more noise than should be there due to underexposure of some shots. In practice I tend to make black and white images from these situations instead of colour, but the people I am shooting for would often like colour versions as well. I hope Canon hasn't patented the technology in such a way that prevents other manufacturers from doing something similar.</p>

<p>BTW I agree that 16MP, 8fps DX would probably be fine for many <em>photographers</em> but a lot of people seem to be hung up on specifications (unfortunately) and motivated to buy/upgrade when specifications show a leap. I think Nikon cannot ignore this if they want to make a commercially successful "D400". A 16MP 8fps DX would have been perfect in 2012. Three years later, expectations on specifications have advanced even if the real world performance of such a camera would be perfect for many action shooters. I think 24MP 8fps would be ok today, or 18-20MP 9-10fps. Even a 16MP 8fps model might be fine but the price would have to be somewhere intermediate between the D7100 and 7D Mk II prices. Either which way I am sure that the 7D Mark II has pushed Nikon to work hard on some interesting new high end DX cameras that hopefully satisfy user needs.</p>

<blockquote>

<p> I think it is more likely that Nikon felt it was necessary (in response to customer feedback) to make a major effort on high resolution to compete with Canon (in which I think they succeeded) </p>

</blockquote>

<p>If (BIG IF!) pixel count counted as bragging rites, then Nikon have just sunk without trace. Not saying I want a 5Ds either!</p>

<p>I think Nikon should make more 'niche' cameras like the D810A. I'd guess the marketing budget for the D810A was much higher than the R&D to make it happen!</p>

<p>Just noticed the D810A has a dark-frame subtraction in RAW, I never thought about it before, but is that new or not????</p>

<p>I must say, I quite like Andrews' Bayer-Free monochrome sensor. It fits into the 'Colour is an Artificial Construct' way of thinking...:-)</p>

<p>D810T (for Thermal) anyone? I gather an FX thermal sensor might be a <em>little</em> bit expensive though! DX maybe?</p>

<p>+1 on the de-flicker. </p>

<blockquote>

<p>Just noticed the D810A has a dark-frame subtraction in RAW, I never thought about it before, but is that new or not????</p>

</blockquote>

<p>Long Exposure Noise Reduction (LENR) has always included doing a dark-frame subtraction. Very necessary on the Exmor cameras. Have they changed the settings on this in the D810a?</p>

<blockquote>

<p>I found the info about it happening specifically in RAW here..<br>

<br>

http://www.dpreview.com/articles/5786202951/nikon-d810a-astrophotography-camera<br>

<br>

Additional features include a Long Exposure M mode that provides selectable shutter speeds from 4 to 900 sec (15 minutes), before having to use bulb or time mode. A virtual exposure preview in bulb and time settings, similar to Olympus’ livetime mode, lets users essentially see in the dark, so that they can frame and focus.<strong> And a dark frame subtraction mode (called ‘astro noise reduction’ in Raw), should help to keep noise in check when aiming for the stars.</strong> Other changes come in the form of a red virtual horizon, so as not to interfere with night vision, and the ability to dim the viewfinder’s OLED. <br>

</p>

</blockquote>

<p>I'm not entirely sure what a virtual exposure preview in B or T settings is either?<br>

<br>

I quite like the Red Virtual Horizon and the dimming OLEDs.<br>

</p>

<p>Hmm. Is "dark frame subtraction mode" only available in "Long Exposure Mode"? I've always thought that dark frame subtraction should be a full-time option with this camera. The medium format cameras that use the Sony sensor, at least the PhaseOne IQ-250 that I know of, does a mandatory dark-frame subtraction on every exposure, no exceptions.</p>

<p>Luke: I'm actually quite sympathetic to that, especially with the apparent thermal noise on the D800 at high ISO settings. I've done a very short exposure at very high ISO that would really have benefitted from a dark frame. I care, if anything, <i>less</i> about long exposures, because I'm more likely to do one 30-second exposure for multiple shots. And thank you for sharing the technical details about the sensor.<br />

<br />

On Nikon's ability to hit a faster frame rate, as I've noted before:</p>

<ul>

<li>D4, "Expeed 3", 11fps, 16MP, 176MP/s</li>

<li>D800, "Expeed 3", 4fps, 36MP, 144MP/s</li>

<li>D610, "Expeed 3", 6fps, 24MP, 144MP/s</li>

<li>D810, "Expeed 4", 5fps, 36MP, 180MP/s</li>

<li>D750, "Expeed 4", 6.5fps, 24MP, 156MP/s</li>

<li>1Dx, "2xDigic 5+", 14fps, 18MP, 252MP/s</li>

<li>7D2, "2xDigic 6", 10fps, 20MP, 200MP/s</li>

<li><strike>"D400", 8fps, 24MP, 192MP/s</strike></li>

</ul>

 

