Nikon D300 Notes

(Digital SLR Still Camera)

 

Intro / Disclaimer

Me: Amateur Photographer. Once upon a time, my curiousity about DSLRs got the best of me, and I purchased a D300. I assumed the picture quality would be much better than with my compacts. I mean it costs 10 times as much, weighs 10 times as much, and is 10 times bigger, with big giant lenses instead of little tiny ones. Initially, I was very disappointed. I kept trying to figure out what I was doing wrong, and wondered if my camera was OK. Since then I've lowered my expectations and learned to appreciate the true advantages the D300 has over my compacts, and I've made peace with selective focus - often love it, sometimes still fight with it. But, my D300 taught me how amazing the little compacts are, and I'm glad I didn't throw them away when I bought the D300. But also, I've learned to deeply appreciate the advantages of the D300, and am now really enjoying shooting with it. This document covers some things about shooting with the D300, and includes some comparisons with compacts - pros and cons. What's here comes from a combination of fact/experience and logic/theory. Still with me? - Read on...

 

Definitions

Focus Distance	There is exactly one thin plane of focus at a certain distance from the camera. Everything at that distance will be in maximum focus, everything nearby distance-wise will be slightly out of focus, everything much nearer or much farther away will be way out of focus.
Depth-Of-Field	The rate at which things at growing distances from the focus distance become out of focus. The greater the depth-of-field the slower the rate, meaning more things are more in focus. Depth of field is defined by actual focal length (not equivalent focal length), aperture, and focus distance. Point-n-shoots have much shorter focal lengths, which is why they are biased toward greater depth of field.
Selective Focus	The opposite of great depth of field, meaning as things are further from the focus distance they become out of focus rapidly, so fewer things are in focus.
Compact/Point-N-Shoot	Compact all-electronic cameras with fixed lenses and great depth-of-field. Pictures are most often shot using the LCD, although some have electronic viewfinders. I include Bridge/Super-zooms in this category too.
DSLR	Digital versions of the old film cameras using Single-Lens-Reflex technology - they have physically bigger sensors (much fatter pixels - not much higher resolution) and more selective focus. Pictures are most often shot looking through an optical viewfinder, although some allow focusing/picture taking while looking at the LCD. Interchangeable lenses.
High ISO	Uping the gain on the sensors pixel amplifiers, resulting in faster shutter speeds, but more noise.
Low ISO	Reduced gain on the sensors pixel amplifiers resulting in cleaner pictures, but reduced shutter speeds.
Aperture	Size of opening through which picture light passes. Note: Lenses perform best in the middle. Wide-open causes chromatic aberration and softening and stuff. Stopping down causes diffraction resulting in softening, loss of light, and stuff. The optimal range depends on the camera and the lens.
Stop Down	Decrease aperture (larger f-stop numbers).
1 Stop	A term thrown around a lot - often incorrectly. It means increase or decrease of light by a factor of 2. For example, stopping down the aperture from 2.8 to 4 results in a 1-stop reduction in light (aperture math is non-linear). And, increasing the ISO from 200 to 400 results in a shutter speed increase equivalent to 1 stop (ISO math is linear).
Reproduction Ratio	1:1 means object the size of the sensor would occupy an entire uncropped photo - about the height/width of a dime/quarter. 1:5 means object 5 times bigger than the sensor would occupy an entire uncropped photo, and so forth. 5:1 means object 1/5 the size of the sensor would occupy an entire uncropped photo. Synonymous with Magnification Ratio.
Macro	For the purpose of this document, any close-up that would not be possible with a non-macro lens, or teleconverter / extension tubes. Approximately 5:1 - 1:5.
Extreme Macro	Sometimes just a synonym for macro, sometimes more specifically means any reproduction ratio greater than 1:1, for example 5:1.
Lens Speed	A misnomer. Refers to the maximum and/or sweetest aperture. "Faster" lenses can be shot more open resulting in faster shutter speeds - thus the term: "Fast Lens". Also, autofocus may be faster, since focusing is done at maximum aperture, and fast lenses often have fast autofocus motors.
Ambient	The contribution of light that is not from Flash.
Ghost	Secondary image component resulting from relative movement of subject/camera while shutter is open. Could be due to subject movement or camera movement, or both. Mostly refers to flash photography when primary image component is illuminated sharply by flash, but secondary ghosting is present due ambient illumination and slow shutter speed.
Hyperfocal Distance	Focus distance that results in near objects as well as farthest objects being maximally focused.
Circle Of Confusion	In general, its a measure of how out-of-focus a point can be before it matters. As it relates specifically to a DSLR camera body, it can be thought of as a measure of how sharp a lens needs to be for your sensor, or how far you can stop down before diffraction causes softening, or at what point focused-enough is indistinguishable from perfectly-focused. Coarse sensors need less-sharp lenses, can be stopped down further without diffraction penalty, and may be more forgiving focus-wise. The D300's circle-of-confusion is about .02mm. (D3 CoC is .03, and Canon S5 is .005).
Non-destructive Edit	Image data is not altered. Instead, a list of change instructions is maintained in a separate file, or a database, or as metadata in the same file as the unaltered image data.
Destructive Edit	Image data is altered, and edit history is discarded.
Color Profile	Maps image data to colors.
Neutral Color Profile	Renders colors in a manner closest to what the human eye would see, which has lower contrast and less saturation than any other profile. All raw converters should render neutral profile exactly the same, color-wise. This is sometimes the best profile to start post-processing with because it may have the fewest coloration anomalies.
Standard Color Profile	Not standard. More contrast and saturation than neutral - generally more pleasing than neutral and often a good starting point for post processing. Some raw converters have their own "standard", as well as trying to imitate the camera manufacturers version of "standard".
Vivid/Landscape Profile	More contrast and saturation than "standard". Again, Some raw converters have their own "vivid", as well as trying to imitate the camera manufacturers version of "vivid". Cooler white balance - tries to avoid oversaturating yellows.
Noise	Sensors are imperfect - pixel readings vary between "spot-on" and completely erroneous.
Color Noise	Imperfect pixel readings that results in color being affected more than luminance. For reasons I don't understand, the predominant erroneous colors are often cyan or magenta.
Luminance Noise	Imperfect pixel readings that results in luminance being affected more than color.
RAW Conversion	Conversion of sensor data to RGB. Includes assignment of a color profile and estimate of white balance. In its purest form, excludes noise reduction and sharpening, although different algorithms have differing amounts and qualities of minimal noise reduction and sharpening built in.
RAW Converter	Computer Software (or firmware in the camera) that converts raw sensor data to RGB. Raw conversion software always include the capability to do noise reduction and sharpening as well - in some you can turn most of it off, in some you cant - be-aware.
Capture Sharpening	True capture sharpening is (hopefully) done during raw conversion, or shortly thereafter, and uses a very different algorithm than creative and output sharpeners. Purpose is to make up for lack of sharpness due to lens/sensor/focus/camera-shake.
Creative Sharpening	Performed after capture sharpening, typically using an edge transition exageration technique, e.g. unsharp mask. Generally provides masking/threshold & detail controls which should be adjusted in a fashion unique to each photo and desired effect.
Output Sharpening	Additional sharpening generally global - governed by target output medium and expected viewing distance range.
Flash Sync Speed	Fastest shutter speed that won't produce a dark band when flash fires only once.
High Speed Flash Sync	Firing the flash multiple times as slit between front and rear curtain races in front of the sensor.

 

In These Ways, a Point-n-shoot is better than a D300

Note: FOR THE PURPOSE OF THIS DOCUMENT, I'M TALKING ABOUT COMPACTS IN THE $200-$500 RANGE THAT HAVE PRO QUALITY LENSES, TOP QUALITY SENSORS, & SOPHISTICATED FIRMWARE. Not talking about off-brand $50-$100 cheapies where quality has been compromised.

• The obvious:
  • Smaller/lighter/cheaper.
• Not so obvious:
  • They have more depth-of-field. This sometimes yields an apparently sharper picture, albeit sometimes a bit flat, and is the reason I now call these "2D cameras" instead of "point-n-shoot". When maximum depth of field is paramount, a point-n-shoot may take a sharper picture than a D300 (reason: D300 needs to be stopped down to match depth of field of 2D camera, which results in softening due to diffraction and slows down the shutter if light is primarily ambient). Possible Examples: Macros, group photos, documentation/information photos, some landscape shots.
  • You never get dirt/dust on your sensor.
  • They can take pictures silently (DSLRs unavoidable shutter sound can annoy people and scare animals).
  • They have limited video capabilities.
  • Have high-level convenient automatic modes, like scene modes.
  • They draw less attention in use.
  • Can be used for taking pictures of the underside of mushrooms without pulling them from the ground (get into tight places better).
  • LCD monitor shifts brightness in response to ambient light.
  • Histogram is displayed before you shoot, so you can adjust exposure before you shoot. On DSLRs histograms are only after-the-fact, so you have to shoot, check histogram, then shoot again.
  • When shooting using LCD instead of viewfinder (called Live-View on DSLRs), the LCD stays on until the shot is taken. DSLRs either dont have a live view mode, or the LCD shuts off before focusing & subsequent shot. D300 has a tripod mode where the LCD stays on, but its not suitable for hand-held operation.
  • Sweet spot of the lens at full zoom is f/5 - ish. Sweet spot of telephoto lenses for the D300 (that cost less than $1000) is f/10 - ish - a good stop down, or two.
  • Flash sync speed is faster for popup flash than the D300.

 

In These Ways, a Point-n-shoot is the same as a D300

• Provided the shooting conditions are within limits, especially plenty of light, and when selective focus is not desired, a point-n-shoot will take as good a picture as a D300, for the most part.

 

In These Ways, a D300 is better than a point-n-shoot

• The obvious:
  • optical viewfinder
  • interchangeable lenses.
• The biggest differences:
  • They have more selective focus (the opposite of more depth-of-field). This is sometimes good, sometimes a pain in the arse, and is the reason I call this my 3D camera.
  • The sensor has much better signal to noise ratio. At low ISOs when there is plenty of light (plenty of signal) this doesn't matter much, but when the light is low or maximum shutter speeds are desired, it matters a lot.
  • More sophisticated auto-focus system.
  • More precise manual focus.
• Other major differences, although in my opinion less important than the above:
• Better, more accurate, more sensitive & smooth color recording.
• Greater dynamic range, meaning more detail is possible in the shadows before clipping to black, and more details are possible in the highlights before clipping to white.
• Note: In practice, these benefits are often lessened by the fact that higher ISOs are more often used which results in decreased color performance and reduced dynamic range.
• A whole host of more subtle things:
  • More controls can be twiddled without having to remove your eye from your target.
  • Viewfinder allows much more precise focusing, especially manually.
  • Better built-in flash - way less red-eye.
  • Better support in 3rd party post-processing software, e.g. lens correction.
  • Supports multiple remote speedlights.
  • Supports higher shutter speeds, with and without flash.
  • Supports higher burst rates.
  • More sophisticated auto-focus system, which results in better auto-focus when you have it set right and use it right, but may also screw up a shot when its set wrong or used wrong - this happens to me all the time ;-)
  • Allows trap focusing.
  • Better low-light focusing.
  • Better support for accessories, like lighting and radio extenders.
  • Better support on the web and in print by other photo enthusiasts.
  • More things can be twiddled using the menus.
  • Instant power-up.
  • Short shutter lag.
  • RAW support
  • Everything faster.
  • Manual modes. Although some compact/p&s/2D cameras have full manual support.
  • Big male appendage syndrome. I ain't kidding - you feel very powerful with one of these things in your hands.
  • So I hear, you can attach a $10,000 super-telephoto lens on it and get better performance for shooting sports and wildlife - I wouldn't know.
  • I also hear you can put $2000 shorter lenses on it for better quality / performance - I believe that's true, albeit often overstated.
  • High-speed flash sync supported via external flash.

 

What you can do to make a point-n-shoot more like a DSLR

• If the light is low and the subject ain't moving, use a tripod, since upping the ISO destroys quality.
• Open up the aperture to minimize depth-of-field (maximize selective focus).
  • Note: Pictures will not be as sharp, and exhibit more chromatic aberration when shooting wide open - do chromatic aberration correction in software afterward.
• Get closer to your subject and/or zoom in, and dont crop much afterward.
• Use external lighting, if your compact supports it.
• Use adapters / add-on lenses for wide-angle, increased telephoto reach, macro magnification, and fish-eye.

 

What you can do to make a DSLR more like a Point-n-shoot

• Stop down to increase depth of field. Note:
  • Pictures may not be as sharp due to diffraction.
  • A tripod may be required since shutter speed will go down, or:
  • Up the ISO, and/or use your flash or something if possible to make up for lost light.
• Get further from the subject and/or zoom out, and crop afterward.
• Only take your kit lens with you.

 

My Compacts

1. Canon S5: 38-450mm (equivalent) - A DSLR-like camera, but using compact-digicam technology, includes: super-sharp Carl Zeis Lens, with super-zoom, super-macro, great video/audio, nice pop-up flash, external lighting support, electronic viewfinder, articulating lcd, adapter / add-on lens detection.
2. Panasonic Lumix: Featuring super-sharp Leica 24mm (equivalent) wide angle lens.

Before buying a DSLR, I recognized a primary weakness in the Canon S5-IS being no wide-angle, and purchased a Panasonic Lumix to fill the void. I decided not to buy any of the accessories for the Canon, since they can easily end up costing 5 times as much as the camera itself did - I was holding out for the DSLR decision.

 

Which is sharper - A High Quality Compact, or a D300?

When I first got my D300 and took some test shots to compare image quality to my compacts, the compacts looked sharper. Since then, I've come to realize that both cameras are about the same sharpness at the focal plane. The compacts look sharper because of greater depth-of-field (and aggressive in-camera sharpening). But, if you stop the D300 down to have the same depth of field as a compact, it wont be as sharp as the compact. It is because of this that I say the biggest difference between the two is in selective focus. If you dont want selective focus, but you do want sharpness, consider shooting it with your compact instead. Note: The best quality lenses are not any sharper than mid-priced lenses once stopped way down.

 

Sharpness / Resolution - Bounded by Sensor or Lens?

First picture I saw from a Digital Camera was 640 by 480 - full resolution - it looked terrible. That was like 15 years ago, when lenses were about as sharp as they are today. Sharpness / Resolution was clearly bounded by the sensor, not the lens. Since then, digital sensors have increased in resolution (and quality) at a rate of about 1MP per year. Sensor resolution has now evolved to the point where many sensors can resolve more detail than many lenses, and proper camera design should balance the two. For example, the D300 sensor at 12MP can resolve as much detail as most if not all Nikon Lenses can throw at it. There are lenses however that have headroom when used with the D3, for example the 14-24. Full advantage of that lens can only be had with a D3x. One way to tell how well your sensor can resolve lens detail is to compare (on a computer monitor at 100%) a closeup photograph (of something flat, taken straight on, like a test chart) taken under ideal light at one of the lenses sharpest apertures, say f/8, to one taken at f/22. The lens image should be about the same, except for some diffraction. If the picture taken at f/22 looks noticeably softer, then your sensor, and your eyes, are able to resolve the difference between the lenses sharpest and least sharp images. Try the test again only using f/16 instead of f/22 - can you still see a difference? - the point at which you can no longer see a difference is a measure of sensor versus optical resolution, and your eyes ;-} Another way to prove to yourself that your sensor is capable of resolving sharpness differences between lenses is to compare different lenses at the same focal length and aperture and compare the difference (at 100%). Ken Rockwell compares the Tokina 11-16 with the Nikon 12-24 this way, and the results you can see here. Also, check out the references below for more information on digital camera sharpness / resolution.

 

Pop Quiz - Which Camera For Which Purpose Is Best? Canon S5 or equivalent, or Nikon D300 or equivalent?

Not always stated, but there is a pervasive trend that high-ISO yields much poorer results on Canon S5 than D300. And on the D300: greater depth-of-field is accompanied by a decrease in sharpness, and requires more light. Depth-of-field issues wane when shooting things far away, or using wide angle lens. Likewise, high-ISO issues are moot when light is abundent.

Warning: These scenarios are grossly oversimplified. The intent is not to provide a set of rules to follow, but to help illuminate by examples a set of principles that will allow intelligent decisions to be made.

Macro notes: Canon-S5 handles macros by switching to widest angle and doing some internal magic that allows close focusing (like an extension tube, maybe?). Because of the wide angle, the camera has to be very close to the subject. On the D300, macros are shot just like any other picture, except if you have a macro lens you can focus closer; if you dont have a macro lens you can mount an ordinary lens on an extension tube and get similar performance.

