Ophrys Photography

---

Wildlife photography by John Devries, Kent UK. Inspirational images from nature.

Let's get a couple of things straight. Firstly I have no allegiance to any camera manufacturer, I just want the best tool for the job. I have always used Canon cameras, not because I am a "fan boy" to use the expression used on many internet forums, but because (until recently at least) I have always considered Canon to be that bit better than Nikon for my needs and also I have a very heavy investment in expensive "L" lenses which would make me think long and hard before switching brands. Secondly, this article is not meant to be a review or a scientific test, I'm certainly not qualified to comment on heavy technical aspects about these cameras - there are plenty of other technical reviews full of graphs, measurements and formulae written by folks much smarter than me, that will do that if you require it. Instead, I want to offer my practical opinions as a photographer based on using two, great (in all sense of the word) full-frame cameras from the top rival brands - namely the Canon 1Ds MK II and Nikon D3.

Is this a strange choice of models to choose to compare ? Maybe- after all, the 1DsmkII is not even a current camera - it was superceded by the 1Ds mk III some time ago and the D3 is still current. I do not have access to a 1Dsmk III so that kind of rules it out. Also it has a 21 Mp sensor vs the Nikon D3's 12Mp - so comparing images becomes problematic unless you interpolate them up or down to match size. Finally the price of the 1Dsmk III flagship is much higher - the 24Mp Nikon D3X is the obvious competitor in price and performance. If you would like to read a D3 vs 1DsmkIII comparison click here.

The 1DmkIII that I usually use matches the D3 far better in both price (both around £3100 currently) and also features ( they have similar Mp counts, 10 for the Canon and 12 for the Nikon, both shoot at a very fast rate (9-10 fps) use a similar battery, both have advanced focusing systems, live view and so on. However, the1D mk III has a 1.3x crop sensor which gives a different field of view with the same lens at the same shooting distance, so once again it becomes difficult to do direct comparisons without resorting to image resizing. More importantly, I did not have a 1DmkIII available to test either, as I had recently written-off my camera in an accident on a trip to Scotland.


Red grouse - One of the last shots I took using my Canon EOS 1DmkIII plus 500mm f4 lens

Thanks to the accident, I now have the opportunity to put the insurance money towards replacing the 1Dmk III or to jump ship to Nikon and buy a D3- as some other sports and wildlife pros have done with no apparent regret - prompted mainly by the 1DmkIII's checkered history of autofocus inconsistencies. My 1DmkIII was the best focusing Canon camera that I have owned, but it did still throw up soft frames in most 10fps burst of images for no apparent reason. Sod's law says that the best frame would be the soft one. I wondered if the Nikon really could do better.
I decided that the only way to find out was to hire a Nikon D3 ( from Jacobs in London) along with a couple of lenses that I also happened to own Canon direct equivalents of - namely a 100mm f2.8 macro (vs the105mm Micro Nikor ED VR, ) and a 70-200 f2.8 IS (vs a Nikon 7-200 f2.8 AF-S VR.)
I also own a 1DsmkII as a second camera which also has a full frame sensor and a few more pixels than the D3 (4.6 more to be precise). I have always rated the image quality from this old girl as being state of the art. Perhaps the latest 1Dsmk III and 5DmkII with 21Mp each are better still - I don't know, I haven't used them.

So, for better or for worse, the 1DsmkII is what I used for the comparison, but you will also find the article littered with references to the 1Dmk III too. I appreciate that the 1DsmkII can't compete with the much newer D3 in terms of frame rate (it can only manage a 4 frames per second) and has a pathetic viewing screen which looks like an out of focus postage stamp against the D3's wonderfully bright 922,000 pixel screen. It also has a horrible dual button press system which was intended to prevent the camera accidentally being knocked onto another setting (which it did) but also made operating it fiddly and irritating. Also the menu systems are far less intuitive than the D3 or mk III cameras. Despite all this, the IDs II autofocus system has never come in for the criticism that the 1Dmk III's has received - and it continues to produce wonderful quality results..I was now all set to do a comparative full frame shootout , " Whoopie - at last ! " you say.

