|
Nikon D300: Impressions from a D200 user
By John Cazolis
Autofocusing with the D300
Coming from the D200, the autofocusing on the new camera certainly feels much better and a lot more accurate. And the more I use it, the more I realize that there really isn’t any comparison between the two. The old 11 point system of the D200 was quite OK for most work but the autofocusing supplied with the D300 is way ahead in every respect. Compared to the D2x the focusing under good lighting conditions feels about the same but as soon as the light level starts to drop, the D300 starts to shine.
I think most of the positive feedback from the AF system can be attributed to the less hunting, especially in low light, which would also help speed things up in the AF department. Also impressive are the new 3D-tracking technology and Scene Recognition system (more on those later).
Although I haven’t used any instruments to measure the differences, I get the feeling that the autofocusing motor for screwdriver lenses has more torque than the D200. I say this because my AF 80-400mm VR lens, which is one of the slower lenses in the Nikon line-up, seems to focus faster on the D300. Mind you the 80-400VR is still a slow focusing lens but compared to its previous performance on the D200, it does feel as if the new camera is driving the lens with more confidence, where before hunting was more frequent. As mentioned at the start of this paragraph, I didn’t measure any of this in a lab, but autofocusing definitely feels better with a lot less hunting and I’m sure all new users of the D300 will be left with the same impressions.
Custom settings menu AF options
In the Custom Settings menu, option a3 will help professional photographers or others who shoot a certain way, to setup the autofocusing in the best way suited for their style or subject. So a sports photographer who frequents a bike track might select the 9 point system or if shooting soccer the 21 point system or if shooting birds the 51 point system. The 3D system can be impressive but under true working conditions can get fooled a lot so is only recommended for easy scenes as described further down.
I recommend new users avoid choosing the 51 point 3D AF system until first having used the other options so as to get better acquainted with the new AF technology that the D300 provides. Starting off with the 9 point system for a few hours or days will give you the chance to understand its abilities and shortcomings. Then select the 21 point system, use it, progress to the 51 point system and then use the Auto system last.
Dynamic Area AF
On the back of the camera there is a selector with three options: Single point AF (great choice for 90% of situations – static subjects), Dynamic Area AF (great for action and most non-stationary subjects) and Auto (impressive but limited).
Lets forget about the Single point AF and Auto options so we can concentrate on Dynamic Area AF. With Dynamic Area AF the photographer selects the initial AF point and as long as the camera is in Continuous shooting mode, it will track the subject using 9, 21 or all 51 points. Any of those can be selected manually or automatically. Dynamic Area AF is intended for subjects that are non-stationary – let’s take a closer look:
9 point AF system
Usually the quickest and most accurate autofocusing is rewarded to single point AF systems. Amongst the options in the Dynamic AF Area modes, the 9 point seems to be the closest thing to spot focusing. The nine points are grouped tightly and due to the less coverage you may need to work harder at being more precise with lens placement as you follow your subject. This system is recommended for predictable action such as runners, bikes, cars, or anything else that is moving towards or away from the camera or is moving across the frame such as in panning.
There is a good reason for preferring fewer points. The camera needs to crunch the data coming from all points, so crunching data for 9 points will be done quicker compared to 51 points, so the less points, the faster the focusing and this is noticeable in the field.
I personally shoot a lot of wildlife and have found that the 9 point system is pretty good at tracking small birds in bushes amongst the many interfering branches and other obstacles (Custom setting a4 set to normal). That is of course with an AFS lens (eg AFS 70-200mm), since the effectiveness of the Dynamic AF modes are less capable with slower and less responsive lenses such as the AF 80-400mm. I find the 9 point system good for subjects that are small in the frame and that are somewhat predictable in their movements.
21 point AF system
After using the 9 point system for a while, the 21 point will seem similar but with a wider range of coverage. The 21 point system is good for when the subject is larger in the frame such as people close up or when the movement of the subject is non-predictable. The erratic movements of a soccer player or a wild animal might be tracked easier by the larger 21 point system.