Scenario	Camera	Comments
1:1 Macros Of Small Animals & Insects	D300	Canon S5 1:1 macro requires subject nearly touching lens - Insects & Small Animals wont tolerate this.
1:1 Macros Of Inanimate Objects	D300	Canon S5 1:1 macro requires subject nearly touching lens - lens blocks all the light. Exception: If subject is translucent, and can be backlit, S5 may take a better picture.
1:2 Macros Of Inanimate Objects	Depends	If extreme selective focus desired: D300. If extreme depth-of-field desired: Canon S5. If somewhere in between, either can be used, however D300 will require macro flash supplement, since it will have to be stopped way down. Likewise, if ambient low, Canon S5 will require a flash-ring or other creative light injection to avoid high-ISO - on-camera flash won't reach around the lens. Its may also be possible to use tripod and skip the lights, if subject is really still.
1:5 Macro Of Whatever	Depends	If extreme selective focus desired: D300. If extreme depth-of-field desired: Canon S5. If somewhere in between, either can be used, however D300 may require macro flash supplement, since it will have to be stopped down significantly. If ambient low, Canon S5 can use on-camera flash to avoid high-ISO. Note: Canon -S5 still needs to be closer to subject for 1:5 macro than D300. If this would be prohibitive to subject, D300 required. If subject is still, its possible a tripod may be used instead of or in conjunction with flash.
5:1 Macro	D300	Requires reverse mounting a relatively wide-angle lens over a macro or telephoto lens. Use teleconverter for more magnification, extension tubes for closer focus - these things may be joined together without hacking, although I did have to hack a mount for my macro lights for this case. Note: focus is extremely selective at all apertures. It may also be possible to hack things together for super-extreme macros on S5, but not as supported via normal mechanisms. Again, depth-of-field will probably be greater with the S5 than the D300, depending on what you've put together.
Indoor Sports	D300	Need fast shutter in low-light - High ISO required. Selective focus generally good as well. If space is extremely well lit, the S5 can be used, but photos will be much more 2D. Still, D300 has other advantages as well (see Outdoor Sports).
Outdoor Sports	D300	I give the edge to the D300, due to auto-focus system, selective focus is generally desirable and is not extreme anyway at theses distances like it is for macros, and may still want to up the ISO a bit to reduce blur.
Quiet Indoor Events	Canon S5	You wont be able to take non-blurry pictures when people are moving, but at least you can get some shots. D300 shutter makes too much noise for some events. I'm assuming flash also disallowed, or you're too far from subjects for it to help.
Loud Indoor Events	D300	Photographing a Rock Concert? - Use the D300. Selective focus is good, and shutter sound is not a problem. Jack up the ISO if ya have to, and maybe perform noise reduction in post. Warning: Avoid mosh pits and vigorous dancing.
Landscapes In Abundent Light	Depends	If you want top quality prints, or will be cropping a lot: D300. If just viewing on computer monitor, either one about as good.
Landscapes In Low Light	Depends	If you have a tripod, you can get good results with a Canon S5. If handheld: D300, since you can up the ISO with less penalty. Of course if you need to stop down because you want more depth of field, you're now in the same boat with the Canon S5 - needing a tripod.
Close-up of a person, with a portion of the Golden Gate Bridge or Eifel Tower in focus in the background.	Depends	If ambient only: Canon S5, since great depth-of-field required. If flash can adequately illuminate subject, and tripod available to eliminate backgound blur / foreground ghosting: D300 - stopped down and using slow-sync. If light is low, or subject(s) moving, it could go either way. Canon-S5 can be shot more open, but ISO can be up'd on the D300.
Not so close up of person with the entire golden gate bridge in partial focus in the background.	D300	This is a wider angle shot, which have inherently less selective focus. Exception: You forgot your wide angle lens - have to shoot with the S5 and stitch a panorama together.
Group Photo in Small Room	Canon S5	Assumes people are both near the camera and far from the camera and all must be in focus. To get the same shot with a D300 you would have to stop it way down and add lots of light, and the results will be softer, which may or may not be OK.
Group Dining At Long Table	Canon S5	See 'Group Photo in Small Room'.
Group Dining At Round Table	D300	I'm assuming you have an ultra-wide angle lens for it. If not opt for the S5.
Group Photo in Large Room	D300	Assumes people are lined up wider, instead of deeper. Use wide-angle lens and get further away.
Closeup of Cobra Snake	D300	Generally want to isolate subject from background, whilst not getting too close. Same holds true for alligators, and poisonous frogs... Canon S5 would also do a good job of capturing the snake, since it has super-zoom, and this would not be a true "macro" (requiring macro mode) - assuming enough light, of course. The biggest reason to choose D300 is selective focus. If snake background focus desired, the Canon-S5 may be better.
Birds In Trees	D300	For Auto-focus & Selective Focus.
Birds In Air	D300	Although I would favor the D300, if you could get a focus lock on it and/or track it adequately with an S5, it would also do a fine job. Selective focus or not - doesn't matter. In either case you want to shoot as open as necessary given the light, but without being so open as to exhibit chromatic aberration and other too-open performance phenomena.
Portraits	D300	D300 produces truer skin tones. You can take great portraits with an S5 if there's enough light - and skin tones may be OK, but they may be hard to correct. People are very sensitive to and intolerant of their skin tone being off. Also, you probably want more selective focus, although not necessarily.
Weddings	D300	See "Portraits" above, but also because you need to appear professional. That, and you may want / need selective focus. Also, the clients may want top-quality prints and D300 pics may look better printed, especially large. But bring your compact too ;-}
Vey Low Light / Very Long Shutter Exposures	D300	Usually we talk of noise due to the higher-ISO needed for faster shutters, but there is also noise due to reduced signal during extended shutters (low-ISO). The D300s bigger pixel sensors make for a much cleaner picture in this case as well.
Surfing	D300	Better dynamic focus tracking.
Racing	D300	More accurate predictive focus.
Moving Subject(s), Low Light	D300	But a D3 would be even better. (Assuming you want sharpness not blur). This is a case where you need fast shutter, open aperture, and high-ISO.
When too lazy to deal with D300 or its not with you.	Canon-S5	Or even smaller pocket camera.
For extreme depth of field and maximum sharpness.	Canon-S5	After stopping my D300 (usually macro) lens down far enough to match the Canon-S5's depth of field, the D300 image will be less sharp and detailed than the Canon-S5. Exception: if everything in picture is way out in the distance, especially if shot with wide angle, depth of field not really an issue.

As the above scenarios suggest, often the D300 is a better choice for a particular situation, and sometimes either one will do about the same job, although you might have to shoot differently depending on camera used. Occasionally, the Compact P&S is actually a better tool for the job. Note: Neither camera can take a sharp picture of moving subjects in really low light (theoretically you could if subject was close enough for flash, but you still have the problem of focusing).

Tripods

Definition of a tripod - anything that keeps the camera still (includes the ground, trees, rocks, furniture, ...). Compacts require tripods to avoid high-ISO; D300 requires a tripod to avoid overly selective focus. Of course they both require tripods if light is really low (unless you want or dont mind a blurry picture), but tripods only helps with camera movement, subject movement is only affected by shutter speed. You can sometimes circumvent the need for a tripod by using more light, e.g. Flash. One thing I like about the D300 is I rarely need a tripod. If stopping way down for a macro I use Flash. If shooting indoors - flash. Outside in low light - crank up the ISO. Truth is, I love the freedom to move the camera around until I've found just the right angle before shooting - tripod kinks my style. On the other hand, for sharp, high-quality, low-light photos, like the later portion of a sunset - gotta have a tripod. All my lenses except the wide-angle ones have stabilization, and I almost always use it when shooting hand-held.

Full Frame (FX) versus cropped frame (DX) Sharpness / Resolution

Disclaimer: This analysis assumes same resolution sensor in both cases, e.g. D3/D700 versus D300, and ignores image quality issues due to larger / smaller pixel sensor photo-sites, which is fair to do in the case of abundent light and low ISO.

Full frame is better if you're taking pictures of something that does not require full-zoom and a post crop. Why? because you can zoom in further preserving more detail for the same shot. Lets take, for example, a certain angle of view of a sunset, shot with the same lens on an FX versus DX, final picture to be uncropped in software afterwards. Both will be 12MP resolution. You would have to zoom out the lens on the DX camera to get the same shot as the FX, thereby losing optical detail. Or conversely, you would need to zoom in the lens on the FX camera to get the same shot as the DX, thereby increasing optical detail. Another thing in this case: D3 would have shallower depth of field (more selective focus, assuming same aperture), since it depends on actual focal length, not equivalent focal length, and the D3 would have the larger focal length.

However, if you're trying to get as close a shot as you can of something and thus would be zoomed in all the way on either camera, like say a bird, the theoretical and sometimes practical advantage shifts to the DX, and here's why: you are projecting the same image from the lens onto more pixels. Optical detail is the same in either case, but the DX camera would record it at higher resolution. Whether or not the D300 actually makes a better picture, depends on whether at that zoom range, and distance to subject, and lighting..., the optical detail meets or exceeds the sensor resolution (subject to the constraints outlayed in the disclaimer above). And therefore, in this case, the D300 may be more sensitive to the quality of the lens than the D3 - would be if the lens was the limiting factor in optical resolution. So, if you've been following along, you'd realize that you need better lenses for the D300 than the D3 - the D300's finely packed sensor wont cover for less sharp lenses like the D3's fatter more loosely packed pixels. This is incidentally, contrary to what most people believe.

In the case of a D3X which has a 24MP sensor, none of this applies - it will take a better picture than a D300 in all circumstances. Note: The D300 pixel sensor sites are about the same size as the D3X's, yet the D3X uses a newer, better, sensor technology which can record cleaner pixels with the same size pixel sensors. Lastly, the best lenses do have more optical resolution than a D3's sensor (e.g. Nikon 14-24). If they didn't, there would be little point in the D3X's 24MP sensor, except for the advantage in printing very large. So, in order for D3X pictures to be sharper than D3, the best lenses must be used.

 

 

Full Frame (FX) versus cropped frame (DX) Focal Length Equivalence

Ever heard people say: "My 300mm FX lens is equivalent to a 450mm lens on my DX camera."? Well, I find it misleading. The focal length is not equivalent to 450mm - its 300mm on both, period. What is equivalent is the field of view - the 300mm lens on an DX camera has the same angle of view as a 450mm lens has on a FX camera, and thats not an advantage for the DX camera. The point is: you don't get any more magnification, no greater zoom, no clearer or more detailed image is presented to the sensor... There may still be an advantage on the DX camera at maximum focal length, but only if the DX camera has greater pixel density - it has nothing to do with the crop factor. For example, when shooting a bird and cropping to make the bird bigger, there is an advantage for the D300 over the D3 because the D300 pixel density is greater, however there is no advantage on a D40 versus a D3 since the pixel density is lower.

 

My Gripes About the D300

• Some functions can not be accessed without re-assigning buttons. e.g. Bracketing is only available via a button, as is depth-of-field preview, as is Flash Exposure Lock, but all of these things can not be assigned to a button at the same time. Thus I have to keep switching the button assignments around in a musical chairs fashion - very lame.
• ISO is not included on the LCD display along with shutter speed and aperture. This is a ridiculous hold-over from film days which is absurd for a digital camera, and represents a lack of innovation in modern camera user interface. Nikon should contract with Apple to re-think the user interface.
• Auto-ISO should be WAY more sophisticated - presently you select min-ISO (200+), max-ISO, and minimum shutter speed, and instead of dipping the shutter speed too low it bumps the ISO. I never use it. You should be able to just specify a list of priorities and let 'er rip. It should be capable of going lower as well as higher too. I mean the camera can detect camera and subject movement both. If my subject or camera is moving I need to maximize shutter speed, but if neither is moving I'd prefer to keep ISO low. Likewise, if I'm shooting at 500mm, then shutter speed is more important than if I'm shooting at 35mm. None of these things is taken into consideration, even though the information is available to the firmware, and so obviously affects what the chosen ISO should be. Auto-ISO operates in manual exposure mode too - this could be considered a feature, but I almost never want it, and need to remember to go to a menu to turn it off. All ISO related stuff should be given a place right next to shutter speed & aperture - repeat after me: digital sensors are not film.
• Like Auto-ISO, auto exposure should be more customizable - defining sharpest apertures, and weighting aperture against shutter speed - should detect focal length and focus distance and camera movement and act accordingly.
• Let me say again, the camera should detect when the camera is very still, like when its on a tripod, but also when its sitting on the ground, and factor that in to metering & ISO decisions, as well as live-view mode and lens stabilization. Likewise, the camera should detect subject movement (it already does) and factor that into decisions about metering and ISO. Does anyone really disagree with this? - Come On Nikon! (and Canon, and Sony, and Olympus, and Panasonic, ...) Likewise, the camera should detect focal length and focus distance, and factor those into its metering and ISO decisions. I should be able to set the camera down and take a picture, or pick it up handheld and take a picture, zoom-in or zoom-out and take a picture, and track a moving subject and take a picture, without having to go into camera menus and adjust settings for optimal results. Of course manual override should be manual override, but auto modes need improvement.
• Customizable settings should be allowed on a partial basis, instead of having banks of complete settings, and changes to camera settings should not affect the presets unless saved. That way, you can change the settings without worrying about them being permanent, and you can also make them permanent, plus you could mix and match saved settings at will, for different conditions.
• There really should be a way to return the camera to user default, instead of just factory default.
• LCD Monitor should brighten in the sun automatically. Does anyone not like this? - If so, make an option for manual override. Drives me crazy having to go into a menu item everytime I walk into the sun for a picture, then go back into the menu when I step into the shade, then back to the menu when I go inside... - downright embarassing!
• There's no wireless support, and flash CLS uses infra-red instead of radio waves. Yes, if you're willing to hang a $700 wireless transmitter and a $200 PocketWizard/RadioPopper off your camera, and another $200 PocketWizard/RadioPopper off each flash, you can accomplish the same thing, but medium-range wireless should be built in IMHO, with option for an external antenna and maybe a power booster for long-range.

 

So, Which Lenses?

The three most defining things are:

• Focal Length(s)
• Max (and in the case of macros: minimum) Aperture(s)
• Closest Focus Distance

it is those three things that define what you may be able to do with the lens, other important things are:

• Image Stabilization
• AutoFocus Speed.

Those two added to the above three define what you will be able to do with the lens. Next question is: how well? - which is defined by a whole host of lens quality related things, like:

• Sharpness
• Colors
• Distortion
• Aberrations

And, things that matter to photographer, but not photograph:

• Build Quality
• Handling

 

Zooms versus Primes

Once upon a time, all lenses were primes, then came zooms whose image quality was inferior but convenience was compelling. Now, image quality of zooms has mostly caught up with primes (top-quality, shorter zoom range zooms, are about as sharp as primes; mid-priced & long zoom range lenses are still lagging primes), therefore image quality is not the main advantage of a prime. Maximum / sweet-spot aperture is the main advantage of a prime. Primes generally have greater maximum aperture. A potential benefit of a prime is lower distortion (e.g. barrel / pincushion), and whats there is more easily correctable since it tends to have fixed characterestics (I say potential because PTLens & DxO do great jobs of removing even complex distortions for supported lenses). Distortion doesn't much matter for landscapes and even people, but matters a lot for architecture or other things with straight lines that should be straight. Also, standard & wide-angle primes are smaller & lighter than their zooming brethren (although once you have a multitude of primes to cover the zoom's focal range the advantage may have dissipated). Prime super-telephotos are big, heavy, expensive; likewise: top-quality super-telephoto zooms are also big, heavy, and expensive.

 

Sharpness Matters, But Not As Much As You Might Think

This discussion is confined to good through top quality lenses, i.e. real glass, reputable manufacturer (e.g. Nikon, Canon, Sigma, Tokina, Tamron, Leica, Zeiss), costing at least a hundred bucks. Sharpness differences between good and top quality lenses matter when:

• Everything else is near perfect: close to subject, good light, good atmospheric conditions, camera steady, focus spot-on, aperture at or near lens's sharpest.
• You have a fine grained sensor, e.g. 6-8MP compact, or 10-24MP DSLR.
• And, then only right at the focus plane. Since 99% of the content of real pictures is not at the focus plane, this is the primary reason absolute lens sharpness matters less than you might think. OK, if you are doing photo reproduction, or shooting test charts, it matters a lot, but not so much for landscapes or people or even macros.

Teleconverters

Disclaimer: Based on theory and experience, not physics; not confirmed:

Teleconverters are magnifying glasses. Thus a perfect teleconverter will still reduce sharpness, because it works by fanning out the image presented to it by the lens - it necessarily reduces information density, i.e. resolution/sharpness. Its like a hardware/optical version of enlargemnt via software/image interpolation. A sort-of result of this is that if someone asked me:

"Should I get a 300/2.8 and a 1.4 teleconverter or a 400/4?"

I would answer:

"Evaluate the lenses indendently of the teleconverter, then get a teleconverter to go with whichever lens you got."

It makes more sense to pick a teleconverter based on which lens you'll be using it with, rather than picking a lens based on its performance with a teleconverter versus another lens without a teleconverter.

My teleconverter motto: shoot it without a teleconverter if you can, use a teleconverter if you must - perhaps biggest reason to use a teleconverter is by blowing up the image presented to the autofocus subsystem and the viewfinder you can focus more accurately on the object of interest. i.e. without a teleconverter you may only be able to focus on a birds head, whereas with a teleconverter you may be able to focus on a birds eyes. If you got lucky enough to get focus on the birds eyes without the teleconverter the difference in image quality using software crop enlargement versus optical magnification ala teleconverter may be small, but you probably wouldn't be that lucky.

My Lenses & Recommendations

If you only are gonna have one lens, it should be an 18-200, period. It has large range of focal lengths, focuses fairly closely, has image stabilization, fast auto-focus, and good image quality (especially the wider half of the range). It was my intention when I first bought the D300 to only have one lens and I so I bought the 18-200. Then I bumped up against its limitations and started buying more lenses, the chronology went something like this:

• In order to match my Canon S5's ability to take macros (1:1), I purchased a Nikon 105mm/2.8 macro lens. I love macros and this lens gives them to me, and also serves as a 105mm low-light prime for non-macro work.
• In order to match my Canon S5's telephoto reach, I purchased a Nikon 70-300. I would have preferred an 80-400, but I was stuck on Nikon lenses and their 80-400 had ultra-slow autofocus, which takes a lot of the fun out of a telephoto lens. All other Nikon lenses that reached to 300mm, 400mm, or beyond were too expensive. It has good image quality up to 200mm, but 300mm performance is mediocre.
• Nikon 10.5mm fisheye - I had run up against the 18mm (27mm equivalent) wide-angle wall, and decided to go for the fisheye which can undergo rectilinear correction to satisfy ultra-wide need as well. It was either this or the Nikon 12-24 which cost about twice as much. Note: The specs for the 10.5mm fisheye lead one to believe the angle of view is 180 degrees, but it is not. That may be what it is on an FX camera, but after the DX sensor crops it, it is less.
• I think it was about at this point I bought the 50mm/1.8 prime, since the lens door had now been opened and everyone else agreed this was a must have - cheap, fast, small/light, sharp. Allows photos of performers in small night clubs without using flash, or actors in plays, or the late segment of sunsets/sunrises - handheld, ... (I also use it reversed over my telephoto lens(es) for extreme macros, or with extension tubes for normal macros)
• Sigma 150-500. I realized that Nikon was not the only option for lenses and really wanted more than 300mm. This is a big heavy lens. If I'd known I was going to get it, I would have foregone the Nikon 70-300, although I can fit the 70-300 in my camera bag with the rest of my lenses; not so for the 150-500, so I guess the 70-300 still has its place. PS - @2009/01 DxO Optics does not have support for it. Not sure about PTLens - I've never needed distortion correction with this lens, just Chromatic Aberration correction which can be done nearly as well, or better, manually.
• Tokina 11-16. Much cheaper than Nikon 12-24 and improves convenience, quality, and ability to compose rectilinear ultra-wide angle photos (over the 10.5 fisheye crop & correction). Note: the 10.5mm fisheye focuses way closer, and because of this, if I had to give up one of these two lenses, it would be this one (and keep the 10.5 fish). I'm not a big fan of the fisheye distortion, but it can be somewhat corrected. Still, if the 11-16 allowed me to focus as close as the 10.5mm fisheye, I'd give up the 10.5mm and keep the 11-16, if forced. PS - PTLens supports Tokina 11-16, but @2009/01 DxO Optics does not.