I won't bore you with all the test details I employed throughout this comparison - but trust me, I have treated both camera systems identically in each comparison - e.g. same tripod, cable release, shooting distance, aperture, ISO etc and identically post-processed all the RAW images. As the cameras are both full frame and have similar numbers of pixels, I have done no resampling up or down of any images - so what you see is what you would get if you used these cameras. The images from the action sequences were taken hand held - both with the image stabiliser (IS) or vibration reduction (VR) systems switched on for each lens using sufficient shutter speed to freeze motion and eliminate camera shake.

Round 1 Image quality

We all think we know what image quality means, but when it came down to it, I had trouble defining it. A boffin will probably be able to do this readily, but I felt that it must come down to a number of factors that I have considered here:

The amount of detail that the camera plus lens can resolve (Resolution), how it reproduces colour, in terms of accuracy, brightness and intensity (Colour rendition) - although with a RAW converter you can adjust colours infinitely anyway. How sharp the image appears (presumably a combination of resolution and micro-contrast at the edges) and how much sharpening the image will tolerate in post processing before degenerating into artifacts such as halos etc. The ability to reproduce all the tonal values in a scene from the very darkest to the very brightest - thus being able to show good detail in both the shadows and the highlights (Dynamic range). Finally the amount of unwanted granularity (noise) present in an image at each ISO setting (Noise performance). I have not investigated vignetting, distortion, ghosting, edge sharpness, diffraction etc as these are probably more attributable to lenses than cameras, although at this level of excellence and at the middling apertures that I employed, I just didn't notice them anyhow.

- Resolution

Roof tiles- starting image


Roof - 100% crop. Canon 1Ds mkII 1/500sec f8 ISO 200 70-200mm lens at 100mm.


Roof - 100% crop. Nikon 1/500sec f8 ISO 200, 70-200mm - lens at 100mm


Conclusion - Look at the lichen on the roof at the bottom left. There is a just a tad more detail in the Canon image. Not much, but it is there. This is probably down to the extra 4.5MP of resolution or possibly lens differences.

Irises - starting image using 70-200mm lenses at 200mm, 200 ISO f8


Iris 100% crop - Canon (look at the pollen grains on the petal)


Iris 100% crop - Nikon



Iris Starting image using 100mm (and 105mm ) macro lenses. ISO 200 f8


Iris group 100% - Canon. Look closely at the texture in the closest drop petal



Iris group 100% - Nikon. More saturated blues than the Canon

Conclusion - In each case there are tiny differences in Canon's favour. These differences are hard to illustrate here at low resolution - even if you are using a high quality, calibrated screen - they are quite a bit more noticeable on the high-res images.

I have had extensive opportunity to explore the image quality produced from the 1DmkIII camera and have illustrated my findings in my Full frame vs crop article. In summary, I found it to be better than the EOS 40D but inferior to the 1DsmkII. I haven't been able to compare the 1DmkIII side by side with the D3, but as the 1DsmkII has been present in both comparisons as a control, it seems reasonable to extrapolate that the D3 would come out mid way between the 1DsmkII and the 1DmkIII in image quality at low to moderate ISO but I'm sure it would be a very close call. You can draw more conclusions yourself by using the excellent "camera comparometer" on the Imaging Resources website. You will need to look at the still life images at equivalent ISO settings at full size. Also DP review have compared the D3 against the 1DmkIII and the differences are negligible.

So this test is won by the Canon 1DsmkII - just. But when not being viewed side by side, these differences become pretty insignificant between all three cameras mentioned. As I only had the D3 for a weekend hire, I did not have time to faff round doing the lens micro-adjustments that this camera (and the 1DmkIII) are capable of. The lenses were obviously focusing extremely well without the adjustments, but it is just possible that investing the time to do this tweaking might just close the image quality gap with the Canon 1DsII to zero.
I certainly found that my 1DmkIII required significant micro-adjustments depending on the lens fitted, but interestingly, my old 40D and the 1Ds mk II were fine on these same lenses without any adjustment (just as well as you can't adjust it yourself anyway) but an interesting/curious observation.

- Colour rendition
Both cameras produce accurate colours when the white balance is adjusted correctly. The Nikon has a tendency to produce more saturated images, a bit like Fujichrome vs Sensia films. The Canon image tends to look a little more "Sensia". If you shoot RAW and at low ISO, you can make both camera's images look very similar by playing around with the hue and saturation. I presume that this can also be done in jpeg via the picture styles, but I did not try.

I'll call this round a draw.