51 point AF system
The 51 point system is a different ball game. The area of coverage is nearly 2/3 of the frame and acquisitions sometimes can seem to take some “noticeable” time. The AUTO version of this system does not indicate on what its focusing on while the 3D version does. The 3D version also seems more capable as well (depending on lens, subject and framing). I feel that both versions work better with normal to tele lenses than with wide angle lenses – especially the AUTO version. This may be because longer lenses throw the background out of focus easier. The intention of the 51-point Auto system is to let the subject wonder within the frame as you concentrate on the composition instead of trying to track the subject by keeping the lens centered on it
51point AF with 3D tracking
The 3D-tracking option uses scene recognition technology to follow the chosen subject within the frame. This new technology uses color, brightness and other information obtained from the 1005-pixel RGB sensor to analyze the subject prior to capture and applies the results to achieving greater accuracy in AF, AE and AWB control. The primary objectives of this technology are subject identification and subject tracking. Initially the camera tries to identify the background from the foreground and then the people (if any in the frame) by their color and other parameters. At this point the camera analyzes the position and size of the person. This information is used to focus on the person in the Auto mode. Even if the subject you are tracking moves out of the 51-point AF range, the subject might still be tracked based on the information coming from the larger 1005-pixel RGB sensor.
The 51-point 3D tracking technology is a bit of a hit and miss situation especially when shooting complex scenes that have busy backgrounds or that are made up of low contrast parts. On the other hand it is very impressive in well defined and easy compositions such as BIFs (birds in flight), portraits and other “stand out” subjects against neutral backgrounds (eg: tennis player, people in a studio setup, motor-racing etc). The key is, if your subject is well defined, the 3D tracking sensor can find it. Even eyes (human and animal) seem to be tracked by the sensor even though Nikon engineers claim that no such technology has been utilized so this success must be attributed to the contrast and color part of the scene recognition system.
Apart from action I also find the 3D-tracking technology useful for semi-stationary subjects as well. I can focus on my subject with whatever point is already selected and then as I recompose the sensor will track the initial focus point across the screen. If the subject happens to move the camera will detect it and follow it while focusing. Pretty cool stuff but still quite limited in its use. This technology is certainly a step in the right direction, but I feel it has to mature a bit more. Obviously the next generation of Nikon cameras will have a more mature version available so the future is looking bright.
My personal setup
The fact that you have a focus mode selector on the front of the camera, an AF area mode selector on the back and other options within the menu system can overwhelm most users. I like to simplify my life so I have eliminated the use of the front selector by leaving it on Continuous focus all the time. Thankfully most of my lenses have an easy to use manual focusing override switch (M/A) which is more convenient for those rare occasions that it’s needed. In the menus I have selected 9 points as my standard Dynamic range AF system of choice. So when I’m out shooting I can easily just select one of the AF area modes (single point, 9 point and 51 point Auto) in an instant with the thumb of my right hand. I don’t seem to ever use the first option of Auto and usually have the middle option of 9 points selected for most of my photography. When I have a static type of subject I might switch to single point if the other area modes don’t help.
There are a lot of options for customizing focusing to your personal liking but some experience with the system will ensure you choose what’s best for your style. As I gain more experience with the new AF technology I will be posting updates on my blog: (http://nikonglass.blogspot.com/)
IMAGES:
Not exactly award winning images, but they illustrate how good the new AF is. This active warbler was kept in focus by the 9-point Dynamic area AF even though it kept jumping from branch to branch. The AF was not fooled by the branches, pretty amazing!
Feedback on this review:
:
[ 1] 2
 Re: Review: Nikon D300 by John Cazolis
1: January 12, 2008, 02:24:50 PM by eschurri very much enjoyed this article. One thing i'd love some more info on is how you use auto focus. I've had a D200 for two years, and i love it, but i've never figured out the real differences between some of the group/dynamic autofocus modes.
i've settled on locking the focus in the center area and using the focus lock when composing a picture. that doesn't mean i put my subject in the center of the picture (egad!) -- rather, i aim the center of the camera at the target image, focus, hold the focus lock (usually with the shutter half-depressed), and then recompose the image. I did this for all the shots i took of wildlife in Africa and it worked great.