Another option, if you're willing to spend more for better quality / faster lenses, instead of the Tokina 11-16, Nikon 18-200, 50/1.8, 70-300, 7 Sigma 150-500:

• Nikon 10-24 - And maybe skip the 10.5 fisheye.
  • Option: Nikon 14-24 and maybe the 16mm fisheye. - With this option you'd be ready to migrate from DX to FX without skipping a (lens) beat.
• Nikon 50mm/1.4D - Dont get the 'G' or you have to do a hardware mod to be able to reverse mount it for extreme macros.
• Nikon 24-70/2.8 - performance exceeding middle third of an 18-200.
• Nikon 70-200/2.8 - performance exceeding long half of an 18-200.
• Nikon 1.7 Teleconverter - if you're only going to get one TC, this is the one to get. Performance at 300mm exceeds the 70-300, I'm "sure".
  • Option - get two teleconverters:
    • Nikon 1.4 Teleconverter (small range boost with better low-light performance - performance with 70-200 exceeding the Nikon 70-300 lens)
    • Nikon 2.0 Teleconverter (big range boost for using in broad daylight - performance with 70-200 rivalling Sigma 150-500)
  • Option2 - get all three.
• Toss in a super-telephoto prime or 200-400 zoom and you'd really be having things.
• I suppose you could get the 18-200 too for walk-around and have the option for maximum quality versus maximum convenience and odds of getting the shot.

This would be a higher quality but more expensive way to accomplish about the same thing. This combo would require more lens changes, and reduce the odds of having an appropriate lens already mounted when a fleeting shot came along (unless you have a second body - pros and cons of this approach left as an exercise for the reader). Also, some people supplement their DSLR/Lens choice with a compact :-}

 

Lens Notes

Sigma 150-500

One of my most frequently used lenses, especially for shooting near home.

Pros

• Only way to reach 500mm, with lens stabilization, without spending multiple thousands.
• Small & Light compared to the multi-thousand dollar options - works fine on ordinary tripod with ordinary head.
• Way sharper / more detail than by shooting with a 70-300 and blowing it up, provided you can keep camera shake down - of course.
• Lens stabilization works well, although not as well as Nikon's (see cons below).

Cons

• Maximum aperture exceeds Nikon spec (5.6) at 250-500mm (6-6.3). Even so, autofocus works well in medium to bright light, although it back-focuses at minimum focus distance at 500mm - thus manual focusing required at closest focus distances.
• Big & Heavy compared to 70-300 (or compact megazoom camera+lens).
• Lens stabilizer makes a sound that is annoying to photographer and skiddish subjects. Photographer may get used to it, but skiddish subjects don't.
• Lens attachment has play, like kenko extension tubes and teleconverter. Doesn't affect performance (I hope) but its annoying, although I've sort-of gotten used to it by now.
• I prefer Nikon's Normal (auto pan detect) & Active (great for shooting while driving...) lens stabilization modes to Sigma's Pan/No-pan modes (no active mode equivalent).
• Perhaps my biggest gripe: zoom and focus rings work in the opposite direction as all my other lenses - ggrrrrhhh - the chosen direction is consistent with Canon lenses and I suppose by not having to manufacture their lenses to work in both directions Sigma keeps cost down, - still annoying...
• Like 70-300 its performance with teleconverter is not very good (manual focus, image quality sufferage needs down stoppage for partial mitigation).

Conclusion:

Better than a Nikon 70-300, although bigger & heavier. Cheaper than a Nikon 200-400, or Nikon 300+TC, or 400/2.8, or 500/4..., plus smaller and lighter. Best option I know of for super-telephoto on a cost or size/weight budget.

PS - Nikon 80-400 is under $2k and is sharp, similar size/weight as Sigma 150-500, but auto-focuses much more slowly and thus is a good option for non-action super-telephoto, but not for wildlife or sport shooting.

 

Nikkor 70-300(VR) Lens Test (with and without teleconverter)

I was recently posting in the forum at nikonians.org and erroneously stated that the Nikkor 70-300 (VR) could produce better results with a Kenko (Pro-300) 1.4 teleconverter (provided focus could be achieved and sufficient light) than by shooting without the teleconverter and enlarging. That claim was based on preliminary informal testing. After more rigorous formal testing, I then concluded that was incorrect - that it is better to shoot the 70-300 without the teleconverter, period. After more informal experience, I now conclude that, although it is possible to get better results with a teleconverter, I usually don't ***. With teleconverter, results are noticeably softer and chromatic-aberration/purple-fringing is worse, plus tone & colors are a little off. Here are some samples - you be the judge. Note: the image taken with the 70-300 with TC is 100%, the 105 and 70-300-Plain versions have been enlarged to be the same size as the 70-300 with TC. Image tone has been tweaked a tiny bit to make images look more similar. All images are unsharpened. the 105 version is included as a reference. I also conclude that the 70-300 plain is much softer than the 105 prime. I suppose that should be expected. These samples are shots of a really crappy printout of a test chart, shot with my D300 at f9, using a tripod of course (70-300 versions shot at 300mm from about 6 feet away, 105mm version shot from about half that). I also shot many other things but the results were the same.

*** Better results are achievable with teleconverter by stopping down more (than f9, like f13,14, or even 16), thus I now reserve 70-300+TC for cases when subject is still and I've got a tripod.

105mm prime (for reference comparison):

70-300 Plain:

70-300 with Kenko (Pro-300) 1.4 Teleconverter:

 

Update @2009-06-22: After even more shooting the 70-300 with and without the teleconverter, it seems that occasionally shooting the teleconverter may prove better than without. For example, when chromatic aberration is not a problem, or is corrected. For me personally, I think the case for shooting with the teleconverter versus without is when I can achieve more accurate focus, manually, with a bigger target in the viewfinder to work with. This case is more common with close-up subjects than far-away, and is why I use the teleconverter more for macros than wildlife.

 

Macros

Some love shootin' 'em, some dont. Depth of field runs very shallow shooting macros, especially with a DSLR. This can be difficult to work with, but also yields some very satisfying results when ya "nail it", meaning having a composition, focus, aperture, and angle that play nice together artistically.

Extension Tubes

First off I'd like to say: if you are into close-ups/macros, get extension tubes. These work great with ALL lenses for shooting closer up. When I first got mine, the only lenses I used them with was my primes, especially my macro lens, for shooting closer up. On a prime, it goes like this: more extension means you can get closer to your subject and still attain focus (thus increasing effective magnification whilst maintaining full optical image quality of the lens, unlike with a teleconverter) - focus range is miniscule and anyway if you're not wanting to focus at near closest distance with the tube then you ought to take it off. Recently (@6/2009) I discovered the very different and useful behavior of extension tubes on zooms: More extension means much closer focusing at wide-angle setting, and much less closer focusing but increased working distance at full zoom. The effect of this is that you can put an extension tube on a zoom and get the largest reproduction ratio at the widest setting (much to my surprise, although perhaps it shouldn't have surprised me since thats exactly the behavior of a compact super-macro setting). But also, you can increase focal length and move back, without sacrificing very much magnification. For you Genesis and irony fans: "Gotta zoom in to zoom out". This is really good news for those of us that like to go out with a telephoto zoom, without the size/weight of their macro lens. and still be able to shoot macros at whatever magnification whilst simultaneously from whatever working distance. In this regard, a telephoto zoom with an extension tube is better than a true (prime) macro lens.

Teleconverters for Macros

Same principle as for using teleconverter for long shots (see above): Leave it off if you can, put it on if you must. More specifically: if you can get adequate magnification and you can get focus plane where you want it without the teleconverter, then do it. If you can't get adequate magnification without it, or you can't get focus plane where you want it without a teleconverter, then use one.

 

Using Extension Tubes & Teleconverters Simultaneously

Use extension tubes without teleconverter if sufficient magnification attainable and working distance acceptable, because image quality is best this way. If you need more working distance or you need more magnification and you are out of extension tubes, add the teleconverter. For best magnification, but requiring closest focus distance, put the teleconverter on the camera, and the extension tubes closest to the lens. For lesser magnification boost, but greater working distance - put the extension tubes on the camera and teleconverter closest to the lens. For something in between, sandwich the teleconverter in between the extension tubes.

 

Reverse Mounting a Second Lens (with aperture ring) Over The First Lens

Ebay has a variety of "macro coupling ring" sizes for connecting the two lenses together front-to-front, or Adorama has a 52-52 which can be used in conjunction with a step-down ring if necessary.

Manification Ratio is focal length of first lens divided by focal length of second (reversed) lens. For example, if you reverse a 50 over your 105 you get about 2:1 reproduction. If you reverse a 35mm over your 200 you get about 6:1. Set reversed aperture wide open and focus at infinity, then adjust final aperture and focus using first lens / camera. Image quality is not as good as plain extension tubes nor as good as reverse mounting a single lens directly on the camera, due to glass stacking, but greater magnification is attainable, and auto-metering and auto-focus still work.

 

Reverse Mounting a Lens (with aperture ring) on your Camera

An old trick, cherished by crusty old photographers, poo-poo'd by pampered new ones: makes taking pictures harder, and doesn't improve reproduction ratio as much as reverse mounting a wider lens over a less-wide lens.

Wider lenses means more magnification - the opposite of forward mounting. Working distance: very short.

The up side is that there's no light loss, and image quality is good since there's no stacking of glass. The down side is that you have to adjust everything blindly, but you can get some great shots if you iterate. More specifically: You have to set exposure manually with no help from camera metering, and focus manually - with aperture already stopped down (read: dark), or focus and then stop down (read: tripod).

You can also reverse a slightly telephotic lens but the magnification ain't as much - reserve this for when you forgot to bring your extension tubes or other macro equipment.

Reversing a very telephotic lens makes no sense - reserve this for when you're really bored ;-?

Ebay has the hoo-de-kai for mounting the lens for under $10. Nikon's is usually $40. - I have the cheap one, which so far is fine for reverse mounting a 50/1.8. If I were going to hang a heavy and expensive lens off it I'd feel more secure with the Nikon brand.

PS - You can hold the lens to the camera with your hands and/or gaffers tape, if you're desparate and without reversing adapter.

True Macros Lenses

A true macro lens is one whose primary design criteria was close focusing. As such, they have the best close focusing performance and highest image quality for pictures shot close up. However, you can get equal or nearly equal image quality just by using extension tubes with any other sharp lens, although you won't have the same close focusing performance, in terms of ability to focus, minimum aperture, and loss of light that needs to be made up for some how.

 

Macro Rig Recommendations

For increased magnfication, first:

• Use extension tubes

then, for more magnification:

• Add teleconverter, or
• Reverse mount wider lens over less-wide lens

and for maximum:

• Do both (see below).

If you're a real man (or a macho woman), you could also try:

• Reverse mounting a standard or wide-angle lens (with aperture ring) directly on your camera.

If you're a real woman (or a really really macho man), you could also try:

• Reverse mounting a standard or wide-angle lens without an aperture ring directly on your camera (requires either taking pictures with minimum aperture, or a hardware mod to the lens...)

 

Example Macro Rigs

For extreme macros, the following configuration is awesome:

Camera -> teleconverter -> extension tubes -> 105mm macro lens -> 50mm reverse mounted -> subject.

- focus the 50mm at infinity and open wide.
- Use the 105 for focusing the shot & setting aperture.
- I use a 1.4 teleconverter, but any length will work.

This yields a reproduction ratio of about 5:1. Only need 4:1? - shorten the extension tubes. Only need 3:1? - leave off the extension tubes. Only need 2:1? - leave off the teleconverter and extension tubes. Only need 1.4:1? - leave off the extension tubes and 50mm reversed...

I made a bracket from a 105mm lens hood that holds the Nikon macro flash mounting ring in place whilst the 50mm reversing ring hogs the threads of the 105. This way I can inject light onto a subject whilst the 105 is stopped way down for some great extreme macro shots.

Premium extreme macro rig:

 

Setup without Macro Lens (for when travelling light)

I've found the 50mm/1.8 lens a great asset when travelling light - in addition to being small & light, its sharp, fast, and verstatile for macros:

MFD is minimum focus distance, which is measured from focal plane to sensor. Distance from front of lens to focal plane will be anywhere from a few to several inches less.

Combo	Ratio	MFD	Comments
50mm Lens Alone	1:6.6	1.5 feet	Not suitable for close-ups macros.
With 12mm extension tube	1:4.2	~1 foot	Good for big flowers or groups of little ones...
50mm With 20mm extension tube	1:2.5	~9 inches	Good for small flowers and safe non-skiddish insects.
With 36mm extension tube	1:1.4	~8 inches	Now were into portions of flowers and small safe non-skiddish insects.

Note: the above combinations allow pop-up flash to be used since there is still enough room to illuminate subject. Also, ambient shots may still be doable. Adding additional extension almost certainly means macro ring flash since the lens will be very close to the subject, light is lost, and depth-of-field is very shallow. When travelling ultra-light the macro capability stops here. Note: a teleconverter can be placed between the extension tube and the camera, for greater magnification, but this only makes sense if the reduced working distance of adding another extension tube is intolerable, since otherwise adding another extension tube would accomplish the same thing with less image quality degradation. Of course, adding the teleconverter after you're all out of extension tubes is an option, but I'm not sure if the quality would be better or worse than reversing the 50 over a telephoto lens, which if set at 85mm - magnification would be about the same, as is working distance, but 70-300 with 50 reversed vignettes heavily at shorter focal lengths.

Ultra-light travel macro setup:

The above rig takes beautiful 1:1 macros, has about 3-4 inch working distance from front of lens, and pop-up flash has plenty of oomph and angle to illuminate subjects at minimum aperture & ISO. Add a lumiquest popup diffuser if you like softer light.

 

If ring flash is available, extreme-macros are possible thusly:

68mm tube	~1.36:1	~7 inches	Extreme Macro - usually requires f22 (minimum aperture) and ring flash, unless extremely shallow depth-of-field is desired (not usually).
50mm reversed	~1:2		Hardest to work with since you have to use trial & error metering and try to focus with aperture already adjusted manually. However, there is no loss of light due to the extension tubes that would be required to accomplish the same thing. Same image quality as lens + extension tube, I think.
70-300 at 70mm with 50mm stacked.	1.4:1	1 inch in front of lens	Very hard to find subject, and very shallow depth of field. Very similar to working with 68mm extenstion tube, except aperture of 70-300 can go down to 32 instead of the 50's f22 minimum. Most compelling reason to do this instead of extension tubes is if the full 68mm complement not brought (travelling ultra light), or the 70-300 already had the 50 stacked on it for more extreme macros and shot desired without switching lenses around. Heavy vignetting.
70-300 at 200mm with 50mm stacked.	4:1	1 inch in front of lens.	Very, very hard to find subject, and very very shallow depth of feild, even stopped down to f40 (70-300 minimum at 200mm). You can take full-size pictures of bug faces if you can focus on them before they disappear. - Try to focus on the eyes ;-} Ring flash usually required.

Extreme Macro Rig, Light:

At 300mm, this rig takes 6:1 extreme macros, but you could substitute whatever long lens you are carrying for the 70-300. At 70mm you've got 1.4:1. At these magnifications, minimum aperture and a macro flash is good. Note: Macro flash bracket capability added to 70-300 lens hood by strategically drilling out 4 divits into which the 4 tabs of the macro ring click (not shown, but trust me - they're there).

 

Conclusion #1: 50mm/1.8 with extension tubes, within its limits, takes macros as good as my big heavy expensive 105mm/2.8(VR). Those limits are:

• Loss of light due to tubes - if enough ambient, or flash supplementation is OK, this doesn't matter.
• Minimum aperture is less - if f22 minimum is good enough, this doesn't matter.
• Range of focus is far less - to change magnification by more than a little, you have to change tubes.
• Working distance is less - flowers dont care, but bees do. This can also be a plus, if only small space available.
• Angle of view is wider - busier backgrounds. - Bokeh of 50mm with tubes seems fine, but I dont have the mileage yet to really know.
• Autofocus doesn't work as well - doesn't usually matter since most macros are shot using manual focus.
• No VR - doesn't matter as much at 50mm as it does at 105mm (or 500mm).

Put another way: Once the proper tube is in place for the desired magnification, and provided there is enough light / shutter speed will be high enough. (Point: if significant ambient contribution, shutter speed may be high enough to hold camera steady enough, but not fast enough to freeze motion adequately), and neither longer working distance nor very small apertures is desired, the 50mm/1.8 plus extension tube(s) will do as good as a true macro lens.

And, last time: If you dont have quite enough light for the 50mm/1.8 with tubes, or need maximum depth of field, or greater working distance, or really want a narrow field-of-view in the background, or manual focus is problematic, or if you want to be shooting at different magnifications without missing a beat (or risking dust on your sensor), or need VR to keep steady, the 105mm/2.8(VR) is a better lens to have mounted, but if its not in the bag, you may be able to get just as good, or almost as good shots without it.

Conclusion #2: If you also bring along your macro ring lights, you can also rig up a very nice extreme macro setup whilst still travelling light.

 

General Settings Descriptions (an incomplete list)

See also: My Default Settings.
See also: My Default Settings.
See also: My Default Settings.

Common Lens Settings (General Descriptions)

• AF/M - autofocus to start, with option to finish manually. Manual finish should only be used with AF-S, unless you use AF-ON only for focusing, and then get your finger off it.
• M - manual: Probably should just always use setting on camera-body instead of lens, since not all lenses have it, but camera body always works.
• VR
  • Normal for almost all hand-held shots: could be turned off for wide angle and/or fast shutter speed photos, but then I have to remember to turn it back on.
  • Active - need to remember to use this when shooting from car. Also used for shooting from horse-back, or helicopter...
• OS (Sigma Lens)
  • '1' - Normal/Active
  • '2' - stabilize vertically only - used for shooting objects moving mainly horizontally whilst panning. Not sure what happens if ya forget and leave it on '1' then pan.
• Focus Limit - Restricts range to reduce hunting.