- Image Sharpness
Both cameras/lenses can produce equally sharp images, but you often have to work the Canon images a little harder in post-processing to get there. The Nikon RAW images look a little sharper and brighter straight out of the camera to me. Neither of the cameras produced obviously digital or plasticy images ( a Canon criticism levelled by Nikon fan boys apparently) and neither did they show artifacts when viewed at high magnification.
Any difference in sharpness after post-processing is down to good user technique. Both systems can produce excellent images.

I'll call this round a draw too


- Dynamic range

I have no way of measuring this other than subjectively. DXO have done some scientific studies on sensors and you can do some fascinating side-by side camera comparisons if you click on the DXO link. According to DXO, the D3 has a score of 12.2 against the Canon's 11.3. I don't know how significant this is in real world use as both cameras provided bags of shadow and highlight detail but both will clip highlights and shadows in difficult scenes such as a bride in a white wedding dress standing beside a groom in a black suit. Both cameras seem to clip intense reds and yellows very readily, making photographing a yellow iris or a poppy very challenging. The dynamic range of sensors used in digital photography is apparently many times less than that of the human eye, so although exceptional cameras, neither of them is yet adequate in this respect compared to your eye and brain's image processing system. This is why High Dynamic Range (HDR) photography is currently in vogue.

I did not have time to play with the active D-Lighting settings on the D3 which should theoretically dig out more detail at the extremes. It is not clear from reading the manual whether this only works on jpegs or is also relevant to RAW images though. The mk III Canon's shoot in 14 bit mode all the time, but the D3 can shoot in either 12 or 14 bit modes. The extra bits supposedly provide finer gradation of colour and shadow detail. I can't say I have noticed any difference (and nor have many other trusted photographers on internet forums) so I think that this must be marketing hype.

Dynamic range usually reduces with ISO, so image quality goes down as ISO increases. The 1DsII still holds up very well against the Nikon in this respect despite greater noise, but I am getting ahead of my self....

- Noise performance

Here are some noise comparisons. I have always felt that the 1Ds mk II has similar noise performance to the 1Dmk III, which is about a stop better than an EOS 40D or 1DmkII. The D3 has a reputation for being very good on noise performance - so let's see.

Starting image (Common-spotted orchid)

ISO	Canon 1Ds mk II	Nikon D3
200
400
800
1600
3200
6400
8000
10,000
12800

D3, ISO 12800 Image after Neat Image noise reduction - perfectly adequte for web-use at this size:

Conclusions on noise

What we are hoping for here is low noise, particularly in the shadows (look at the bottom right portion of the 100% images) and high detail retention with no smearing in the subject itself (as this would be indicative of obvious onboard noise reduction having taken place. No in-camera high ISO noise reduction was selected in either case.

At 200 ISO there is no obvious difference in noise between the cameras.
At 400 ISO the Canon image is excellent, but the D3's background already looks smoother with no loss of detail.
At 800 ISO the Canon is still perfectly useable (remember this is pixel peeping to the extreme) but the D3 is still looking as good as the Canon at 200 ISO (a two stop difference).
At 1600 ISO the Canon is showing obvious noise in the background, but the image would clean up well in a good noise reduction software such as Neat Image. The D3 is still producing totally useable images, but the first hint of noise is creeping in. Both cameras have retained good image quality at this ISO with no obvious evidence of detail smearing or loss of colour saturation
At 3200 ISO The Canon image is pretty degraded but would be OK for web-use or work where the image quality is not paramount (such as a newspaper article). I did not test the Canon at it's "H setting" which is an extended setting of 6400 ISO as I felt that the camera had already called "uncle". The D3 was still going strong at 3200 ISO and a whiff of Neat image would again give a publishable quality image if necessary.
At 6400 ISO there is obvious degradation present at 100% viewing but with care (exposing accurately to the right) the camera can produce good images at this extraordinarily high ISO that would clean up well in Neat Image. I would not use it at any higher ISO personally. I did not try setting the in-camera High ISO noise reduction on, but this would be interesting to try on another occasion.The Nikon goes on all the way to 25,600 ISO - which I didn't even actually try, but I expect it would give you a reasonable record shot if you were desperate.