In my experience, using the autofocus usually causes the camera to focus on something other than what i want. i find my method above faster than using my thumb on the selector dial on the back of the camera.
Re: Review: Nikon D300 by John Cazolis
2: January 12, 2008, 09:21:56 PM by Nikkoreschurr, I think maybe you missed the second part ( http://nikongear.com/nef/reviews/D300_JC/D300_review_2_John_Cazolis.php) which is about AF. Your observation about single point AF being easier to use and being more accurate more often is correct. Dynamic Area AF is for action scenes where we as photographers cant keep up. We all want a good camera that can cope at following the subject, but this is quite hard. Each model of camera has improved on the previous model, but AF is still a hard thing to master. A good understanding of the camera and its AF mechanism, some prefocusing, a bit of experience with all the hardware and the subject and a bit of luck is usually the best prescription for success. Happy shooting 
D300: 14 versus 12 bit
3: January 13, 2008, 01:29:09 AM by NikkorI've received a lot of questions regarding whether to upgrade or not and whether 14 bit is worth shooting over 12bit. So here is my take on the 14 bit issue. Enjoy  I only shoot RAW and it’s not so much for the quality but for the convenience as well. This is why I prefer compressed RAW for most of my photography. But even so, it’s nice to know that I have the choice to shoot a landscape or some other critical image in uncompressed 14-bit for maximum capture of tones giving me more leeway in post processing thus maximizing the output quality. And I say this even though I know that i cant really see any “practical” difference between 12 and 14 bit, or between compressed or uncompressed in the final print, let alone in publications. Sure if you go pixel peeping, the extra 2 bits offer some extra info both in the highlights (and in the shadows) as shown in the example pics on my blog, but I honestly think that 12-bit compressed is the way to go. If you think the extra info as shown is of use to you, then go ahead and select 14 bits. The examples have had 2 stops of exposure added to magnify the advantage of 14 bits. You may want to download both pics or open them in new tabs to compare them. http://nikonglass.blogspot.com/2008/01/d300-14-versus-12-bit.html
Re: Review: Nikon D300 by John Cazolis
4: January 14, 2008, 01:41:38 PM by NikosRYou have chosen to provide a highlight example, while I would expect than any perceivable difference would be more easily be manifested in the shadows. Any highlight gain might be somewhat evident when using lossy compression rather than losless or no compression at all, but a suitable test must be performed before we can distinguish any difference between 12 and 14 bits in the highlights and most importantly, in highlight recovery.
Re: Review: Nikon D300 by John Cazolis
5: January 18, 2008, 08:26:33 PM by paulcAs an adjunct to the 12 bit vs. 14 bit discussion, I am curious if anyone has any evidence of quality differences between "lossless compress sed" and "compressed?"
And yes I am aware that in many things, stuff measured with instruments can show very different results than just observation by the human eye. Our eyes really are amazing instruments that can automatically compensate for a lot of stuff. Thus there MAY be a defined difference if measured via instruments that can NOT be seen by the naked eye.
Re: Review: Nikon D300 by John Cazolis
6: January 18, 2008, 08:40:57 PM by ajAs an adjunct to the 12 bit vs. 14 bit discussion, I am curious if anyone has any evidence of quality differences between "lossless compress sed" and "compressed?"
If this is true lossless compression, there should not be any difference at all. The other compression option might show some difference.
Re: Review: Nikon D300 by John Cazolis
7: January 18, 2008, 09:49:18 PM by NikkorYou have chosen to provide a highlight example, while I would expect than any perceivable difference would be more easily be manifested in the shadows. Any highlight gain might be somewhat evident when using lossy compression rather than losless or no compression at all, but a suitable test must be performed before we can distinguish any difference between 12 and 14 bits in the highlights and most importantly, in highlight recovery.