Camera Body Settings (General Descriptions)

• Focus Mode (Lever) Switch (On Front By Lens)
  • AF-C - Continuous servo. This setting required for dynamic focus tracking, otherwise it functions same as single point.
  • AF-S - Adjust focus until first lock then stay.
  • M - Same as setting lens to manual focus, if it has that setting - bypasses autofocus entirely. Camera always operates in Release-Only priority when manual focus.
• Shooting Mode Dial
  • S - Single shot.
  • cL - Continuous Low Speed
  • cH - Continuous High Speed
  • Lv - Live View, meaning use LCD like a point-n-shoot.
  • Clock Icon (Self Timer). Note: Performs Autofocus and metering upon initiation, just waits to flip mirror and release shutter. I often use exposure delay with self timer when using telephoto lens to reduce mirror vibration.
  • MUp (Mirror Up) - First press does focus and metering and raises the mirror, second press or 30 seconds later the shutter flys.
• Exposure Mode Button
  • P - Balance shutter speed with aperture.
  • A - User selects aperture, camera sets shutter speed for correct exposure.
  • S - User selects shutter speed, camera sets aperture for correct exposure.
  • M - Manual shutter & aperture. Note: Auto-Flash and Auto-ISO still function in manual exposure mode.
• Flash Compensation Button - Increase or decrease computed flash output by this amount. Note: its affect on global exposure depends on its relative contribution. If entire exposure is due to flash, then it behaves the same as exposure compensation, except has affect in manual exposure mode too. If only supplying a little fill, will only affect the fill.
• Exposure Compensation Button - Does not affect shutter nor aperture in manual exposure mode, but WILL affect flash output if flash is firing. In 'A' mode, will affect shutter speed, in 'S' mode will affect aperture, and in 'P' mode may affect both.
• Flash Enable Button - Pops Up the built-in Flash. On D300 if flash is up it will fire, always. Magnitude may be miniscule or massive depending on settings, but it will fire.
• Depth-of-Field Preview Button - Re-assignable, defaults to depth-of-field-preview. Can sometimes work with dial too.
• Func Button - Re-assignable, defaults to nothing. Can sometimes work with dial too.
• ISO Button - Adjusts fixed ISO value (non-Auto), or ISO minimum value (Auto). For Auto-ISO setup, menus are required.
• WB Button - Change white balance setting - only used in JPEG-Only mode (rarely).
• Qual Button - Change image target file type - may be used to select JPEG-Only mode (rarely).
• Metering Mode Dial - Selects basis for automatic exposure (shutter/aperture) settings - ignored in manual exposure mode.
  • Center Weighted - Base on central area of frame (fill flash)
  • Matrix - Base on whole frame and engage scene recognition intelligence (fill flash).
  • Spot - Base on small area around focus point (probably "actual" but maybe "selected" - dunno). Flash for spot area only - ignore everything outside.
• AutoFocus Mode (Lever) Switch (On Back By LCD)
  • Auto-Area - Let Camera Choose Focus Point, based on Closest Subject, Face Recognition, Scene Recognition, and whatever else Nikon thinks up. Discussed some more below.
  • Dynamic Area Tracking - User selects initial focus point, camera tries to track subject as dictated by relevant user settings. Discussed in much more detail below.
  • Single Point - Focus on object at selected focus point only.
• Focus Point Selector (rocker) - Select focus point
  • Auto-Area - does nothing.
  • Dynamic Area Tracking - select initial/central focus point, actual focus point moves as focus tracks.
  • Single Point - Selects actual (fixed) focus point.
• Focus Point Selection Lock (Lever) - Keep selected focus point from changing - ignored in Auto-Area mode.
• AF-ON - Initiates Auto-Focus - ignored in manual focus mode. Usually not used (although not ignored) when auto-focus assigned to shutter release, but can be used just to check focus without shooting.

Menu Settings (General Descriptions - Very Incomplete)

• Banks (4) - shooting & custom - great for businesses where up to four different people share a camera, but not very useful for setting up scene modes for a single user. See D300 Gripes above (or not).
• Auto ISO - Select maximum ISO and minimum shutter speed. Camera cranks ISO instead of allowing shutter to fall below minimum. Shutter speed will fall below minimum if ISO greater than max would be needed, to avoid dark pictures. I have my gripes about the way this is implemented on the D300, but it has its usefulness.

• See My Default Menu Settings below.

 

My Default Settings (see above for general descriptions):

Guiding light: Set to maximize odds of getting a fairly good shot if suddenly and fleetingly presented.

Lens Settings (My Defaults)

• Focus: A/M. I rarely change this. Since AF-ON used for activating autofocus, manual focus can be used without changing this setting - just be keeping finger off AF-ON and twist the focus ring. Only time I change this to 'M' is if I've changed AF-S to focus priority for trap focus and I want to postpone changing it back for a shot or three, in which case 'M' must be selected to fire a shot that is not considered focused at the selected point.
• VR: On/Normal. I shoot mostly handheld on land, so I shoot mostly with VR:On/Normal. Every now and then I turn this off if I'm shooting still subjects in bright light with open apertures, but usually not, since doing so runs the risk of forgetting to turn it back on. When I'm at my best I remember to put it on Active when I'm shooting from a moving car...
  • Sigma Lens: OS: '1' - stabilize vertical + horizontal. So far, I havent use mode '2' (stabilize vertically only) except experimentally.
• Zoom: Locked. I'd rather have the lens locked automatically when camera is off, and unlocked when camera is on, but thats not an option. Locked state keeps lens from creepin' & slidin' during handling & carrying.
• Focus: Unlimited. By default, I want to be able to pick up the camera and focus at any distance. Sometimes I even remember to flick this switch when I'm only shooting at a limited distance ;-}

Camera Body Settings (My Defaults)

• ISO - depends on lens and expected light, but manual and biased high, say 400 for wide-angle lens, 500 for mid-range zoom, 640 or 800 for telephoto. If bright conditions, halfen; if dark conditions, double... I presently reserve Auto-ISO for special conditions.
• AF-C. This supports continuous shooting or single shot, moving or still.
• Exposure Compensation: 0. Sometimes increase, sometimes decrease before shooting - never know.
• Flash Compensation: 0. Sometimes reduce flash before shooting, rarely raise it.
• Shooting Dial: Continuous-Low (instead of 'S'). This way I can squeeze out a single shot, or keep down for multiple.
• Mode Button: P. Biased towards fairly high shutter speed, and once reasonably high, starts shutting down the aperture. I used to keep it on aperture priority all the way open, since this does actually maximize odds of getting some kind of shot thats in focus, but lens performance is so much better stopped down a bit that its usually a good idea if there's enought light, even if you dont desire more depth-of-field.
• Metering: Matrix. I used to use spot but spot is really 3mm circle, and results vary wildly for a small subject depending on how much of the 3mm circle is occupied. Matrix is most often right.
• Selected Focus Point: Center (Unlocked).
• AutoFocus Mode: Dynamic. In case subject moves.
• Flash Mode: Rear (Slow). The main reason I do this is so the camera doesn't choose a fixed minimum shutter speed in 'P' & 'A' exposure modes when flash is used; It just adds a bit of fill flash. I often snap my first flash shot this way, then switch to manual mode and down-throttle the shutter for decreased ambient as desired and/or down-throttle the exposure/flash-exposure. A secondary benefit is that ambient trails fall behind moving objects where they usually belong, instead of leading them. The down side is that sometimes the first shot is over-exposed, since it always adds flash but never subtracts ambient, but you can anticipate this and compensate. Another down side is that sometimes there is ambient blur at lower shutter speeds. Again, this can be anticipated and compensated, and can also be viewed as a feature since it captures motion blur. Its less of a feature when it captures camera shake ;-)
• ISO (Again, For Emphasis): depends on lens and expected light, but manual and biased high, say 400 for wide-angle lens, 500 for mid-range zoom, 640 or 800 for telephoto. If bright conditions, halfen; if dark conditions, double... I presently reserve Auto-ISO for special conditions.

Camera Menu Settings (My Defaults)

• Banks (A-D) - Can mix and match main settings with custom settings, but I use them in pairs, if I bother at all.
1. A - Point & Shoot settings for when I hand the camera to someone else to shoot with. I keep these banks most like a 2D camera.
2. B - My Primary camera settings - almost always use this, even for experimenting. You have to manually put settings back when your done anyway since there's no "reset to user default" so there's hardly any point to the experimental banks.
3. C - My Secondary - Never used.
4. D - Experimental - For goofing around / learning, although I mostly use my primary settings for experimenting now-a-days.
• AF-ON for focusing (instead of shutter-release). Reasons:
  • I can lock focus just be taking my finger off the AF-ON button.
  • I can skip autofocus altogether and just use manual focus without having to switch modes, just by not pressing the AF-ON before shooting. This to work, release priorty must be set for AF-S.
• I can re-engage focusing without re-engaging VR (which is tied inextricably to the shutter release button).
• I can use trap focus, as long as I remember to set AF-S to focus priority then switch it back.
• The down side is I have to switch it to shutter release to use remote shutter release, then remember to switch it back.
• PS - Pressing two buttons to take a picture instead of one becomes second nature after a while.
• AF-C Priority: Release + Focus (instead of release-only priority), so it takes fewer pictures in a burst, but more of them are focused on something (not necessarily focused on what you want).
• AF-S Priority: Release (instead of focus priority). There's a dot that always says whether I'm in focus or not - this way I can choose to release, or not. Also, it allows me to do a focus, recompose, shoot sequence which can not otherwise be done. Admittedly you can change the selected focus point but I find it can be less convenient, more limited, and takes me longer.
• Adobe RGB (instead of Standard RGB) - Does not affect RAW files, but does make JPEG previews look more like what the RAW file would be.
• High ISO NR: Low. The default of medium is way too much for real photos - its for marketing noise comparisons.
• Func button: Flash-Off. Its the only way to quickly shoot one without flash if pop-up is raised or an SB-600 is on and ready. (You have to hold the On/Off switch for a full half second before it will turn off, which may be an eternity when its time to shoot pronto).
• Depth-of-field Preview Button: Flash Lock. I switched from default when I had a certain subject that always blinked upon pre-flash. Now if subject reacts adversely upon pre-flash I just hit the button, then shoot again. Depth-of-field preview is only marginally useful anyway.
• NEF Recording 12-bits, lossless compression. This way, I can use high-speed burst by flicking a switch instead of having to go into a menu.
• Live View: Handheld, Continuous-Low. I mostly shoot handheld. Continuous-Low covers one shot about as well as a good burst.
• Dynamic AF area: 21 (instead of 9). Not sure this is a good idea. Truth is: the more points the more likely a surrounding object will steal focus. On the other hand if ya cant keep it within the point group, you lose focus anyway - its a trade off.
  • Note: This setting should be on the body, an in fact sometimes I program the 'Func' button with it when I'm not using bracketing. It is critical to dynamic tracking success rate.
• Focus Tracking with Lock On: Normal. Reminder: 'Off' should only be used if not tracking a specific subject but you just want to focus on whatever wanders under the selected focus point, regardless of what it is. Most people misunderstand this setting, and if they claim that 'Off' is better then it means they are happy enough to have something in focus, even if its the wrong thing. Again: if 'Off' seems to improve tracking, its because you already lost the lock on your original subject and the camera has chosen another (wrong) one. Last time: This setting only comes into play when subject focus has already been lost.
  • PS - When learning, its best to leave this setting on 'Long' - makes it easier to tell when you've lost tracking.
• AF Activation: Off. Setting should be called "Autofocus Via AF-ON Only". See above discussion.
• Self-timer delay: Value doesn't matter, and in fact I rarely use it preferring a wireless remote shutter release to a timed one. I only use this when lazy or remote shutter release not with me.
• Exposure Delay Mode: Off. I regularly turn this on for still shots from tripod instead of mirror-up, since it requires one less click to take a picture. I always think my camera is broken when I forget to turn it back off.
• Flash cntrl for built-in flash: TTL. Change to Commander mode for off-camera flash. Rarely use manual preferring compensation instead. Although perhaps I could use manual to alleviate need for flash-lock. Hmmm... An alternative is to leave in Commander mode since the flash behavior is the same when the built-in is enabled, except that the built-in flash amount will depend on the sum of the global flash comp value and commander set value, so its a bit more complicated and subject to goofs, plus it issues IR pulses to communicate with all its remotes which aren't there, which bothers me aesthetically.
• ISO (Yet Again, For More Emphasis) - depends on lens and expected light, but manual and biased high, say 400 for wide-angle lens, 500 for mid-range zoom, 640 or 800 for telephoto. If bright conditions, halfen; if dark conditions, double... I presently reserve Auto-ISO for special occasions.

 

RAW Files, Raw Image Data, & Conversion

Included in every RAW File:

• JPEG Thumbnail
• JPEG Large Preview
• Camera Settings
• Raw Image Data - This is really RAW - it does not even have colors assigned to pixel sensor readings yet.

When a picture is taken, the camera stores the camera settings & unadulterated raw image data, and also applies camera settings to the raw image data and stores the result in a couple JPEGs that can be extracted for quick previewing. These JPEGs are what the camera plays back, and is what computer software displays while they're busy converting the raw image data into viewable form, which takes most computers several seconds to do. Some software will also cache raw conversions to speed up display, but the first one displayed is always one of the JPEGs.

 

Which Settings Affect Raw Image Data

Assuming a given shot has been composed and focused.

Directly:

• Shutter Speed
• Aperture
• ISO

Indirectly:

• Active D-Lighting (assuming an auto-exposure mode). - The after-the-fact adjustments that are part of Active D-Lighting are not applied to Raw Image data, but exposure may have geen adjusted to pave the way for applying active d-lighting to the JPEG image.

Not At All:

• All other settings. Note: Other settings will affect the JPEG preview image embedded in the Raw File - you cant stop this, but they do not affect the raw image data. Note Also: ViewNX & Capture NX2 give the illusion that other settings affect the RAW file, but its just because they affect the rendering after conversion. Other raw converters ignore the various camera settings. I dont think they even know how to read them.

 

RAW Conversion

Raw conversion is the process of converting raw image data into viewable form. Nikon software folds camera settings in with the RAW data during conversion, 3rd party software just folds their own software settings in with the RAW data during conversion. Nikon software allows you to change camera settings after the picture is taken. 3rd party software just ignores camera settings - I'm guessing they dont even know how to read them. Some people think that Nikon also knows how to interpret the raw image data better than 3rd party software (Nikon does not claim this). Other people think 3rd party software understands how to interpret raw image data as well as Nikon software. To the best of my knowlege, Adobe makes the only 3rd party conversion software that knows how to properly assign colors to the converted data (via Camera Model Specific Color Profiles). Other differences in conversion relate to things like sharpening, but do not relate to raw data interpretation. Update: I think PhaseOne now (@6/2009) has camera-specific conversion profiles - if I remember correctly.

 

The D300 Sensor

Technology in a Nutshell

CCD sensors generate analog voltages that must be converted off-chip to digital. CMOS sensors generate digital values directly (on-chip). CCD sensors used to produce better image quality overall, but CMOS sensor technology has advanced in recent years to be able to produce better images. Other differences have to do with hardware/system considerations which dont fit in this nutshell. The D300 uses a CMOS sensor which was made by Sony.

12 bit sensor? 14-bit sensor?, or switchable??

There is some controversy around this, but my thoughts are: Its 12-bits, period. Same sensor as Sony uses in their (12-bit only) Alpha 700 (I think that was the model). I think the 14-bit mode is just a software trick. The reason I think this is that it is very CPU intensive to shoot 14-bits. It could be that there is a hardware shift used in conjunction with software processing - I really dunno, and from the best I can tell its a well-kept secret between Nikon and Sony. In any case, its definitely not a 14-bit sensor that you can down-throttle to 12-bits to save space. I think if it really had a 14-bit hardware mode, Nikon would be boasting about the technology, and Sony would be using it too. As it stands, Sony does not use the 14-bit mode of this or like sensor. I think 14-bit mode is kinda like digital zoom - a convenience thing more than a true-value thing. Actually, if I'm right (that its just an after-the-fact software thing), one would expect to be able to switch from 12 to 14 bits after a picture has been taken using Nikon software. On the other hand, the marketing people may have kept the software engineering team from implementing this after-the-fact in software, hmmm... I think I just switched to believing there is probably a hardware component to the 14-bit mode, although I still think its of the kludgy/not-quite-ready-for-prime-time variety.

*** This just in (2009-05-18): I recently returned from a trip to Zion National Park where I switched to 14-bit mode whilst shooting narrow dark canyons with bright sky overhead (read: high-dynamic-range). I underexposed the shots to keep the highlights from clipping so bad (and to keep from having to combine multiple images shot at different exposures). It seemed like I was able to add exposure to the dark areas during post with less image quality reduction than I would have had in 12-bit mode - could be my imagination - wish I'd a shot a few 12-bit'ers for comparison - oh well...

 

Taking the Picture (Generally)

Before a shooting spree, consider:

• Battery Charge
• Shots Remaining
• Sensor Clean
• Lenses Clean
• Lens Hood
• Lens Cap

and whether camera has been returned to default state.

1. Consider the Shot.
2. Consider (changing) the Lens, including:
   1. TC
   2. Extensions
   3. Filters
   4. Hood
   5. Clean
   6. If lens changed, clean sensor.
3. Consider Lens Cap.
4. Consider Lens Settings.
5. Consider Camera Body Settings.
6. Consider Settings On Menus.
7. Stabilize, Focus, Shoot

Acronym: SLCSSFS - Shot, Lens, Cap, Settings (x3), Stabilize, Focus, Shoot.
Short Lovers Congregate (on) Some Sultry Fridays (through) Sundays

Also: Lights (and metering/exposure settings), (Other) Camera Settings, Action (including AutoFocus and VR).

 

Focusing

Composition aside, focusing (and camera stabilization) is one of the most important parts of taking a picture (along with ISO/lighting/exposure), not that out-of-focus pictures cant be fun too, but a good photographer should be able to focus when demanded, yet also be able to enjoy lack-of-focus when the out-of-focus magic occurs. Also, rather than providing detailed information about how its various focusing modes work, Nikon leaves it more as an exercise for the user to figure out / experience. And it really requires practice and experience in real-world scenarios to master focusing on moving subjects - very hard to get the hang of it by shooting your waving hands and flailing feet. Hint: you can pretty much see whats in focus and what not, so you dont necessarily have to shoot pictures to experiment with focusing.

 

Auto-Area Focus Discussion

This mode is useful in both AF-C (continuous-servo) and AF-S (single-servo) modes.

This should be first consideration, since under the right conditions it is the easiest and most accurate and reliable way to focus. Use when you want to shoot the things it tends to prefer, like close things, and people. This mode is preferrable to dynamic tracking if the subject will be recognized and it would be too hard to get the selected focus point on the subject anyway. It is my favorite mode to shoot flying birds when they are not recessed in trees, e.g. open sky. I still need more practice with this mode to see what it does and does not focus on. Its a poor choice when you know exactly what you want to focus on, and can focus on it more accurately using an other mode.

Good Applications

• Group Photos / Party Shots
• Birds Flying in Open Sky
• Other animals in open conditions.
• Landscapes / Sky-scapes when nothing undesirable will steal focus.