What we have here in the D3 is a camera that I would definitely use up to 3200 ISO in everyday use in order to freeze motion if necessary or to maintain shutterspeed in low light. Even 6400 iso is usable with care if you use Neat Image to clean the image up a little.
This is an incredible achievement and is really exciting for the action photographer as this ability to up the ISO with gay abandon, combined with a Vibration Reduction lens (VR) which enables hand-holding at up to 3 stops lower shutterspeed could provide some unique opportunities. I could envisage setting the camera to a selected shutterspeed and using the auto iso facility on this camera set to 200 ISO minimum and 3200 maximum to photograph birds in flight. The camera would provide the maximum aperture possible to maintain depth of field by adjusting the ISO - wow.
Higher ISO settings also enable a high shutterspeed to be obtained when working at small apertures in macro work without running into subject motion blur or user/camera shake .I really enjoyed happy snapping macros using the D3 and 105mm VR macro at small apertures and high ISO.

So, all things considered this round 1 goes to the D3. The image quality is so close to the Canon, it matches it in the other factors, but it is supreme on noise performance. This opens up exciting new opportunities for the wildlife action photographer.

Round 2 Metering accuracy in matrix /evaluative metering mode

Again, no fancy measurements, just observations gained by taking pictures of real subjects and looking at the camera histogram. I have always felt that my Canon cameras underexpose by around 1/3 of a stop. On my 1DII and III, I often dialled in +1/3 exposure compensation in evaluative metering mode as a default setting. In good light the evaluative metering (that I always use) is very good, but on dull days or under difficult lighting conditions, I often have to dial in positive exposure compensation in order to "expose to the right ", rarely do I find myself applying negative exposure compensation unless purposely under-exposing an "arty image". I rate the 1DsII as being the Canon with the best and most reliable metering that I have tried, but I frequently override it.

The Nikon D3 seems to better fit with my style of shooting as it seems designed to consistently expose to the right whenever possible. If it gets the exposure wrong - it has usually over-exposed it a tad rather than under-expose. Although by shooting RAW, this can usually be recovered in post-processing. I must say though, that I photographed a wide range of difficult subjects - ranging from white flowers to small subjects against black backgrounds and I found the metering to be uncannily accurate - and hardly ever had to override it in 2 days of use. Impressive stuff.

This round clearly goes to the Nikon

Round 3 Autofocus ability in continuous/AI servo autofocus mode

Rob Galbraith (an accomplished US sports pro photographer) has become famous/notorious for taking Canon to task over the issues with the 1Dmk III. If you want to read his very lengthy, thorough but fascinating investigation into the 1Dkmk III AF please click here. According to RG the 1DmkIII does not focus as well as his 1DmkIIN camera. In it's latest submirror fixed/AF check and adjusted /updated firmware guise, he feels that the camera autofocuses just fine but it's Achilles heel is it's inconsistent tracking ability .

I would have loved to have put the 1DmkIII directly up against the D3, for the next test, but I had to make do with the 1Dsmk II (but which has the mk II AF system that RG prefers). I used the 70-200 zoom lenses handheld, set to 135 mm for these tests with IS/VR switched on. In test 1 I was able to achieve a shutter speed of 1/1250 sec at f4 , 1/2000 sec at f2.8 in test 2 and 1/5000sec at f2.8 in test 3, all at 800 ISO. Motion blur and camera shake would not be expected to come into the equation under these conditions.

I used and recommend using jpeg large/fine for these tests to maximise the number of shots taken until the camera buffer fills. The D3's buffer filled at 39 frames at 9fps (sounds a bit low - perhaps I had active D-Lighting or 14 bit selected which slows processing) and the 1DsmkII's buffer filled at just 15 frames, but remember it has larger files and an older digic 2 processor. The latest generation digic III - which is on the mk III cameras is much quicker, and the 1DmkIII can often fill a memory card before the buffer fills. Buffer capacity is very important when using the high speed motordrive for action - there is nothing worse than hitting the buffer and having to stop shooting at a critical time.

Dual Test 1 plain top f4	D3 Test 2 contrasty top f2.8	D3 Test 3 Car tests f2.8

In the dual camera test -test 1, the runner was wearing a plain top which is challenging to the AF system as it likes some contrast to get it's teeth into. The lenses were stopped down to f4, which provides a little more depth of field, and would be just a little more forgiving of AF errors than at f2.8 (lens wide open).

This was the dual camera comparison test, but I went on to push the D3 a bit harder and carried out two further challenges at an aperture of f2.8 - where the depth of field is very shallow and any focussing errors will be all too obvious.