<p>There is some evidence that the "Expeed 4" is a minor rehash of the Expeed 3. Not least that the one in the D4s has the same part number printed on it as the one in the D4, but also that the additional capabilities seem pretty minor for a "new chip". (I'm reasonably sure there was a change, but I expect a metal fix or at most a process shrink.) All the FX Nikons top out within a reasonable range of data transfer, and it's much lower than the Canons. I'm perfectly happy to believe it's not the Expeed, and is in fact the sensor (other than Sony's ability to hit high frame rates with a similar 24MP DX sensor), but there's a limit here somewhere, and the numbers are suspiciously similar. The shutter on the D800 and D810 can clearly go faster, because it does in crop mode.<br />

<br />

Hitting 8fps with a 24MP body requires more pixel pushing power than Nikon have so far put in a DSLR. Let alone hitting 10fps. Yes, at 16MP, they could do it, but since 16MP from a DX body offers no reach advantage over a D810, that might be a harder marketing sell. Though it probably wouldn't have been a few years ago, I agree. Roll on, the Expeed 5 (hence Ilkka's "around the same time as the D5", I guess) - or whatever the limiting factor is.<br />

<br />

(And yes, I know the 1 series manage much larger pixel rates. I have to assume something different is going on in the processing.)</p>

<p>I don't think that the DSP engine, whether Expeed 3 or 4, is a rate-limiting factor. I think in the case of the D4/s the combination of the physical shutter and mirror are practical rate-limiting factors, and to a lesser extent, the sensor read. </p>

<p>As for the Sony sensor cameras, the entire point of putting an A-D unit at the bottom of every column was to allow for slow, sub mHz readout in parallel. If the Sony sensors could be read any faster without a corresponding degradation in quality, Nikon could and would provide as much DSP as is needed to handle the limits of the sensor.</p>

<p>Andrew, I took liberty and put a strike on your "D400" listing since that is a non-existing camera. In the future, if someone finds your post via some search engine and could be very confused by the specs you posted.</p>

<p>I see certain rumor sites continue to spread rumors on some D400. IMO they would only bring more disappointment and frustration to those who are still unwise enough to wait for such camera. Looking back, these threads from a few years ago look silly:</p>

<ul>

<li>Nikon D400 specs, posted from Ukrainian/Russian website http://www.photo.net/nikon-camera-forum/00X4A5</li>

<li>Is D400 definitely coming? http://www.photo.net/nikon-camera-forum/00bh47</li>

</ul>

<p>Belatedly, no problem Shun. Sorry to have caused confusion - I'd deliberately put it in quotes, but I agree that may not have been enough to indicate it's fictional nature. Honestly, I'm sure Nikon <i>considered</i> a camera that people might have thought of as a "D400" - big companies consider a lot of products that don't hit the market - so articles based on spotting prototypes may not have been complete nonsense except in retrospect. Considering a product and having it ready to ship are two very different things, however.<br />

<br />

Luke: I trust that you have more internal knowledge on the sensor workings than I do. I'm still a little surprised that there isn't more differentiation between the Expeed 3 and 4, even if pixel read is not included - but I'm a bit influenced by the tear-down which indicated the same part number on the D4's Expeed 3 and the D4s's Expeed 4. Sony <i>do</i> manage to read the same sensor faster (the A6000 can do 11fps from a 24MP crop sensor), but I can't argue that it seems to lose a little accuracy compared to Nikon - though how much of that is the frame rate trade-off and how much is other processing, I've no idea. I suspect <a href="http://www.photo.net/nikon-camera-forum/00d8vV">those after a high-speed DX Nikon</a> would like Nikon to offer a different trade-off in at least one body, just as the D4 can't quite match the D800 at minimum ISO - and Samsung seem to manage 15fps with a 28MP DX crop, with some quality questions - but it might be a completely different sensor than the current DX bodies, and therefore not economical.<br />

<br />

And yes, the D4s may well be limited by the speed the mirror moves. I don't claim there's much practical difference between the D4 and 1Dx despite their frame rate numbers, but some of the other bodies do seem to be limited by more than the mechanicals, At least, the D800/D810, which gain speed if you crop the sensor. I imagine the read speed limitation more as applying to the high-end DX bodies (partly because they've got more pixels than the D4, currently), since the D300s shutter could mechanically hit 8fps.</p>

<p><em>Sony do manage to read the same sensor faster (the A6000 can do 11fps from a 24MP crop sensor),</em></p>

<p>I'm not sure which Nikon camera you're referring to? E.g. D5300 is similar but has better dynamic range than the A6000 (at ISO 100, 200); D7100 Toshiba sensor has better dynamic range than A6000 at ISO 100-6400. A6000 has some advantage at ISO 12800-25600. Different performance, different feature set (PDAF / no PDAF), different speed => what evidence is there that the sensor is the same? In any case the difference in dynamic range suggests that this technology is not well suited to high fps applications. A high pixel density APS-C camera is likely used at low ISO because the level of detail that is possible at low ISO (and that same detail is responsible for what advantage there is using the high density sensor for "reach") is smeared at high ISO by the in practice necessary high ISO noise reduction algorithms to get an acceptable to view image. So Nikon's choice of optimizing the image quality at low ISO seems to be very relevant in applications wishing to take advantage of the DX sensor's reach. It seems their choice not to offer high fps in DX is tied to engineering and image quality reasons. The window of opportunity to take advantage of the "reach" in a high fps APS-C model seems something that might be possible in bright (slightly diffused) daylight at ISO 400. If it is flat out sunlight then every bit of dynamic range you can get from the sensor is useful => but to optimize dynamic range, they can not offer high fps. At ISO 800-1600 and above the smearing of detail starts due to noise reduction, so e.g. D4s (800 and higher) or D810 (up to maybe ISO 2000) would give better results. This gives a very narrow window of lighting conditions and applications where high fps APS-C camera gives the best results of all camera types. I am sure such conditions exist, but do they justify making a separate camera just for these conditions and applications? By making DX slower and higher quality (higher DR) Nikon is expanding the envelope of the use of a DX camera for reach applications into brighter lighting conditions where high contrast is often a problem as well as improving the image quality of landscape type imagery.</p>