 

Dynamic Area Auto-Focus Tracking Discussion

Only tracks in AF-C (continuous servo) mode. In AF-S (single-servo) mode, dynamic mode is same as single point auto-focus since it wont track.

Best when you know exactly what you want to focus on, AND you can get it in the cross-hairs (within selected focus point region), AND the camera is able to lock focus on the proper subject (at the selected focus point), AND it is moving or may start moving, AND you would not be able to track it with a single point. Optimal number of points to use is quite simply the minimum number that you can keep the subject in, or return the subject to, and get a focus fix on, before the AF Tracking Timer expires (see Auto-'Focus tracking with lock-on' setting). Unfortunately, this is often not known ahead of time, so if your subject hasn't flown away, or gone into the dugout yet, you can change point count and retry. Experience dictates that more points improves your odds of keeping the subject contained in the focus group, but also increases the odds that focus is stolen by a surrounding object. Likewise, less points improves your odds of keeping a subject in focus that only moves away from the selected point by a little, but increases the odds of losing focus altogether if the subject moves outside the focus points group.

Good Applications

• Specific Players in Sports
• Specific Objects Racing
• Birds recessed in trees, sitting or moving around on branches.
• Other animals scurrying around in their habitats.

 

Dynamic Area Auto-Focus Tracking - Under The Hood

After 5 months of owning a D300 I have the following thoughts about the Dynamic Auto-Focus tracking system, based on a combination of experience and speculation. I'd love to know if anyone else has more detailed or accurate information about how it works - it helps me to have a model to go by. This post is about the middle auto-focus mode switch (cross-hairs on camera back) and AF-C on camera front (continuous servo). Also, when I say focus button, I mean shutter release or AF-ON. Another definition: I call the focus point indicated by the visible rectangle the "selected focus point", and the others "supporting focus points", when the distinction needs to be made. Also: "sensor" and "focus point" are synonymous.

Background: the D300 AF system hardware consists of 51 focus points which provide the following information to the firmware: how far the lens needs to move and in which direction to achieve focus at that point, or else a fault condition meaning the sensor doesn't know. When the magnitude is zero then the point is considered to be in focus. The hardware has no intelligence other than to supply this information to the firmware. Also, the firmware has available the distance at which the camera is focused, or would be if there was something discernable at that distance.

Disclaimer: The following is a simplified model which can explain any autofocus behavior I have seen so far, maybe. It is not knowledge.

Regarding 9, 21, & 51 (NOT 3D): When you press the focus button, the camera reads the selected sensor and if not in a fault condition, issues a command to the lens to move by the amount and direction dictated by the sensor. If it is in a fault condition it issues a command to the lens to just keep moving one direction or the other in the hopes of getting a proper reading at the selected sensor along the way. This is whats known as "hunting". After the lens moves, another reading is taken at the selected sensor and if the magnitude returned is zero, then the camera is considered to be focused, and is indicated by a solid dot in the viewfinder (note: since it took two readings it now knows speed of approach or recession). So far so good, and in AF-C mode, it just keeps doing the above. So, what if it had focus at the selected focus point, but now it doesn't? - Well, it checks the other 8, 20, or 50 supporting points in a presently unspecified order. First one not in a fault condition and whose distance is reasonable wins, or maybe it checks 'em all and takes the most likely one; in any case the lens then moved a little, if necessary as dictated by the chosen support point. Again, the supporting point is checked and if still good, then the camera is considered focused. So, what if all sensors are in a fault condition, or would be focused at unreasonable distances? - It depends on "a4 - Focus tracking with lock-on" setting. If set to "Off", it declares "Game Over" and starts again trying to gain focus at the selected point. This is why you should never select "Off" if it is your intention to track a subject under less than perfect tracking conditions (read: ALMOST ALWAYS), i.e. if it misses re-focusing after one iteration of the sensors, it just focuses on the background or whatever is handy at the originally selected focus point, which is NOT what you want. So, if a4 is set to anything else, it continues to survey the supporting points in the hopes that one of them will not be in a fault state, AND will be in focus at a distance that is not too far from what may reasonably be expected (the camera continually updates subject velocity and direction, as well as distance). Note: It does not just idle, which is why you want to choose some non-Off setting for a4 - so it keeps trying to find your subject, or perhaps an acceptable substitute. If it does this for a while and never finds anything it considers worth focusing on, THEN it declares "Game Over" and starts again trying to gain focus at the selected focus point.

Some things to note:
- Cross type sensors are better at determining focus than horizontal ones. This is one reason dynamic auto-focus tests that use 51 points sometimes fail - the outer points aren't as good at determining focus as the inner 15 points.
- Supporting sensors require good focus conditions to not be in a fault condition when they are checked. This is one reason dynamic auto-focus tests using 9 or 21 points often fail when performed indoors and/or using marginal subjects to track, especially when a4 is set to "Off".
- The camera may switch subjects as it tracks, even though this was not your intention. This is more likely the more supporting points you use or should I say the more potential subject matter in supporting points. This is why fewer points may be better if your subject is moving around other things that are at a close distance and moving at a close velocity. If this happens, your best recourse is to release the focus button until you can get the selected focus point over the correct subject, then press it again. (This is also a good reason to use AF-ON instead of shutter release for focusing - better to continue with VR while waiting to get your subject in your sites). The only other option I can think of is to keep trying to aim your selected focus point at the intended subject in the hope that the camera will eventually favor your intentions for a new subject - this however is more likely to result in continuing to track a wrong subject or finding a new wrong subject. If the latter is the technique you use, then a4 set to Off will more quickly result in a proper new subject, and is why some people use it, but is not what it was intended for. a4-Off is intended to be used when you really are NOT trying to track a specific subject, but really want to focus on whatever subject happens to be at a certain point.
- The camera uses distance AND velocity to determine which sensor is most likely to be on your intended subject if it is not in a fault condition. This is another reason why autofocus tests may fail - if you establish focus with no movement, then move, it may not track. This is also the reason the camera is really good at maintaining focus on a flying bird as it flys through the trees, but really bad at maintaining focus on a bird sitting in a tree that suddenly starts flying.
- Focus points are not really points but small areas or zones, which may cover two different competing subjects - it may pick the wrong subject at a point than you would like. This is especially problematic when your preferred subject is very small, or harder to focus on than other subject matter in the point zone area, or is very nearly at the same distance...
- It is my understanding that the D3 uses the same hardware and same autofocus firmware as the D300, but has a faster processor and hence better dynamic autofocus tracking, since it can more accurately assess changing subject distance and velocity and more quickly track - more likely to stay on the correct subject, and less likely to lose tracking.
- If you happen to be shaking the camera when a sensor is checked it may result in a sensor fault condition. This is another mechanism whereby tracking may be lost or shift to a wrong subject. Ditto for other focusing anomalies.

Regarding 51-3D: I haven't used this too much, since it seems so squirrelly and usually ends up focusing on something other than what I intended. Still, I think its like 51-plain except when the focus jumps sensors it also changes the selected focus point and theoretically its using additional information from metering subsystem. It behaves as if it wont take no-focus-point for an answer and uses a best-fit scheme for choosing a new focus point after checking the supporting focus points just once, instead of re-checking them to try again for what may be more likely to be your true subject - dunno really. Truth is I often end up misfocused using 51-plain too and it may be that this mode actually tracks better, but is just unnerving with the focus point jumping around, potentially revealing the error of its ways.

Regarding Auto-Area focus: I think its like 51-3D except favors closest subject and faces and such instead of previous subject. I mostly use this for one-shots when I'm not too picky about the exact focus point / subject, instead of bursts where I'm tracking a particular subject. e.g. sometimes I set the camera up high in the corner of a room with a wide angle lens at a party and use a remote trigger to shoot. I just want whoever is closest to be in best focus and the ones at the other side of the room to not claim focus. Its also the only way to automatically focus on a subject that you just cant keep in the selected focus point long enough to lock on, like a bird thats flyin around dippin and tuckin.

 

51/3D

Jury is not in on this one. My impression so far is: should work better than 51/Plain, but doesn't, and is unnerving the way the focus point keeps jumping around. Still evaluating...

Single Point Auto-Focus Discussion

Note: In AF-C mode, it functions like a 1-point version of dynamic area tracking mode. In AF-S (single-servo) mode, single point auto-focus is same as dynamic mode.

When you know exactly what you want to focus on, AND its moving in such a way that you can track it good enough with a single point, AND there are other objects nearby that would steal focus if selected focus point wandered off for a microsecond. Note: If the subject is not moving at all, I usually just leave it on dynamic-area-tracking/continuous-servo; not sure if there may be a case for exception.

Good Applications

• Turtle Racing? - Honestly, I never use this mode, but perhaps I should, since 9/21/51 point mode often ends up misfocusing.

2010 update: I use this mode more often now - precisely for the reasons indicated above.

Manual Focus Discussion

Use Manual Focus when Auto-Focus simply is not working, or is not precise enough.

Good Applications

• Extreme Macros, since autofocus may not be precise enough.
• When light is too low for autofocus system. A special case of this is when using teleconverters or extension tubes.
• When focusing through glass, or a screen, or fence, which may confound autofocus, although another way to remedy this is to move focus way past the fooling thing either manually or focusing on something else far away, and then autofocusing.
• When objects are continually coming between the subject and the lens, like waving grass or people. Note: the autofocus system can handle this somewhat, but you still have to get an initial focus, and focus may still get stolen. Manual focus is never hi-jacked.
• Pre-focusing for action shots - but use Trap Focus instead if it will work.
• When object being focused on is elusive - has no texture or contrast, or is transparent. A special case of this is when you actually do want to shoot glass, or a screen, or a fence...

 

Continuous Servo (AF-C) versus Single Servo (AF-S) Discussion

Continuous is intended to be used when the subject is moving (or the subject is still but the camera is moving), but usually works just fine when they're still too. So, I leave my camera in this mode most of the time. The #1 reason I sometimes switch to Single Servo mode is to get AF-Illumination light, which only works in Single Servo mode. Another reason I sometimes switch is when I've programmed a different release priority for it (usually focus priority to take a single shot using trap focusing), although my present default is release priority for both continuous and single servo. Another side-benefit of switching to AF-S (single servo) mode is so focus points are displayed when doing auto-area focus.

Reminder: In Single Servo mode, once gained, focus stays locked by shutter release, even in AF-ON only mode. This is not the case in Continuous Servo mode. This mostly doesn't matter, but matters a lot when trap focusing in continuous shooting modes (cL & cH).

 

Trap Focus Discussion

In some cases, autofocus doesn't work very well, or not at all, ditto for manual focus, but you know there is eventually going to ge something interesting to shoot at a certain distance, at some time in the hopefully not too distant future. In this case:

• Set focus priority for AF-C or AF-S
  • Choose AF-C and select Continuous-Low or Continuous-High shooting mode if you want to shoot a burst, once triggered, that ends when nothing is in focus anymore. Reminder: with AF-C, focus is not locked by shutter release.
  • Choose AF-S and select Single-Shot mode if you want a single shot. Reminder: with AF-S, focus is locked by shutter release, so if you choose a continuous shooting mode it will keep shooting, once triggered, as long as you have your finger down, even if nothing is in focus, since focus gets locked on. Thats an option too, but not usually what you want.
• Set camera to not focus by shutter release, but only AF-ON ( 'AF Activation: Off')
• Set autofocus mode to dynamic-area, or single-point - doesn't matter (Auto-Area can also be used but its usually not what you want). Reminder: dynamic-area wont track and ignores all but selected focus point, since AF-ON must be depressed to continue focus servo / tracking but must be released before triggering. Trap focus is a special case where dynamic and single point focus are same in both AF-C and AF-S servo modes.
• Focus manually or using autofocus at a desired point/distance.
• Aim the camera at an eventual point of interest that does not presently have an object that will be in focus.
• Press the shutter release. - It shouldn't fire a picture until something comes into focus at the pre-focus distance, then it will fire one or more frames, depending on settings.

Hot tip: use a remote trigger that can hold the shutter-release down for you, then you can engage it and walk away and it will collect shots of everything that goes in front of the lens (slowly enough and with enough light / contrast so that focus is detected) at the pre-focus distance, return for your surprise shots, god-willing. Requires Continuous-Low or Continuous-High shooting mode (so it will take more than one picture) and AF-C (so that focus isn't locked after triggering. Warning: If you put it on Continuous shooting mode and AF-S, it will shoot "forever" once triggered (until your card is full or number of shots per burst is reached) since focus will be locked and shutter-release will be down.

 

Other ways to Focus/Trigger photographs

If I spent more time making money and less time futzing around with photography, I'd buy a laser trigger so I could spend more time futzing around with photographing hard to photograph things. The laser trigger is like you see in James Bond style movies where a laser beam when interrupted generates a shutter release signal to the camera. This is a much more flexible and reliable alternative to trap focus shooting, since an object need only break a line and not a point, and doesn't require focus detection to trigger. There's a sound trigger too for when you want to trigger off sound instead of optically. One day, maybe...

Another option which I might yet rig is a set of contacts that will trigger. This way one could have a picture a moment after someone walks through a door (using self-timer or hardware delay module), or shoot a burst of photos the moment the mouse springs the mouse-trap... I would think both "break-to-trigger" and "make-to-trigger" would be useful.

Image Quality

The primary determiners of raw image quality, camera-wise, assuming you've got the camera steady enough, exposed properly, and properly focused..., are:

• Distance to subject - closer the better, but not too close ;-}
• Zoom - Has similar advantages as distance-to-subject, but first thing is get closer, then zoom in.
• Lighting
• ISO, and do not be fooled by Nikon marketing. They say 200 is as good as 100 and 100 is just for slower shutter, but its not true. ISO-100 makes for the best quality images - better than ISO-200. Sometimes the difference is negligible, but sometimes it is substantial.
• Aperture - Middle apertures are sharpest, and exhibit least optical anomalies.

Secondary things:

• Lens. I can hear many of you kicking & screaming as you read that, but it is my experience that you can take a better picture when you're close enough, with good light, low ISO, middle aperture, with a decent lens, than you can if you are lacking even one of those things, but have a better lens. Granted, if you have all those things and a better lens, you can achieve better quality still, thats why I've listed it here.
• 14-bits (instead of 12). If you enhance shadows (and maybe in highlight recovery - not sure), and look carefully, you can see a difference. There is a theoretical improvement in smoothness of mid-tones as well, but I cant see it.

 

Exposure Modes (P*,A,S,M), Auto-ISO, & Flash

Definitions:

• Aperture Priority (A) - Fixed Aperture, variable shutter speed. Auto-ISO interprets base ISO as minimum.
• Program Mode (P*) - Camera chooses both shutter speed & aperture according to a secret algorithm that unfortunately does not take focal length into consideration. Program Mode includes a Program Shift(*) feature that allows you to let camera choose both shutter speed & aperture, but can be biased towards faster shutter or more depth of field. Auto-ISO interprets base ISO as minimum.
• Shutter Priority - Fixed Shutter speed, variable aperture, Auto-ISO interprets base ISO as minimum.
• Manual - Fixed shutter speed & aperture. Auto-ISO ignores base ISO setting and will lower or raise ISO in accordance with exposure compensation setting. In manual ISO mode (without Flash), exposure compensation setting is ignored.

Note: Camera computes shutter speed & aperture based on ambient light, mode, iso/auto-iso, & flash settings / potential-calibrated-through-the-lens flash contribution, as follows:

• Shutter speed & aperture computed based on ambient, mode, and exposure compensation, if not manual mode.
• If flash enabled, flash output computed according to what would be needed for exposure compensation, then adjusted by flash compensation amount.
• If Auto-ISO enabled, ISO will be raised if needed to meet target (compensated) exposure, and if manual mode, it may be lowered.

Some consequences of this are:

• ISO will never be raised (via auto-iso) if there is enough light via ambient/flash.
• If flash is primary light source, exposure compensation & flash compensation are additive and both affect flash contribution.
• If ambient is primary light source, exposure compensation only affects shutter speed & aperture, and flash compensation only affects flash contribution.

(@6/2009) My usage:

• 50% Aperture Priority
• 30% Program Mode*
• 20% Manual
• 0-1% Shutter Priority

Arguably, (and once shot composed, lit, focused, ...) aperture is the single most important exposure setting. It determines sharpness & depth of field. Thus, its rarely a "dont care". Shutter speed is more often a dont care, provided its fast enough. Usually if I care about exact shutter speed, I use manual mode to set shutter speed, and set the aperture to what I want and use flash and/or auto-iso to make up the difference. Shutter priority mode makes me very uncomfortable :-} If I only care about minimum shutter speed I use aperture priority mode and set the aperture to what I want and rely on auto-iso to make up the difference. I use program mode when I'd rather have aperture opened than ISO increased when light is waning, or when I'd rather have aperture stopped down than shutter quickened when light is abundent.

Flash Modes

Default flash mode is front-curtain fast-sync, which triggers the flash once shutter is open, leaves it open for a short time, then closes it. How much ambient comes in depends on how long shutter is open, which if not set explicitly via manual or shutter priority modes, will be fixed by the flash shutter speed setting (e2), in P & A modes, or shortened to prevent ambient overexposure. I prefer rear-curtain slow-sync, reason: if ambient contribution is low front-curtain bout same as rear-curtain. However, if ambient contribution is substantial, rear-curtain ambient ghosts trail flashed subjects, which is more natural. Front-curtain ambient ghosts lead flashed subjects, which looks less natural, usually. Shutter speed is same as without flash, which is why I often switch to manual mode for flash photos (so I control explicitly the flash/ambient ratio), or in auto-modes I combine negative exposure compensation with positive flash compensation to reduce the ambient / flash ratio.

Hyperfocal Focusing

A "sister" consideration of aperture/depth-of-field is hyperfocal focusing. Aperture/depth-of-field is the consideration when there are near objects you want focused, and nearer/farther objects you don't. Hyperfocal focusing is when you want both near and farthest objects maximally focused, and you dont want to stop down any further than necessary to do it. For example, you want a car & the mountain behind it to be focused.

Two options:

• not optimal: Focus on the car and stop down so mountains are mostly focused too. Waists depth-of-field on objects in front of the car.
• not optimal: Focus on the mountains and stop down so car is mostly focused too. Waists depth-of-field on distant objects.

The optimal option is to focus in between and stop down less. So the questions are:

• where to focus
• how far to stop down

which depends on:

• focal length
• distance to nearest object
• how coarse your sensor is (defines "Circle Of Confusion" / what maximally focused means).

Qualitatively:

• wider the angle, nearer the hyperfocal distance
• further the nearest object, further the hyperfocal distance
• greater the aperture, further is hyperfocal distance.