In test 2 the runner changed her top to a higher contrast black and white shirt which should make things a bit easier for AF system. I used single AF point focus (centre point only) and the focus point was kept on the neck of the runner's shirt as best I could. I checked whether I had succeeded in Breezebrowser Pro (I had) using a useful feature that superimposes the selected AF point in red over the image.

In test 3 (which I would recommend including in any initial testing of a new camera purchase) is to get an assistant to drive a car towards you at a steady 30mph and focus on the number plate.

Results:

D3:
Test 1 - 1/39 images soft at f4 (97% hit rate).
Test 2 - 2/39 images soft at f2.8 (95% hit rate)
Test 3 - 39/39 images sharp at f2.8 (100% hit rate).

NB I should add that the soft images were still usable for non-critical work, particularly after sharpening a little more firmly in post process.

1DsmkII:
Test 1 - 3/15 images out of focus at f4 (80% hit rate) a further 2 images were soft (total 66% hit rate)

NB The two out of focus images were totally unusable

1Dmk III:
When I first got my 1Dmk III back from Canon after the latest round of AF checks and firmware, I did a quick test with my cat walking (not running towards me). I was a bit disappointed to find that there were still out of focus frames in the sequence as usual, although there were plenty of in-focus shots in there too to select from. Although nothing like as thorough or scientific my findings were therefore pretty much in line with Rob Galbraith's.

Update on Canon 1Dmk III AF performance - 19 July 2009

The acquisition of another Canon 1DmkIII enabled me to do a comparative test with the a car driving towards the camera at aperture f2.8 . This is equivalent to test 3 above in which the Nikon scored 100%.
The Canon 1DmkIII scored 88% and 70% if you include the slightly soft but still just useable images. Looking back at the images from the D3, I was surprised at how much better the image qualtiy of the Canon was when it did nail the shot.

Further update on Canon 1Dmk III AF performance - 21st July 2009

Finally I repeated test 1 (runner at aperture of f4). The Canon's buffer filled at 45 images vs the D3's 39 so there were a few more images to assess in the calculations. The results were that 2 out of the 45 were out of focus (4.44%) and a further 3 were soft (6.66%) or to put it another way, 89% of the images were totally sharp. This compared with 97% sharp for the D3.

So the Canon Af is much improved of late, but the Nikon D3 still definitely has the better tracking ability it seems.

Using the cameras to take real action shots

The D3 and 1DsII was used to take a series of images of my neighbour's beautiful horses, using the 70-200 IS/VR lenses. With the D3, provided I kept the autofocus point on the horse's neck at a good shutterspeed, I obtained wonderfully sharp images literally every time (if I managed to keep the AF point on the right place). Admittedly a horse is somewhat bigger than a bird in flight, the light levels were good and I was using an f2.8 lens - which should turn on the more sensitive cross-type AF sensors for better responsiveness. I tried centre point only (Single point AF) and 9 point expanded Dynamic AF modes with great success. I did not have time to explore the 21 or 51 point Dynamic area tracking, nor the the 3D focus tracking which utilises light and colour data in addition to the usual contrast detection, but they reputedly work very well. I recommend reading Moose Peterson's article if interested.

My impressions from memory against the Canon 1DmkIII are that the D3 does not lock on quite so fast as the 1DmkIII. With a fast-focusing lens, the 1DmkIII feels like an F1 car must do- the D3 is like a road-going Ferrari by comparison - very quick, but not so hyper-responsive.The story doesn't end there though, as once the D3 has locked onto the subject, it appears far more determined to hang on to it - the best analogy I can think of is that the AF feels as though it has a magnetic attraction to the subject once locked on - it's kind of more sticky. The 1DmkIII certainly appears more inclined to jump to the background than the D3 - particularly if it is a busy background - like trees and bushes for example.

Fast moving stallion with the D3



Mare with D3 - all feet off the ground - this is just too easy - yawn


100% crop of the above - pin sharp and bags of detail - look at those whiskers!


D3 : I'll give you some action matey !


IDs mk II image. On this bright day, the 1DsmkII AF was also pretty accurate and the camera had no trouble dealing with locking onto and tracking these horses. In low light, or with smaller subjects, obtaining these images would be much harder. It still threw up the odd inexplicable soft frame (like the 1DmkIII would) though.


D3: My cats did not escape my attention either. They were really going for it here and the camera did very well to get focus lock.The D3's high ISO ability enabled me to use a very high shutterspeed on a dull day to freeze the motion. The focus point hit the cat's back - which is unfortunately sharper than the head - but this was user error - the camera just did what it was told.