The way I see it, i can shoot with 12bit lossy compressed and have a small and "fast" file that still gives me all the benefits of RAW or 14bit uncompressed with maximum performance. I don't concern myself with the in-betweens because the two extremes are all that matter to me. Of course such a test of all combinations would have some interest, but I wont bother it since the differences between the two extremes is so small. Also, my two comparison images are targeted at the highlights since this is what concerns me the most and probably the only reason for ever using 14 bits in the first place. If 14bits can give me more detail in the shadows is of little concern to me (unless it helps reduce noise). What I want in the shadows is less noise, and the D300 has managed that to some point (its better than the D200). As for the testing methodology, it works for me since for I'm a photographer and not a scientist
Re: Review: Nikon D300 by John Cazolis
8: January 18, 2008, 09:55:17 PM by NikkorAs an adjunct to the 12 bit vs. 14 bit discussion, I am curious if anyone has any evidence of quality differences between "lossless compress sed" and "compressed?"
And yes I am aware that in many things, stuff measured with instruments can show very different results than just observation by the human eye. Our eyes really are amazing instruments that can automatically compensate for a lot of stuff. Thus there MAY be a defined difference if measured via instruments that can NOT be seen by the naked eye.
Technically and numerically there is a difference , but who cares? I think we over concern ourselves too much with such issues. Yes, there is a difference but is too small to concern ourselfs with it. You can shoot and print images (both 12 or 14 bits, compressed or uncompressed) but in the end, no one will see the difference because the viewer is looking at the image and not the pixels. So in practical terms there is no difference. Yes, you are correct, the human eye is indeed a fine instrument Happy shooting 
Re: Review: Nikon D300 by John Cazolis
9: January 19, 2008, 04:50:57 AM by NikosRThe way I see it, i can shoot with 12bit lossy compressed and have a small and "fast" file that still gives me all the benefits of RAW or 14bit uncompressed with maximum performance. I don't concern myself with the in-betweens because the two extremes are all that matter to me. Of course such a test of all combinations would have some interest, but I wont bother it since the differences between the two extremes is so small. Also, my two comparison images are targeted at the highlights since this is what concerns me the most and probably the only reason for ever using 14 bits in the first place. If 14bits can give me more detail in the shadows is of little concern to me (unless it helps reduce noise). What I want in the shadows is less noise, and the D300 has managed that to some point (its better than the D200). As for the testing methodology, it works for me since for I'm a photographer and not a scientist You twisted what I've said. The jist of what I said amounts to that if you don't care about shadow detail (and shadow noise BTW, since it seems from samples I've seen that read noise is much better on 14bits) don't bother with 14bits at all. Looking at highlights they way you have will not provide you with any appreciation of the differences. Highlight recovery should be influenced much more by lossy vs lossless than 12 vs 14 bit. Is this photographically clear?
Re: Review: Nikon D300 by John Cazolis
10: January 19, 2008, 08:43:01 AM by NikkorLooking at highlights they way you have will not provide you with any appreciation of the differences.
How can this be? I tested the two extremes (compressed 12bit & uncompressed 14bit) and showed a small but noticeable difference. To me its really not worth the trouble but i may use it as i said for a landscape shot or some other critical image, maybe even a wedding portrait. So looking deeper into the other two options seems like a mute point. Highlight recovery should be influenced much more by lossy vs lossless than 12 vs 14 bit. Is this photographically clear?
You are referring to Nikon's lossless compression that is not really lossless since it sacrifices some bits from the highlights. But testing has shown that the data removed cannot be noticed by the human eye, so in effect and in practical terms its seen as "lossless". On the other hand 12 vs 14 bits can show a "noticeable" difference, yet very small to consider worthwhile. So the bit difference compared to the compression difference seems to offer more noticeable results. So photographically, a landscaper has more to gain moving up to 14 bits than just going from compressed to uncompressed. I hope this clears matters
[ 1] 2 |
Review: Nikon D300 by John Cazolis