Endpoint-wise:

• If you're shooting wide, and nearest object aint too near, you dont have to stop down that much, and you can just focus a little past the nearest object.
• If you're shooting narrow, and nearest object is close, your gonna have to stop way down and focus further past the nearest object.

See the reference section for links to more exact answers.

Here's the chart for a D300 taken from DOFMaster website:

*** Hyperfocal distance in this chart is twice the distance to the nearest object to be in focus.

Examples:

• Shooting 15mm with nearest object at 7 feet, you can shoot a wide aperture of 2.8 and focus at 14 feet.
• Shooting 50mm with nearest object at 9.5 feet, stop down to f/22 and focus at 19 feet.
• Shooting 80mm with nearest object at 50 feet, stop down to f/11 and focus at 100 feet.

The D300 equation is Hyperfocal distance = focal length squared divided by (aperture times 6.1) - focus at hyperfocal distance to get objects half that distance through infinity in focus.

 

Active D-Lighting

What It Does:

• I'd heard "if not in Matrix Metering mode, it does nothing" - I've found it functions in all metering modes.
• Assesses dark-to-light range, and if highlights are being over-clipped, decreases exposure, to reduce highlight clipping. I do not know the exact algorithm for determining when to reduce exposure and by how much, but here are results from a few brief experiments at each setting *** (see below):
  • Low - Zero on very low contrast shots, small on medium-to-high contrast shots.
  • Normal - Zero on very low contrast shots, and up to 2/3 stop in high contrast shots.
  • High - Small even on very low contrast shots, and up to 1 full stop in high-contrast shots.
• To raw data, it may reduce exposure, but other effects are only done to JPEG image data. Shadow detail must be enhanced manually in post-processing when using a Raw (non-CaptureNX) workflow. See CaptureNX specific notes about active d-lighting below.
• To JPEG image, highlights are recovered, & dark areas are lightened.

For the first half year of D300 shooting, I turned Active D-Lighting off to keep things simpler for myself and to better understand the camera's basic exposure / metering system. During that time I found the camera often tended to overexpose high-contrast scenes and I began compensating by underexposing by 1/3 or 2/3 stop regularly in high-contrast / harsh light conditions, then bringing up the darkness via post-processing. Now, after re-visiting Active D-Lighting, I leave it on Normal most the time, and leave my exposure compensation at zero or -1/3. This accomplishes automatically about the same thing I had been doing manually. Hint: Use 14-bit mode to increase quality of dark areas brought lighter in post. I typically use a combination of increased exposure / brightness, fill, recovery, and tone-curve adjustments, to finish the job in raw as good or better than Nikon software does to JPEG in-camera (ViewNX/CaptureNX can do it out-of-camera).

Note: Processing time is increased before (as firmware calculates under-exposure amount) and after shutter is released (as post-processing effects are applied to JPEG images (even in RAW mode, since RAW files include JPEG preview images). The after-effect is negligible compared to the extra time 14-bit mode takes, and neither lag affects the burst rate / burst recovery time significantly, however, Active D-Lighting should be turned off in situations where shutter lag matters, like sensor-triggered high-speed photography.

*** Capture NX2 allows user to convert to Off, Low, Normal, or High as long as initial setting was not 'Off'. If my findings are correct, I would expect one to be restricted to settings that did not require more underexposure. I now wonder if my findings are correct - its possible Nikon just underexposes by a fixed amount (or not), then adjusts brightness/contrast as per setting, or that there is a penalty involved highlight-wise in trying to go to a higher setting after the fact, or a shadow penalty switching to a lower setting. Hmmm...

Using Flash

As I see it, there are 3 basic kinds of flash usage:

1. Fill: Ambient is mostly enough, flash used to fill in shadows.
2. Balanced: Exposure is balanced between ambient & flash, both providing signficant amounts of exposure.
3. Flash-Only: Flash providing all or almost all of the exposure.

The following discussion, relates to rear-curtain/slow-sync, since thats what I'm using 99% of the time.

• For filling in, theoretically, all you have to do is raise (or mount) the flash. However, its been my experience that most of the time fill is desired, its because there is a strong backlight that would tend to result in an overexposed background, and so most of the time I use P or A mode and down the exposure by 2/3 to 1 stop, and up the flash exposure by 1/3 to 2/3 stops. Reminder to self, and you maybe - external flash / high-speed sync required when ambient is bright, unless stopping way down seems OK.
• For balancing, I typically use manual exposure, and use the meter to determine how much ambient is lacking, and flash to taste. Again, remember external flash / high-speed sync required when ambient bright, unless stopping way down seems OK.
• For flash-Only, I typically use manual mode and set the shutter speed to 250 (max sync speed), and set aperture based on desired depth-of-field and/or for sharpness. If flash not strong enough, open the aperture more and/or up the ISO. If more selective focus desired, open the aperture more and use high-speed sync if necessary (up to 320 shutter speed on-camera, or 8000 using external flash). In general, flash-only is not used when ambient is bright, although it is possible using external flash / high-speed sync. This may be desirable when complete light control is desired and/or far-away background not desirable.

Notes:

• Ambient exposure is defined by strength of ambient, ISO, aperture, and shutter speed.
• Flash exposure is defined by strength of flash, ISO, and aperture, but not shutter speed - since flash duration is short compared to shutter time. Exception: Very high shutter speeds and max'd out flashes, because flash duration goes up as flash output power is increased, and at some point may have longer duration than shutter.

 

Lens Distortion Correction

Typically wide angle zoom lenses produce barrel distortion and telephoto zoom lenses produce pin-cusion distortion. In between focal lengths may produce "moustache" or "handlebar" distortion. At widest and furthest focal lengths distortion can be somewhat corrected manually using a generic barrel/pin-cushion distortion correction algorithm. At mid-ranges, a lens-smart algorithm is required. At all focal lengths a lens-smart algorithm will produce the most accurate results. PTLens can correct lens distortion at any point in the workflow, for any of my lenses. DxO can only correct Raw files, and does not support my 3rd-party lenses, yet (fingers-crossed).

 

Perspective Correction

Vertical & Horizontal corrections can be made by PTLens or DxO at any point in the workflow.

 

Volumetric Anamorphosis Correction

Volumetric anamorphosis is exacerbated by distortion/perspective correction and is impossible to correct perfectly, however with DxO improvement can be had at any point in the workflow.

Post-processing Software

Lightroom

Important Pre-requisite Notes

In fall of 2008 Adobe created color profiles that render colors nearly the same as D300 in-camera or ViewNX/CaptureNX2 would. The internet has tons of pre-camera-profiles information that can be misleading.

Lightroom 2 does not sharpen images down-sized for fit/fill display modes, NX2 and others do. Thus, NX2 images will always look sharper at fit/fill viewing sizes. Always view at 100% when comparing image editors.

Unless otherwise stated, the following information applies to Lightroom 2.6, it may or may not apply to previous versions or subsequent versions.

RAW Conversion (in its purest form)

Gets better all the time. Provided by Adobe Camera Raw, same as Photoshop. Adobe is constantly working on it and updates are provided free and independently of either Photoshop or Lightroom.

Algorithm is solid and produces high quality results that are a bit softer-and-smoother than some but with less detail anomalies than some.

Reminder: To compare raw conversion in its purest form, use Neutral Camera Profile and turn sharpening and noise reduction off.

Noise Reduction

A bit of noise reduction is inherent in the raw conversion algorithm. In addition to that Lightroom provides:

Color Noise Reduction

Very good for low-mid ISOs. High ISO photos should either be deleted or handled with external noise reduction (unless you're going for a noisy effect).

A significant amount of luminance noise is also removed in the process - so much so that I rarely use luminance noise reduction in Lightroom 2. The result has some artifacts that adversely affect the appearance of smooth areas in the mid-ISO shots - Local de-sharpening (strongly) and de-clarify (lightly) reduces these artifacts considerably. Luminance noise reduction also reduces them, but is not applyable locally.

The effect depends more on noise present than position of slider. If the slider is set high enough, Lightroom removes all the color noise it can - higher doesn't help, nor hurt.

Algorithm tends to produce additional softening/smoothing the more color noise there is.

Lightroom 3 (beta) handles color noise reduction better, although its still weak at the highest ISOs.

Luminance Noise Reduction

Its a necessary evil sometimes. It seems to me that Lightroom handles it about as well as others I've tried, which is to say "not as good as I'd like".

It can only be applyed globally and causes smearing of details which may be partially recovered via subsequent sharpening, albeit at the cost of re-introducing some of the luminance noise that was removed in the first place.

I often use local de-sharpening (optionally with some de-clarification) as a substitute for or in conjunction with luminance noise reduction.

Lightroom 3's will be different, hopefully better.

Sharpening

Lightroom's raw conversion algorithm does not sharpen much. Additional sharpening is provided by global (adjustable/maskable) sharpener, and locally applyable sharpening - which uses fixed settings (but can be applied with varying intensity).

Global/Adjustable/Maskable Sharpener

Unmasked - sharpens everything including smooth areas - not recommended except for lowest ISO photos and/or photos without smooth areas.

Amount/Radius/Detail all work together, but here are some general guidelines:

Radius - After having experimented with deconvolution software, I've come to think of radius as very much like PSF factor. Do not make the mistake of assuming the units are pixels, or have anything to do with the radius of unsharp mask in other software. So the rule of thumb is to set the radius lower for photos that are sharper and well focused, and set it higher for photos that are less sharp to begin with.

Detail - Meaning surface texture and edge detail. Crank up when photo is sharp, well-focused, and detail contributes to the desired affect of the photo. Lower when photo is not that sharp/well-focused in the first place, or detail is more distracting than contributory. Sometimesincreases appearance of real detail but always exaggerates noise and other artifacts. Use liberally for things that really have a lot of detail you want to enhance, decrease if photo is noisy, or the finest details are not as important as the less-fine details - the latter of which are enhanced better using Clarity.

Balancing Radius & Detail - In general, the larger the Radius is set, the lower Detail should be set. This is because in areas that have a lot of potential inherent detail, the radius will engulf all potential detail anyway, and besides there really isn't much detail obtainable when its not a very sharp and well focused photo, and trying to eek more detail under these conditions generally just exacerbates noise.

Amount - Sortof like an overall opacity or intensity or strength. Raise enough to sharpen significantly, but do not oversharpen! - Photos in Lightroom should still appear slightly soft!! Then, always use output sharpening that takes into account the presentation medium (e.g. display versus print)!!! One technique is to set the amount overly high, like double what you might expect, e.g. 50-100) whilst adjusting Radius, Detail, & Mask - to better see the affect albeit over-sharpened, then lower to a slightly undersharpened value when done.

Mask liberally to avoid noise exacerbation, or unwanted sharpening, and remember: output sharpening for display uses an unmasked, low-radius, detailed sharpening algorithm.

Another useful techinque is to re-export/import to apply first a high-radius/low-detail "deconvoluting" sortof sharpening first, then overlay with low-radius sharpening to tighten up the finer details/edges.

 

Local Sharpener

Uses a small radius, unmasked, low-threshold/detailed sharpening algorithm. It is not tied to the global sharpener settings.

Photo sharpening may be best achieved by going easy on the global sharpener then supplementing with local sharpening, although sometimes the global sharpening is all you need.

Likewise noise reduction may be best achieved by going easy on the luminance noise reduction then supplementing with negative local sharpening. Or not.

 

Output Sharpening

In two words: Use It. Only reason not to is if you or you're camera over-sharpened already.

Lightroom's output sharpening is excellent and makes the difference between RAW files being too soft and just right.

Output sharpening for display uses an unmasked, low-radius, detailed sharpening algorithm. To compensate, it may be best to use a pinch more luminance noise reduction than you think you might need, and/or a pinch more masking, and/or a pinch less detail.

 

Tone Controls

Exposure - shifts white point. Increasing a lot causes highlight clipping, decreasing will recover highlights but also shifts mids and darks down. If you find yourself wanting to go negative, use recovery instead.

Recovery - the equivalent of Fill Light at the high end. Increasing will shift highlights down mostly, and lights some, but has little affect on mids and even less on darks and shadows. If after recovering the highlights the lights are too low, use tone curve to bring 'em back up.

Fill Light - Increases darkest tones most, lightest tones least. If after recovering shadows the darks are too light, use tone curve to drop 'em back down.

Blacks - shifts black point up. The equivalent of Exposure at the low end, except it won't go negative. Increasing a lot results in shadow clipping, decreasing will recover shadows but also shifts the mids up. If you find yourself wanting to go negative, use fill light instead.

Brightness - Increases mid-tones most, darkest and lightest the least.

Contrast - Decreases lows, increases highs, leaves the mid-tones alone. Increasing contrast can add intensity and mid-tone detail, but decreases intra-dark and intra-light detail.

Tone Curve

Lightroom's controversial tone curve is different. Most others use points where you add and subtract points and move them up down or along the curve. Lightroom uses a fixed 4 zone curve. You can adjust Shadows, Darks, Lights, & Highlights. You can also adjust the partition points between the four zones.

Pros: Fast adjustment of individual tones without affecting other tones much. Course tuning is generally all you need and adjustments can be made very quickly using the four value sliders. Shape can be fine tuned by adjusting the region partitions. The curve is always smooth which minimizes tonal artefacts, and the adjustments can be made very quickly once you get the hang of it.

Cons: "Long windy garden hose" shapes can not be created, although if you find yourself wanting to create complicated shapes you may want to consider whether you're using the right tool. Also, people who have become comfortable with the other kind of tone curve sometimes have a hard time adapting.

I love it.

Presence

Clarity

Hard to imagine life without it - increases local mid-tone contrast, making things look clearer, more present, and detailed. Lack of Clarity slider is CaptureNX2's biggest weakness IMHO - granted, Nik Software's ColorEfexPro3 has a Tonal Contrast tool that fills the gap. Lightrooms version sometimes overexagerrates high contrast edges, like mountain tops against the sky. If this happens, either decrease clarity OR use the local brush with negative clarity to re-claim over-baked edges. Negative clarity also has other benefits - e.g. softening skin tone or reducing presence in subject background...

Vibrance

Increases saturation favoring cool, non-skin colors.

Saturation

Increases saturation of all colors equally.

One can decrease vibrance and increase saturation to bring out warm tones, or decrease saturation and increase vibrance to bring out cool tones. One should get white balance as close as possible first before using saturation/vibrance relationships for color favorings. Likewise, consider individual hue adjustments as well as or instead of.

HSL

To shift hue, saturation, and or lightness on a color-by-color basis.

One thing: Increasing the saturation of reds, for example, only affects things that are predominantly red, in other words, it won't turn neutral grays into reddish grays. i.e. its not an RGB color balance thing - only affects colors that are predominantly the targeted color. If a color is predominantly purplish-blue, yet has some green in it, for example, adjusting the purple sliders will affect it, as will the adjusting the blue sliders, however adjusting the green slider won't affect it.

 

White Balance

Use preset and/or adjust manually using Temperature & Tint sliders. If you're a die-hard RGB white balance fan, you can accomplish the same thing using the RGB sliders under Camera Profiles.

One thing to note:

Temperature -> Warmer => Less Blue (& More Red & Green).
Temperature -> Cooler => More Blue (& Less Red & Green).
Tint -> Magenta => Less Green (& More Red & Blue).
Tint -> Green => More Green (& Less Red & Blue).

You only need two sliders to shift the balance between three colors. The Temperature & Tint Sliders can be thought of as a Blue slider and a Green slider, respectively - to increase just red: increase Tint -> Magenta and increase Temperature -> Warmer.

I was originally a 3-slider RGB man, and it took a while to adapt to the 2-slider Temperature/Tint approach. I think Temperature/Tint is more intuitive, but RGB is more intellectual, so to speak. - I can now use either method without even blinking.

Camera Calibration

Ideally one would use a canned profile that nearest suits the photo/desired effect, then adjust white balance, vibrance/saturation, & HSL to fine tune colors (and/or local adjustment, or split-toning). However, if all the colors are a little off, for a particular photo, regardless of the white balance, the camera profile shadow and/or hue sliders may be the best way to fix it. Likewise, if the colors are consistently off for a particular camera, then a custom profile is the answer.

Lightroom 3 - Final Thoughts

Adobe says they're concentrating on image quality over features Develop-Module-wise. Good. Image quality is number one, everything else (except deal-breakers like reliability and compatibility with existing infrastructure...) is secondary. Although the beta is very rough, there are some noticeable (not dramatic in my opinion) improvements in image quality. If they come through there will certainly be no reason to use any other raw converter just for image quality sake.

Capture NX2

Once upon a time it was purported to produce the best quality RAW conversion. Those days are over. While still good, others have now equalled it or surpassed it - some aspects are objective, others more subjective. Capture NX3 may catapult over the others, but there are no signs that its even being worked on, although I assume that it is. Since the guts are common with the cameras, its possible its release will track with a new "Expeed" processing engine in new model cameras.

With the exception of Split-Toning and Clarity, Capture NX2 can do anything Lightroom can do, and a lot more. Its the single most powerful non-destructive editor so far invented. Bibble exposes layers/masking to the user, whereas CaptureNX2 keeps theirs hidden internally, but CaptureNX2 has U-point auto-masking which no other has.

Raw Conversion

Tends to be sharp, detailed, and with mid-size grainy-ness that hide imperfections. DxO, CaptureOne, and Lightroom 3 are just as sharp & detailed or more, with a finer grainyness that hides imperfections. Lightroom 2 is smoother, with a more subtle, coarser, & more irregular grain hiding imperfections.

Noise Reduction

What most people call Color Noise Reduction is performed during raw conversion - you can't turn it off. CaptureNX2 has a Color Noise Reduction slider that does what most people call Luminance Noise reduction. CaptureNX2 has nothing that it calls Luminance Noise Reduction. Its global version comes with a compensatory sharpening slider, however for optimal results, I prefer to minimize both global noise reduction and compensatory sharpening and apply each locally as I see fit. Don't know if the NR-compensatory sharpening is any different than the sharpening invoked by way of Picture Controls, or if its just more of the same.

Sharpening

1. Whatever may be done as part of RAW conversion (demosaicing/bayer-interpolation) algorithm - not exposed to user.

2. Picture Control (global) - single intensity slider, low radius, detailed.

3. Noise Reduction compensatory (global) - single slider, low radius, detailed - may be same algorithm as Picture Control sharpener - dunno.

4. High-Pass filter (local) - Works great with Overlay blending mode for bringing out edges and not affecting textures/smooth areas.

5. Unsharp Mask (local) - traditional: don't know if all algorithms are the same these days or if one's better than the other. They work well in general. May be used on any channel.