D3: This is Monty chasing a piece of string and looking very serious


I would love to use the D3 to photograph birds in flight - these are a serious challenge and really show up AF weaknesses very quickly.

So once again, this round must go to the D3 - by a fair margin. It's is not foolproof, but it has the most accurate and trustworthy system that I have yet encountered in a digital SLR.

Round 5 A life-long Canon user picks up a Nikon and uses it for the first time - likes and dislikes - ergonomics and handling

Out of fairness to Canon, the following comments relate to the mk III cameras - as the 1DsmkII is not a latest generation offering and is really showing it's age in terms of ease of use.

I have used Canon all my photographic career, so inevitably the Nikon would seem strange at first.
I had about an hour to flick through the manual on the train ride home from London to Kent, and after this I only referred to it two or three times again as the camera is pretty intuitive.

Things seem to have been done to purposely make it different to a Canon - the most obvious being the lens mount which does up anticlockwise instead of the Canon's clockwise. How nice it would be if the two brands used a common mount and interchangeable lenses. I'm sure this is a very intentional incompatibility for obvious reasons.. One funny thing - my Kirk quick release mount
which is tailored to fit the 1DmkIII fitted the D3 pretty well. This is pure coincidence as I even had to get a new plate when I changed from the 1Dmk II to the 1Dmk III !

On picking it up, the Nikon fitted my hand well and had a nice high-grip rubbery feel to it. The hand grip felt thinner and more angular the Canon - but that was OK.
I have average sized hands, but one thing I really didn't like on the D3, was that the shutter release button was too much of an upward stretch for my index finger. It is too high for my liking and makes me hold my finger at an unnatural angle which made it ache after a while. When I went back to the Canon it felt far more comfortable and relaxed to me.

I also missed the Canon's big command dial wheel on the back of the camera - it is just there where you want it to be for making exposure compensation adjustments. This is also true when the camera is in the vertical position. I dislike the Nikon sub command wheel on the front of the camera too. This does not protrude enough above the surface of the camera body and I find it to be set too low. Although I suppose you get used to it in time, I was finding myself having to keep looking up from the viewfinder to position my index finger on it.
I also disliked the fact that to make exposure compensation adjustments, I had to press (and keep pressed) the exposure comp. button and then rotate the fiddly wheel simultaneously. On the Canon you just turn that big wheel - albeit you can knock it accidentally I suppose. It's just as well that the Nikon nails the exposure virtually all the time, as I hardly had to resort to using these controls. It wouldn't be so bad if you could just give the wheel a prod and it stayed active for a few seconds while you make the adjustments rather than having to keep it depressed. The D3 exposure compensation control defaults to turn to the left for plus and right for minus. In countries that read left to right this is plain wrong. However this can be reversed in the menu custom functions, in fact I believe that you can reverse or reassign most functions to other buttons if you prefer. This makes the camera very customisable which is great.

The Nikon's ISO button is placed in a daft place on the bottom left of the camera requiring a two handed action to change ISO. The Canon requires a dab on the top-mounted ISO button and you then move the control dial - much easier.

Nikon has fitted lots of switches on the camera which prevents you having to dig into the menus for commonly used functions - a nice idea. I'm not too sure about the CSM switch though - you use this to select single shot, continuous servo or manual AF. It is a bit fiddly and if you wore gloves to shoot you would have to remove them to change positions.

I liked the fact that the release mode dial on the top of the camera had a lock button - this means that settings can't get changed inadvertently. Canon need to take note for the 40/50D - I frequently used to find that this button had changed settings upon removing it from a camera bag and I often had to waste valuable time troubleshooting what was causing the camera to behave strangely as a result. The mode dial also made changing from motordrive high/low/single shot, liveview and mirror lock-up a breeze.

I must mention the quality of the rear screen on the D3 as it is glorious. It is big, bright and very high in definition. It is much higher resolution than the Canon mk III screens and is now being imitated on the Canon 50D and 5DmkII. I expect we shall see it on all future screens. For the first time in history it is good enough to check focus of images on the camera. On the subject of image review, the Nikon has a strange system of zooming in - you have to press the magnifying glass button which brings up a yellow box in screen and then turn the command dial which frames a part of the image with teh box. You can then press the multi selector button to move the image around the screen. The Canon (and apparently the D300) seems much simpler. Additionally, the Canon can be partially zoomed in and then you can scroll through images at this magnification. The D3 system does not support this.