<p>Ilkka - sorry, I'm conflating the Sony sensor in the D7000 with this. I was really thinking of the 12fps 24MP sensor in the Sony SLT-A77 II (or original A77), but didn't want to conflate the sensor with measurements taken through a partial mirror. And perhaps it would be fairer to say that "other manufacturers, including some used by Nikon for sensors in the past, manage high frame rates at 24MP". Yes, they seem to have slightly worse high-ISO performance than Nikon, and I'll believe that this isn't a coincidence if you tell me so, but it's possible. I'm not saying the majority of DX cameras should make this trade-off, but a "sports DX" camera could.<br />

<br />

You make a good point about ISO 400 - I'm just not sure I believe you (at least a lot of the time). :-) At f/4, to make the numbers nice and round, sunny 16 suggests that an ISO 100 sensor can shoot four stops faster than 1/100s, or 1/1600s. Depending on your hummingbird, that's pretty fast - and easily fast enough to show rolling shutter effects. I'm not saying that nobody should shoot at 1/6400, but ISO 400 doesn't always seem necessary. Yes, light isn't always this good - I'm not denying that a D4 has uses for wildlife, especially having tried to shoot bats in Vancouver aquarium recently. But I'd not dismiss the crop just yet.<br />

<br />

But, of course, I'm a full-frame shooter, so what do I know?</p>

<p>Andrew where did I say that you'd shoot at ISO 400 for sunny f/16? I wrote "in bright (slightly diffused) daylight at ISO 400" I guess that could be interpreted differently but what I meant is typical daylight where there are clouds but not full overcast; f/5.6 or f/8 at 1/ISO shutter speed, or if you need 1/1000s for hand-holding and stopping action you're at f/4-f/5.6, ISO 500. Anyway at significantly higher ISO you lose a significant part the "reach" advantage due to noise and possibly low light induced focus errors (in indoor sports), and at lower ISO you would get best results with a camera with slow reads to get as much dynamic range as possible to manage the high contrast of direct sunlight.</p>

<p>I understand why Nikon would prefer to squeeze as much image quality out of the DX sensor as possible, to expand its optimal use envelope. I think it is ironic that they'd get so much criticism for supposedly stifling DX in order to sell D4s when what they're really doing is trying to make DX cameras that give the best possible image quality for the format size, in a broad variety of conditions.</p>

<p>Ilkka, agreed. I'm just saying that you imply that ISOs below 400 are rarely useful for shooting wildlife. I agree that it depends on circumstances, but I'm just suggesting that it's not so unrealistic to expect to be able to use these ISOs. Of course, if we're comparing (roughly) equivalent images for depth of field, you're shooting roughly a stop faster on DX anyway, gaining you back some ISO. 200-400 f/4 -> 120-300 f/2.8, 300 f/2.8 -> 200 f/2, 600 f/4 -> 400 f/2.8, etc. Yes, the faster lenses don't always exist, which is one reason why I shoot FX (135mm f/1.4, anyone?) and matching the aperture gets rid of most of the price difference, but it's not so unlikely. It depends what you shoot - I claim f/4 isn't so preposterous at medium range, but obviously if you're looking at a 500mm then it may be more wide open than you want. Even though everyone is exposed to both conditions, many might buy a camera because they know that <i>most</i> of the time, they'll be at lower ISOs. Which is why I got a D810, not a D750. Besides, direct sunlight is when I find myself needing dynamic range - if the light's diffused, I often don't need the same shadow recovery.<br />

<br />

I completely agree that Nikon needs to offer DX cameras with the best possible image quality, and it would be bad to replace the D3x00, D5x00 and maybe even D7x00 with a camera that has worse image performance - not that it seems to hurt Canon all that much. The criticism is only that, by offering image quality at the apparent cost of not offering (a bit) more frame rate and buffer, Nikon have a gap in their capabilities for which some people would pay them a premium. It would require careful marketing to explain why a D300 successor produced worse (but still probably very good) images than a D7100, but not necessarily more than needed to explain to your manager why a D4s is lower resolution than a D610. And - while I think it would be a harder marketing problem to sell, especially given the 7D2 - enough people have mooted a 16MP high speed camera that clearly there are those who aren't putting image quality first.</p>