Camera/Lens Specific

Active D-Lighting

Presets for reducing contrast by lightening shadows. Works with camera setting that, when On, underexposes to keep highlights from blowing out. If Active D-Lighting was On in the camera NX2 will increase exposure/brighten image a tad to make up for it, as well as lighten the darkest tones. Although only Nikon supports Active D-Lighting directly, similar results can be achieved in any software, just takes a few more sliders. Still, it goes on the list of "nice things" about NX2.

Shadow/Highlight Protection

These are roughly equivalent to Fill-Light & Recovery in Lightroom, and re-claims near-clipped shadows & highlights respectively. Raw Converters that don't have specific adjustments for this usually rely on sliders along the vertical axis of the tone curve which effectively extend the black point past the left-most limit, or extend the white point past the right-most limit.

Moire Correction

NX2 is the only program I know of that attempts to use Nikon Lens database to correct for Moire. Requires original Raw/NEF file. Not sure how well it does, but once Lightroom 3 comes out, it will probably be the only thing that brings me to use NX2 for Raw Conversion instead of Lightroom. Update @2010-01-21: Just discovered a manual Photoshop technique that probably works better for Moire than NX2's auto - we'll see next Moire correction I have to do.

Vignette Control

Single intensity slider that will be initialized for 100% correction for supported lenses (read: Nikon lenses) and Raw/NEF files. Its an NX2 nicety, but its so easy to correct Vignettes using Lightrooms non-lens-smart vignette correction sliders and/or a brush, its not a big draw.

Image Dust Off

Automatic correction of dust spots if you've had the forsight to shoot a reference photo, which I never have. Its easy enough to sync dust spot corrections in Lightroom, so again - not a big draw.

Auto Lateral Color Aberration Correction

i.e. Red/Cyan and/or Blue/Yellow chromatic aberration. Although the automatic correction in NX2 is enabled by default and works well and thus saves the step of having to look for chromatic aberration to correct, I'm not sure it works better than the manual correction in Lightroom, and hence not a big draw.

Axial Color Aberration Correction

a.k.a. Purple fringing. Haven't tried this one yet. Lightroom 2's fringe correction was pretty weak. Dunno if it works better when NEF/Nikon-Lens, but it is available for all. Might try this one if Lightroom 3 doesn't improve this feature.

Auto Red-Eye

Dunno if this is better than Manual Red-Eye, will need to investigate...

Fisheye Distortion Correction

PTLens does a great job at any point in the post-processing workflow, so not a big draw, althrough lately I've been disappointed that volumetric anamorphosis is missing which leads to even more edge/corner-stretch distortion. DxO will do it just as well although only for original Raw/NEF files, and will be required if volumetric anamorphosis correction control is desired to go with that fisheye correction - e.g. to keep things from being too fat at the sides. Actually, I've noticed NX2 does do some volumetric anamorphosis correction that's not too bad - they just don't make a big deal about it, nor allow any control over it.

Camera/Lens Support - Final Thoughts

Some things you can do in Lightroom, albeit manually, some you can not. Of the ones you can not, they can be done in PTLens and/or DxO Optics. The only one that requires Raw Conversion be done in NX2 is Automatic Moire Correction, but I may just learn to do this manually, or get a helper plugin for Photoshop, in which case there will no longer be any reason to use NX2 for Moire correction. If it turns out NX2's auto moire correction does a stellar job, it will remain the only thing that will force me to do Raw conversion in NX2. There will be a dilemma if I need to do corrections in DxO that requires Raw, plus Moire corrections in NX2 that requires Raw - I'll have to pick one or the other since its hard to have both.

CaptureNX2 as general purpose editor

I try to do as much as possible in Lightroom and not resort to external editing unless I have to. Its the most convenient and minimizes the complexity and storage of additional edit files. There may be times when I have to start with NX2 (Moire), but otherwise I use it to finish files that I can't quite get the results I want in Lightroom, or it would require more time. When this happens, the first question I ask myself is: NX2 or Photoshop/Plugin. NX2 has some features that Photoshop does not, and in general its just less tedious to work in a layer-less environment if layers are not adding any real value. So, NX2 is my first choice if layers or plugins are not needed. Here's why:

U-Points

Sometimes its just dam near orgastic to be able to plop down a couple selection points and tie a few adjustments to them to do things that would take a lot longer to do any other way, and may still not work as well. This is the biggest strength of NX2 as a general purpose editor, i.e. raw-conversion quality aside. Nik Software's Photoshop plugins don't even come close to providing the same functionality as NX2 because u-point selections are accessible only to the plugin where they are defined - you can't link them to other plugins or to any other Photoshop adjustments. In NX2, all adjustments & filters can be linked to a common selection - this is a really good thing that I use "all" the time when editing in NX2.

Distraction Removal Brush (Auto Retouch Brush)

Lightroom removes dust spots, NX2 removes tree branches - need I say more?

Other things

There are a number of tools that go beyond what's possible in Lightroom. My intention in writing this is mostly to remind myself when and why to go outside Lightroom and use NX2. To that end I simply list those I can think of here:

• Curves can be defined on a per channel basis - I rarely do this, but it occasionally helps to remove color casts that vary between darks and lights, or to add them for effect.
• Black, White, and Multiple Neutral Control Points. Another way of setting RGB tone levels, and white balance that can vary from region to region.
• Linear gradient whose graduation can vary non-linearly.
• Radial Gradient (circular) - good for adjusting point light source behavior and such.
• High pass filter - another edge sharpening tool that behaves differently than unsharp mask. Has one distinct advantage that you can view exactly which edges will be affected and by how, as a gray-scale overlay.
• Gaussian Blur - I'm normally trying to make things clearer, but sometimes a blur is just whats needed instead.
• HSL adjustments can be targeted more precisely.

Capture NX2 - Final Thoughts

Has camera/lens specific things that are a big reason some people use it instead of a generic raw converter / destructive editor. All of them can be corrected manually nearly as well or better using other tools so these things are not reason enough for me to use NX2 as my primary raw converter / editor. There for a while I was tempted to switch from Lightroom to NX2 for raw conversion quality, but I think Lightroom 3 will be as good or better, once its finished. Even so, its U-point technology and selection tools, auto-retouch brush, and other tools keep me coming back after raw conversion in Lightroom. Biggest deal is that all NX2 adjustments and filters can be applied locally via U-point selection or any other selection method. Tone curves applied locally allow for "Topaz Adjust / Lucis Art" - like effects or for increased clarity in regions that need it.

 

Photoshop

Disclaimer: I'm still a bit intimidated by Photoshop. Its like driving in San Francisco: I may get where I'm going, but I'm so turned around by the time I do that I have just as much difficulty the next time. Like all 3rd-party image processing applications, the only way to edit it from Lightroom is to export a TIF (or PSD) file to it, edit it, then save it (actually Photoshop can read raw files and interpret Lightroom edit-list, but it can't write same and so writes a big tif instead back to Lightroom). I've used it in the past mostly for layer/masked work, and panorama stitching. Because of my personal issues with it, and the fact that it requires a huge intermediate file, I use it only rarely - generally preceded by kicking and screaming. After recent playing (2009-10), it seems noise reduction is as good as Noise Ninja, and even more flexible, since it may be applied on masked layers. Sharpening also may be applied on masked layers. Noise Ninja makes one think its doing a better job because it has camera-specific profiles, but I'm not sure these really help. I mean ultimately the software has to find and fix the noise, which varies quite a bit from photo to photo, on the same camera, even at the same settings. I just read that Dfine noise reduction software just eliminated camera profiles - I guess they agree with me.

 

DxO Optics Pro

 

Update @2010-01-26: DxO does not require RAW/NEF file for lens module support, it merely requires EXIF data that comes from Nikon - either the camera OR Capture NX2. Thus one can save a jpeg file in NX2 and still have access to lens modules in DxO. Care must be taken to not correct things in DxO that have already been corrected in NX2 or else they will be over-corrected.

 

History

In the beginning, I used only Lightroom for photo post-processing. Then, I encountered its limitations and opened the door to external image editing, e.g. Photoshop. Once that door was opened, no reason not to do RAW conversion in NX2 if its going to be used for its advantages in post-processing. Recently, I shot a set of party photos using an ultra-wide lens and ran into the limitations of PTLens - distortion correction can result in exacerbation of stretching that occurs toward the edges. Enter DxO Optics...

DxO Optics In A Nutshell

DxO is primarily a RAW converter and global image editor. Its default conversion renders hues almost exactly like Camera Standard profile ala Nikon and/or Lightroom, albeit more saturated and contrasty. Intra-DxO it can be used non-destructively (edit-list contained in sidecar file), however, like all editors, its non-destructive edit-list is proprietary and can not be interpreted in Lightroom, thus edits must be folded in to the image data and delivered to Lightroom ala dng/tif/jpg in order to be effective there.

DxO Optics Strengths

Generally the quality is good and features for dynamic contrast and detail enhancement are good, and also there are some specific features that are hard-to-impossible to replicate in Lightroom:

• Lens Softness Correction
• Distortion Correction
• ?Auto Chromatic Aberration Correction?
• Vignette Correction

Lens Softness Correction

This is true "capture sharpening". Lightroom has no equivalent, and relies on "creative sharpening" and output sharpening techniques to try to make up for it. Although the result is simply a sharper photo, the algorithm is very different and effective and produces great results automatically. It uses an adaptive convolution technique that is applied during raw conversion that brings the picture into better focus without exacerbating noise. More like focus-magic than edge sharpening. It works really well. Results can be approximated in Lightroom by cranking the edge sharpening up and masking thoroughly and/or desharpening some areas and adding local sharpening to some areas, but its not as good and you really have to work for it.

Auto Chromatic Aberration & Fringe Correction

Initially I thought DxO's Chromatic Aberration correction worked better than Lightroom's, and perhaps on the photo I was working on at the time, it did. But, certainly on the photo I'm working on now Lightroom's correction is better, and so at the moment I'm actually favoring Lightroom's manual corrector over DxO's semi-automatic/lens-module-based correction. Fringe-wise, I think DxO's is better than Lightrooms - jury still out.

Distortion Correction

DxO Optics is the reigning king of low-to-mid-priced distortion correctors. PTLens can do the same thing barrel-pincushion/perspective-wise, but no other program I know of allows parametric control of volumetric anamorphosis correction. Its not perfect, and sometimes I can do a better job manually warping in Photoshop, but often it can make a wide-angle shot look more like a panorama by just checking a box (and maybe dragging a slider or two).

Vignette Correction

The camera/lens-smart vignette correction works perfectly. Its not as big a deal as some of the other things since you can deal almost as effectively with vignettes manually in lightroom too, but its a nice DxO touch.

 

DxO Optics Limitations

• No Local adjustments
• Inferior results when used after raw conversion.
• AdobeRGB is widest gamut colorspace supported.

No Local Adjustments

No gradient, no brush... - No problem. DxO Optics was not designed to be a complete solution, but a raw converter and corrector. Most people use another program for photo management (e.g. keywording), local adjustments, presentation/packaging, ..., if desired.

Inferior Results when used after raw conversion

Even if its working as intended, it loses some of its best features when its not converting the raw for a known camera / lens combo. And I was unable to get proper tone & colors roundtripping from Lightroom via tif intermediary with any color profile other than sRGB. It works fine with a jpeg intermediary and AdobeRGB, but its unsatisfying to have to resort to sRGB or JPEG, and there's no reason to if you use DxO in the beginning - the problem comes when you have the photo nearly perfected in Lightroom but later you realize you want to correct lens distortions - your choices are to start over in DxO, use tif/sRGB, use jpeg/AdobeRGB, or edit the raw in DxO applying minimal edits that overlap whats already been done in Lightroom, then re-apply Lightroom develop settings strategically upon return. Note: local brush strokes will be misaligned if geometric corrections have been made.

AdobeRGB is widest gamut colorspace supported

DxO claims ProPhoto RGB is overkill. It may be overkill as a final output colorspace, this year, but internally whilst editing I think the wider the gamut and the finer the resolution the merrier. Every other raw conversion company agrees with me. DxO may be right, but there are some areas where you need to keep up with the "Joneses" even if the Jones are wrong. My prediction: DxO Optics v7 will support ProPhoto RGB, or DxO Optics will go under.

DxO in a Lightroom-Centric Workflow Update @2010-01-18

After further experimentation, I'm seriously considering standardizing the way I use DxO as follows:

• Only in cases where DxO's special correction features are desired, and after importing into Lightroom, open the the RAW/NEF file directly in DxO Optics (circumvent all the "help" in the form of plugin mode / external editor interface).
• Use Lens Softness correction down-throttled by .5, if lens module supports it. This uses a lens specific adaptive algorithm that produces great RAW pre-sharpening - the default over-sharpens for this stage of the game.
• Use color noise reduction as appropriate. This must not be postponed until back in Lightroom, otherwise there may be uncorrectable artefacts - always check "remove dead pixels".
• Do not do any luminance noise reduction. I really think its best to do luminance noise reduction on a strategic basis, with different settings for different areas - not globally.
• Use Neutral Color Rendering - this makes sure DxO renders colors as accurately as possible - since a DNG file will be used later, RAW Camera Profiles will be selectable in Lightroom.
• Do not change White Balance - DNG file will allow camera white balance settings to be selected in Lightroom, and adjusted from a common baseline.
• Perform distortion/perspective/volumetric-anamorphosis corrections as desired.
• Jury not in regarding chromatic aberration - try DxO & Lightroom and see which is best.
• Do purple fringe correction - unless Lightroom 3 greatly improves this feature.
• Recover highlights and shadows as necessary using the levels settings of the tone curve, and or fan the data out a bit. For some reason, its impossible to do true highlight/shadow recovery on a cooked DNG like it is on a raw NEF in Lightroom. The reason for fanning out a bit is because for some reason the data seems to get compressed a bit by the time it makes it to Lightroom, plus since DxO is a bit "wide-gamut colorspace challenged" its probably good to have the data fanned out a bit. This may take a few iterations to get right.
• Resist the temptation to make any other changes in DxO.
• Save as DNG (not TIF). The colorspace will be AdobeRGB since thats the best DxO has to offer.
• Import into Lightroom and rememember to turn color noise reduction off. If auto-lens-correction was used in DxO, turn sharpening off too, as a starting point.

Doing it this way will make sure that the baseline for each photo is as close as possible to what the original RAW would be, save for special corrections, and maximizes the editing in Lightroom. This produces excellent results and helps maintain sanity (consistency) in whats already a complicated and sometimes confusing workflow.

DxO Optics - Final Thoughts

I dont anticipate using DxO often, because of the complications and increased storage requirements of going outside Lightroom in a Lightroom-centric workflow. Its most frequent use will no doubt be as a comparison tool to see how things look when done with a different tool, which may give me ideas for further improvement in Lightroom. Its most valuable asset will probably be to make corrections that are nearly impossible any other way. Another occasion may be when I use it and get great results that would be too hard or time consuming to try and replicate in Lightroom.

Things requiring raw are:

• Distortion correction, which PTLens can do at any time, albeit sans volumetric anamorphosis correction.
• Vignette correction, which only DxO can do, but can be approximated in Lightroom using manual vignette correction combined with local adjustments, and possibly done better, since there is no guarantee that I want the same amount of correction in each corner.
• Auto lens softness, which only DxO can do, but can be approximated in Lightroom.
• Chromatic Aberration auto-correction, which can be done manually in Lightroom, with results that may or may not be better than what DxO does.

So, although there is potential for better results than Lightroom, there is very little to draw me into using DxO for more than a test comparison, unless I just can't get the chromatic aberration right using Lightroom but DxO is doing the job, or its the kind of photo where auto-lens-softness is being hard to match, or there is photo content that is just not responding to manual vignette correction in Lightroom, well enough. The DxO Lighting and such can generally be handled using Lightroom and/or Topaz Adjust, ..., but again, once I'm going outside Lightroom, whether I use DxO for raw conversion or not doesn't much matter, since its really the extra files/storage and destructive editing that I'm trying to avoid.

 

PTLens

Does as good a job as DxO Optics at distortion and perspective corrections, and supports more lenses, but lacks volumetric anamorphosis correction which is often desirable to go with distortion/perspective correction. Truth be told however, I think manual warping in Photoshop may produce results better than either PTLens or DxO, albeit a bit more work. Problem is you're correcting in 2D distortions that depend on the original placement of objects in 3D so its always a compromise. Using manual warping you at least have some control over which objects are improved and which are worsened. Update @2009-01-26: CaptureNX2's distortion correction includes volumetric anamorphosis correction auto-magically - imperfect, not adjustable, only on NEFs, sometimes you can get better results by Twiddling in DxO or Photoshop, but its the easiest way to get the job done when it works.

 

Noise Ninja

Love it and hate it. No significant advantage at lower ISOs. Can get extreme noise out better than Lightroom, but requires a lot of time tweaking the sliders for optimal results - default settings are way too heavy for my taste. Truth is I generally prefer Capture NX2 for high-ISO shots because I can tie different levels of noise reduction to different sets of auto-masking selection control points. DxO does a phenomenal job at getting noise out whilst retaining detail, but its not maskable (nor is noise reduction maskable in Lightroom, although sharpening and de-sharpening can be applied locally). Yep, noise ninja has good noise reduction algorithm, and supports masking, but the tools for masking and strategic application and compensatory sharpening are weak. Its entirely possible that the best noise reduction could be achieved by using noise ninja at different settings for different layers, with masking courtesy of photoshop/plugins, but this sounds a lot like work to lazy me ;-}

 

Fluid Mask

Considered one of the best extraction masking tools. My experience - hard to use, but great results for hard-to-mask things like hair and smoke. Its edge detection seems to miss too much, but once the keep & delete regions are tightened up it does a super job at blending in the "gray" areas. Can be used for general masking for selective image enhancement, although its fortay is extraction masking.

 

Topaz Suite

Reminds me of Nero Suite - Does lots of things well without breaking the bank, but programs are divided by algorithm used rather than by what they do. For example, there's separate "Clean" and "Simplify" programs that do similar things but differently. And there's "Adjust" & "Detail" that have greatly overlapping functionality - hard to know which to use. They lean more toward special effects than natural photo enhancement, but if applied gently can achieve astounding results au-naturale as well. If the various adjustments could all be tied to a U-point or two it would be perfect.