The histograms on the Nikon are much larger and clearer than the Canon ones - so it is much easier to read the in the field. Some information coming up on the three different inormation panels is difficult to decipher though - as the writing is so small and it is often difficult to think what it is actually telling you in Nikon speak.

Selecting one of the 45 AF points or the auto selection (ring of fire) on the Canon mk II cameras is achieved by first pressing the AF button on the top right of the camera and then using a combination of the big control wheel and the small aperture wheel to scroll through the points. This is not too bad a system but it is a little slow. The system on the mk III is just plain bizarre and frustrating. The AF points dance around in arcs and circles as you move them and I am ashamed to say that I have never mastered it.Fortuntely Canon came up with a firmware update which enabled the points to be selected more conventionally with the multi selector- thank goodness ! The D3 uses a similar method for getting around it's 51 points and even works diagonally to get to your chosen point quicker - excellent.

I like the fact that you can format a memory card without having to go into the menus. You just press the two buttons marked in red (exposure mode/format plus delete) to format the card. At least this is what is supposed to happen. The one glitch that I had when using the D3 is that I could not get it to format my Sandisk 4 and 8Gb Extreme III CF cards. I could delete images individually, but the cards would not format in camera. I tried deleting all files on the computer before formatting in camera - but it would still not have it - most strange. If you put a card containing Canon RAW CR2 files into the D3 it says "no image" , but the 1DsmkII actually played back Nikon NEF files without batting an eyelid !

Despite all the Nikon's abilities it lacks automatic sensor cleaning . This is strange, and has been rectified on the much cheaper but more recent Nikon D700. These systems are not perfect but they are pretty effective at maintaining a clean sensor once cleaned. The sensor on my D3 hire camera was filthy of course.

Although a seemingly minor matter the lens hoods on the Nikon are suerior to the Canon ones as they attach nicely and click into place - frequently with a release button. The darned hood on my 100-400 Canon is pathetic and it is always falling off !

Other considerations for the nature photographer

The shutter on the D3 is a nasty, noisy, clacky affair - just like the 1DmkIII and is sure to scare off any bird when shooting from reasonably close range even in a hide. The 1DmkIII does have one more trick up it's sleeve though - silent shooting mode.This only permits firing off a single frame at a time, but although it is not "silent" it is very much quieter than the normal shutter clatter. Similarly liveview shooting is very quiet with the 1DmkIII, even with the motordrive running. Nikon - take note please, this is important! I'm sure that wedding and sports (golf ?) photographers would appreciate this too.

XXD Canons and all Nikons will retain AF when using extenders down to f5.6. Canon 1 series cameras will AF to f8 (centre AF point only) which is a big advantage in telephoto work - bird photography for example when using f4 or f5.6 lenses instead of the big f2.8 jobs.
Nikon have a 1.7x converter which loses 1.5 stops of light and autofocus appears to be retained on the lenses with an AF motor built into the lens (S - lenses) despite the aperture dropping just below f5.6 as far as I can tell - but it is all very confusing to someone coming into this afresh .

Nikon really appears to have some serious gaps in their lens lineups. Canon is much stronger here as they have far more image-stabilised lenses and all Canon lenses have in-built motors and work well on both full frame and crop cameras. As Nikon are new to full frame, they have many lenses (DX) that work better on a crop camera. A Nikon strength is that you can use a DX lens on an FX (full frame) camera, but the penalty you pay is that the narrow field of view covers a smaller part of the sensor - so only 5 megapixels are available for use. This is useful for current DX owners as a stopgap but anyone coming to Nikon Fx for the first time should avoid purchasing a DX lens.
Nikon have a reputation for making exceptional wide angle lenses that are better than Canon's, but they appear to be seriously high end f2.8 beasties (read big, heavy and expensive) or alternatively consumer grade only (read low end quality and plasticky) - with nothing in between it appears. Where is a Canon 17-40 f4 equivalent for example?
Nikon make a fantastic 70-200 f2.8 VR, but where is the f4 equivalent ? There is a superb 300mm f2.8, 500mm f4 and 600mm f4 VR, but where is the lightweight Canon 400mm f4 DO equivalent or the cheap but good Canon 400mm f5.6 rival ? The nearest equivalent to Canon's incredibly successful 100-400 f4.5-5.6 IS is the 80-400 4.5-5.6 VR. OK, but why would anybody in their right mind release this lens without an in-built motor to provide fast AF ? This lens depends on a slow mechanical drive from the camera instead! The 200-400 f4 VR is said to be a superb lens rivaling the prime lenses, but is a tad short on focal length for bird photography - particularly on a full frame camera, so extenders are necessary. The 1.4x will retain AF, but I'm not certain whether the 1.7x will on the D3 - I think it does but I'm not certain as the effective aperture drops to f6.3, which is smaller than the usual f5.6 cutoff. Unfortunately as versatile as this lens is and as good as it is optically, it is still very big and heavy - almost as heavy as a 500f4 in fact. The 300mm f4 looks interesting, as it is a ahrp but compact S lens, but it lacks VR - and the list goes on....

If I were to change to Nikon, I would struggle a bit to find a range of lenses to cover my all of my needs I think. For the record, on an Fx camera (D3/D700) I would be tempted to go for a 17-35 f2.8 or the superb 24-70mm f2.8 at the wide end, a 70-200 f2.8 plus 1.4x or 1.7x extenders for a midrange zoom , a 105mm f2.8 VR macro for closeups (plus a converter to extend focal length - or go for a Sigma 180mm macro as well if funds permitted). Now it gets tough - what do you use for birds in flight ? A 300mm f4 plus a 1.4x extender - or a 300mm f2.8 plus 1.4x or 1.7x extenders (pretty heavy) ? Perhaps the 70-200 f2.8 plus a 1.7x extender would be OK - but I don't know about AF speed.For a super telephoto a 500mm f4 would be superb for birds, but a 200-400f4VR would be more versatile for larger subjects like mammals. Some tough choices to make then.

Overall Conclusions

The Nikon D3 is probably the best camera currently available for shooting wildlife, particularly action. It's greatest strengths are :

Exceptional performance at high ISO with the benefits of high shutterspeed
Very accurate autofocus
Very high image quality
High frame rate (9fps)

The downsides are few, but it would be nice if it had:
A few more megapixels to enable harder cropping and to make up for the lack of crop factor for telephoto work, but this must not be at the expense of image noise (maybe this is not yet possible with current technology).
A quieter shutter or silent mode so as not to scare wildlife away
It would be nice if Nikon adopted Canon's big command dial thumb wheel and moved the shutter release to a more comfortable position - but perhaps it is just me this annoys.
It would be nice if the camera retained autofocus to f8 like the Canons. This would enable AF to be retained when using teleconverters on lenses slower tha f2.8.
It would be nice if the liveview screen articulated to enable ground level and overhead shots without resorting to an anglefinder. I'm sure this will come.

Nikon also need to plug the serious holes in their lens line up.

Canon need to crack the autofocus tracking problem on the mk III cameras completely to everybody's satisfaction, they are nearly there. The D3 needs to be used as a benchmark to equal or preferably beat. The use of colour information in addition to phase detection (like the Nikon 3D tracking) might improve the AF tracking further. Canon has too many AF custom functions. All we want is a simple AF system that works without constantly fiddling with the settings to try to improve it.

Canon needs to incorporate autofocus into its Liveview - even the slow contrast-detect system in the Nikon is better than nothing.

Canon should forget plans for putting video on all their cameras and chasing silly levels of megapixels on small sensors as this usually just comes with the penalty of worse noise performance and reduced image quality (e.g. the 15Mp Canon 50D vs the 10 Mp 40D for example). These things are just marketing gimmicks. I feel that efforts should be concentrated on reducing high ISO noise to at least match the D3 and preferably raise the bar further by achieving this at around 16-21 MP on a full frame sensor shooting at 9-10fps. It should be technically possible as the 21Mp 5DmkII is already making big inroads to close the gap with the 12 Mp D3 in noise. Canon were high ISO kings for so long, that I am sure they have the potential to leapfrog Nikon once more.

A 16 Mp, low noise Canon 1Dmk IV with the AF fixed to match the D3 performance standards, would give Canon the best action camera for sports and wildlife in the world. As it stands, because of it's unique high ISO performance, very accurate matrix metering and excellent autofocus tracking, I think Nikon currently holds that title and it is very tempting to jump ship !

JD
18.6.09