Topaz Detail

Can emulate Lightroom Clarity, or Nik Software's Tonal Contrast Filter in Color Efex Pro, or Capture One's Clarity, Viveza's Structure, ..., or not. Has more sliders for adjusting stuff and the learning curve is steep, but the quality of the results are second to none. Built-in presets are very heavy handed, and output image is recomputed slowly after each and every slider adjustment. I have developed my own presets and Photoshop actions for using them to make things go quicker. I recommend doing some noise reduction in Topaz Adjust and/or Topaz Detail afterward if small details are enhanced, since this also exacerbates noise. These two de-noisers seem to be very good at not reversing the small detail enhancement entirely whilst still finding the noise. If you use your own de-noiser I recommend downthrottling in the regions where fine detail need be preserved.

Topaz Adjust

Oftened likened to Lucis Art. My experience - much more powerful than Lucis Art. Lucis Art avoids small detail enhancement and works more in the exposure domain, desaturating to avoid funny color business. Topaz Adjust has adjustments for exposure, color, tone, and detail. At the moment, I use Topaz Detail first for detail enhancement and leave the detail adjustments alone in Topaz Adjust, then use some noise reduction in Topaz Adjust to counter noise exacerbated in Topaz Detail. I finish with Topaz Denoise, although now that I think about it, maybe I should de-noise first :-}

Topaz DeNoise

Best luminance de-noiser I've used so far, and I've used more than a half dozen of the biggies. In usual Topaz Labs style, there are lots of sliders to adjust how it works, but I like that it just analyzes the image to determine how to de-noise, rather than following a canned profile. That makes it useful before, during, or after cropping, or the use of other noise reduction tools previously, distortion corrections, ... Also, the sliders make sense to me and I can see the effect of changing them. The de-blur slider works like focus-magic or auto-lens-softness in DxO, only better - has a clarifying effect on slightly blurry objects thats amazing, yet doesn't affect bokeh or sky so much - which is good. Beware of color desaturation - see the 'Advanced' tab and down-throttle color noise reduction if necessary. Slooooow - and doesn't directly support Lightroom plugins or batching, but does support presets and can be automated via Photoshop - I used Actions and/or droplets to apply en-batch.

 

onOne Suite

A combination of tools for both natural photo enhancement and special effect. I must say that the better I get using lower level tools the less support I need from a plugin suite. Still, they can be invaluable in the mean time, or when you are just confused, or just need to get some creative juice flowing.

 

DynamicPhotoHDR (by MediaChance)

I really wanted to prefer Photomatix, but I don't. What can I say? - just as good or better results with an easier to use and more aesthetically pleasing UI. I think the folks most into HDR still prefer Photomatix, but I'm certainly not in that crowd - at least not yet.

 

DynamicAutoPainter (MediaChance)

Gets my vote for funnest plugin to use, since it starts out very coarse and refines as it runs. You can paint the placest you want to be the most realistic/detailed as it runs and stop it when its got the level of detail/refinement that you want. I've used it on photos that would otherwise be thrown away because they're too blurry or just not very good and produced some very pleasing artwork. I also tried Photoshop painting filters and some other plugins (AutoFX is only one I can think of by name), but I like the way DAP results look and its just plain fun to use.

 

AKVIS Sketch

I tried Photoshop's sketch filters, a couple other sketch programs and plugins whose names are escaping me, as well as do-it-yourself Photoshop procedures. AKVIS is just too good and too easy and sketching is too infrequently used to be worth messing with anything else. It has the nice aspect that it tends to keep the shadows dark instead of lightening everything like some sketch programs. Also, there is a sloppiness to the strokes that make it look a little less "computer-anim" than some. Sketch effects can be applied strategically within AKVIS Sketch, but probably would be just as easy to do using Photoshop layers and traditional masking techiniques.

 

ColorEfexPro3 (by Nik Software)

I troubled whether to get the suite, or the version that supports stand-alone/Photoshop plugin, and ultimately decided just to get the NX2 hosted version - I don't regret it. NX2 is the only host that lets you tie the effects to a set of control points along with other effects from NX2.

If you use Bridge/Photoshop for all your photos instead of Lightroom, I'd recommend getting the complete suite for Photoshop as plugins, and use Dfine for strategic noise reduction, Viveza to do what comes built-in with NX2, ColorEfexPro for more special enhancements, SilverEfex for B&W, and Sharpener for sharpening.

But Lightroom already has great noise reduction and creative/output sharpening built in, and black & whites can be accomplished about as well using Lightroom presets, tone curve and other normal editing tools (or NX2), and with NX2/ColorEfex that work well as external editor, there just isn't as much value in having the whole suite in a Lightroom-centric workflow.

 

PhaseOne's CaptureOne

I used the trial for a while. My sense is that it would be more of a Lightroom substitute than a DxO substitute, but I'm committed to Lightroom for its export plugins, and there is nothing CaptureOne has to offer that isn't covered by Lightroom & DxO, except for tethered shooting, which I don't do now. If I want to shoot tethered I imagine I'd just use Nikon's CameraControl, although I'd check out CaptureOne tethered support before laying down the cash. I guess its about as good RAW-converter-wise as Lightroom(ACR) and/or DxO but not better, so no real reason to use it me-thinks, except for the aforementioned or if you own a PhaseOne DigitalBack and lenses to go with it. Oh yeah, I just rememembered a deal-breaker for me: The big print giveth lens specific corrections but the small print taketh away some features, like chromatic aberration auto-correction, lens sharpness, distortion correction. & light fall-off(vignette)) for all but PhaseOne & Mamiya lenses. I can't help but wonder if its a marketing thing to get the most demanding DSLR tweakers to move to a PhaseOne camera & lenses - too cynical? I mean the thing cost 3 times what DxO Optics cost which supports all corrections for most primary brand DSLR lenses (Nikon, Canon, ...) and a few Tokinas, Tamrons, & Sigmas. I mean if the image quality were truely superior to CaptureNX2, DxO, & Lightroom, I'd use it in a heartbeat, but it ain't. Its got some nice features but nothing you can't do with the others, except tethered shooting, which is covered by Nikon CameraControl for a third the price. - Just can't see ever buying CaptureOne unless they change their policy to fully support Nikon shooters/lenses and gain the edge image-quality-wise. Update: After more experience with trial I see some plusses and minuses. But first some rumination: I think one of the challenges of building a raw converter is dealing with imperfect sensor data. I just compared a relatively low ISO shot with disproportionately high luminance noise - Lightroom's conversion was sortof smeary smoothy not unlike a high luminance noise reduction setting, except without any luminance noise reduction applied. CaptureOne produced more detail but included a lot of strange speckles perhaps like random film grain that no amount of noise reduction would remove - clearly considered image data but not really correct. Also, CaptureOne seems to be more conservative/accurate/natural at color assignment. Anyway it just brought home my sense of a fundamental difference between the two, that Lightrooms conversions may not always have the most detail, but generally steer clear of speckles and other artifacts. Likewise, if CaptureOne input is very good, the output tends to be very good and natural. I think I may see why some say CaptureOne is more oriented toward studio/portrait shooters - excellent natural detailed renditions when input is highest quality. If they supported Nikon lenses I would buy it for the exceptional cases when I really wanted the CaptureOne's result, but I'm afraid that is a deal breaker for me, at least at $400. If they sold the Pro version that discounted for lens absence I would buy it, but at $400 I need my lenses to be supported. Oh - one other thing: Unlike DxO & Lightroom, C1's color noise reduction destroys image colors if set too high - its a must to keep it to a minimum. Final note: Like all imaging apps these days, consistent color interpretation between apps (read: how to properly handle color profiles) is problematic. Most C1 output did not produce consistent color when viewed in Lightroom. This combo did: C1 output ala TIF-16, ProPhoto RGB, No Sharpening. Another pisser: C1-5 stores files where it wants to making it very difficult to integrate with existing file storing conventions, and file re-naming is very limited. Note: To have output files directed to same folder as source image, see 'Advanced' tab in Process Recipe tool. If you select 'Image Folder' as output it ignores everything else you've set and despite what its showing you it will do, it outputs the processed image to source image folder instead. Another con: C1-5 does not include check boxes or such to disable a setting to see its effect. You can temporarily reset it, but you don't get to see what those "reset" values will be and the reset setting is the default, not zeros. Actually, it may be possible to set the defaults to zero, but then you wouldn't get the automatic defaults applied, I don't think. On the plus side, its got a better HSL adjuster than DxO or Lightroom, but not as flexible as NX2 although slightly easier to use. There's a part of me that wants CaptureOne in my toolkit, but I've just encountered another peeve - they store all the big preview images in with the sources, instead of in a temp folder - grrrrrrr.. I can live with it, but I shouldn't have to... Yet another not-up-to-snuff thing: I just noticed Chromatic Aberration over correcting, and no way to down-throttle it. DxO Optics corrects it better since its a lens personality thing that CaptureOne hasn't addressed for DSLRs. Wow: CaptureOne has no way to clean up dead pixels left from noise reduction in enhanced shadow areas, unlike Lightroom that just always does it, and DxO that gives the option (unchecked by default) which I always check. Clarity in C1-5 reaches further into the contrast domain than Lightrooms, sort-of like LucisArt, and although this sometimes looks awesome, you can't clarify just the finer details without adding medium-range contrast. I prefer Lightroom's Clarity sticking to the finer details - I can always brush in some Contrast locally if desired.

Capture One Conclusion

I'm beginning to see why people say "...targeted toward studio portraiture..." - its noise reduction is not that good and extreme shadow recovery is not handled well enough. However, if you start with a low-ish ISO image thats properly exposed, it renders the details nice, sharp, and smooth and colors nicely balanced, albeit not particularly vibrant. The color editor allows fine tuning to a level of detail rivaling NX2 - better than Lightroom and way better than DxO. The one thing CaptureOne has that none of the others have is a skin tone optimizer - too bad it effects the skin tone of tree bark as well as people (not locally applyable). For me, there's so little to offer that isn't already covered by Lightroom/NX2/DxO/PTLens/Photoshop+plugins that I really shouldn't buy it. On the other hand, it would be valuable for special images to render them using CaptureOne to get ideas for how to improve them in Lightroom (as example), and if I really can't replicate what I like from CaptureOne - just use the image from CaptureOne instead, or pull it into Photoshop as a layer and mix it with aspects of images rendered using other tools. I'm going to wait until Lightroom 3 is released - if it truly solves some of its present weaknesses, I'll forget all about CaptureOne; if not - I'll probably swing for CaptureOne to add to my toolkit on the quest for the best... Its biggest weakness is leaving "dead/hot-pixel salt" scattered around - it disappears if photo is shrunken for screen, but gets larger than life when cropping or printing - a total deal breaker for me.

 

ACDSee Pro

I used this before Lightroom - before RAW. I was very happy with it, but it seems not too many people are using it for raw editor - not sure why. Again, I'm commited to a Lightroom-centric workflow because I've invested tons in Lightroom compatible export plugins, so I won't be switching anytime soon. Still, I gave it a check recently and it seems quite good. I can only guess there are quality issues that would be observed after a while, otherwise people using it would be constantly saying "ACDSee is just as good a raw converter as Lightroom for half the price" in the forums and what have you - not seeing that. On the other hand, I also don't see people giving reasons for not using it- hmmm. I still use it as my primary jpeg viewer, especially for slideshowing my exported jpegs, since Lightroom aborts a slideshow as soon as I touch anything on my keyboard, and I want it to keep going while I do other stuff with my computer.

 

File-handling in a Lightroom-centric workflow

Pre-requisite Info

About 3rd-party Lightroom integration claims: They are mostly misleading. Lightroom exposes none of its image data to 3rd party applications. 3rd-party plugins may be written to access catalog metadata and have access to a photo file before or after it is rendered, but otherwise there is no access to image data, period. Most image processing applications professing Lightroom integration are just ordinary stand-alone image processing applications with redundent instructions for configuring in Lightroom as an external editor. Its possible they also come with a support plugin, or some very loose Lightroom support, but unlike in Photoshop, none of them are able to manipulate Lightroom image data without Lightroom rolling up a tif file for them. Even Photoshop itself can not read Lightroom edit history, nor can Lightroom read Photoshop edit history - Photoshop is like any other 3rd party external editor to Lightroom, and Photoshop hardly knows about Lightroom's existence, just knows how to get files back and forth.

I import into lightroom, edit, tag, then export for use outside lightroom. In the simplest case, I have the original nef, original jpg (optional), and edit list stored in both Lightroom database and xmp sidecar file. If I decide to use NX2 for raw conversion instead of, or as well as Lightroom, I will have the NX2 edit list added to the original nef, and the previews replaced in the original nef, together with the tif or jpg file saved in NX2 for lightroom. I may use an other 3rd party app for subsequently editing the tif or jpg file saved by NX2, or I may even return to NX2 to edit the tif or jpg, leaving the nef as it was after initial edit. Its also possible I would export the tif again with incremental lightroom edits, creating a _X_X.tif file. Upon return, I would probably delete the source tif and rename the edited tif to maintain consistency with file-naming convention. Likewise, I may decide to use a 3rd party app for editing a tif or jpg file representing the Lightroom raw conversion, or even NX2. NX2 will only save edit list in a nef file - saving as tif or jpg is destructive (edits are folded into image data and edit history is discarded), but Lightroom can tell the difference between raw and cooked nef image. It still can not read NX2 edit list.

Filenaming conventions:

{CameraID}_{ImageID}[-VersionID].{FileType}:

• D300 Camera ID: NIK3
• Image ID: Unique ever-increasing image serial number, minimum of 5 digits, 6 if necessary. D300 actually uses a 4 digit image number and a 3 digit folder number which are combined to create the image id.
• Version ID:
  • A single digit number if more than one physical version has been created.
• File Types:
  • NEF: Original RAW from camera - possibly updated with NX2 edit-list and jpeg preview, or side-car file to a tif master used for storing nx2 edit-list - lightroom can not see changes made to tif and saved as nef, but one can re-save the tif with NX2 edits AND the nef sidecar for editing. It is recommended to delete the extra nef file once photo is finished. Understanding of course that it could always undergo another round of changes, just won't have benefit of edit list.
  • TIF: Conceivably could be an original TIF from camera but I don't imagine ever doing that. Hence it indicates a destructively edited copy, losslessly compressed.
  • JPG: Destructively edited copy compressed (lossily). If it comes from a destructively edited tif, the corresponding tif should be deleted.
  • PSD: Not Used - seems Photoshop can store the same stuff in a tif using less space, no?

Filename Examples:

• NIK3_12345.NEF - original or NX2-converted-and-edited raw file.
• _NIK3_12345.nef - edit-list (sidecar) for non-NX2-converted but subsequently edited non-raw file.
• NIK3_12345.JPG (upper case) - the original un-edited jpeg file - out of camera, as a sidecar - no longer used.
• NIK3_12345.jpg (lower case) - version edited externally.
• NIK3_12345.TIF (upper case) - reserved for original un-edited tiff file - out of camera - never used.
• NIK3_12345.tif (lower case) - version edited externally.
• NIK3_12345-2.tif - another version edited externally.
• NIK3_12345-3.jpg - another version edited externally.
• _NIK3_12345-2.nef - edit-list/sidecar for another version, non-NX2-converted, subsequently edited with NX2.

 

I have defined Custom Metadata in Lightroom to keep track of Edit Notes - especially if a non-lightroom raw converter was used and if any destructive edits have been done outside Lightroom. In case I want to re-edit in the future, or just to have an idea what's been done.

Procedure for Raw Post-Processing

1. Always import into Lightroom first. (When ViewNX was working, I would sometimes cull straight from the card first).
   1. Standard filenaming convention.
   2. No metadata presets nor develop presets - This way, I never run the risk of having them set wrong, and never have to check 'em. I use folder-sync to import quite frequently too, and this way I don't have to worry about presets or even bother with the import dialog when doing it.
2. Apply metadata presets, always.
3. Consider whether this photo warrants use of NX2 or DxO instead of Lightroom for raw conversion.
4. If Lightroom for raw conversion:
   1. Consider applying developer presets.
5. If NX2 for raw conversion:
   1. Try and do minimal initial edits then save TIF for lightroom.
   2. Consider saving edit-list in original NEF.
6. If DxO for raw conversion:
   1. Try and do minimal initial edits then save DNG for lightroom - so white balance presets and color profiles are available to Lightroom.
7. Edit maximally in Lightroom.
8. Consider finishing externally, usually as tif(pro-photo-rgb), or jpg(adobe-rgb) for DxO.
9. If external editing done, save snapshot in lightroom, in case nef dinked with later in lightroom - this way the settings as input for external editing are saved, to aid reverse engineering of destructive edits.
   1. Until committed to finishing, consider saving state in editor of choice, for example:
      1. Save cooked nef in NX2.
      2. Save pdf/tif with layers in Photoshop.
      3. Or, just leave the apps open with photo loaded.
10. Minimize edits to tif intermediates in Lightroom, since they will be lost upon jpg master conversion. Either do them before hand with lightroom/external editor, or to jpg master after conversion.
11. Convert tif intermediate to jpg master.
12. Consider last-minute touch-ups to jpg master.
13. Delete intermediate files. (once upon a time I was extremely reluctant to do anything outside lightroom, mostly because of huge intermediates. since I've begun deleting intermediates I feel much freer to edit externally. Saving snapshot before external editing, and keeping notes after external editing allows for some reverse engineering, if need be, or curiousity strikes. An advantage of jpg master format is subsequent exporting is much faster).

 

Raw Conversion & Workflow - Final Thoughts

When I first heard about Lightroom, Lightroom 2 was in Beta. At the time, the general consensus was that ACR was inferior to NX2 & CaptureOne (not sure where DxO stood or the others). Even so it was picking up steam in a big way and version 2 promised improvement in image quality. After trying Lightroom 2 I became convinced that if it wasn't quite up to snuff yet, it would be soon - and was here to stay. It was the export plugin API that sealed the deal for me, since I use a proprietary Flash App to display my photos and I wanted to be able to format my photos for the web straight from Lightroom. Since that time I've become quite adept at using Lightroom and have learned to enjoy it very much. On the flip side, I've also encountered its weaknesses and limitations. In the quest for the best image quality I've become experienced with NX2, DxO, both of which I now own, & CaptureOne (which I'm hoping to resist buying), not to mention Photoshop and several plugins. All of those products have their strengths & weaknesses, but with the release of Lightroom 3, I suspect I will come full-circle - using it exclusively in all cases except where a u-point, a layer, extensive local work, geometric corrections, or a plugin seems necessary. Regarding evaluation of raw converters, image-quality-wise, you have to be adept enough to obtain almost exactly equal images tone-wise, color-wise, and detail-wise first - only then can you see the true non-correctable differences between them.

 

High-End Photo Editors

I wonder what the people who create movies for George Lucas or James Cameron use for personal photo editing? Anybody know??

 

Canon versus Nikon versus Others

My take is that Nikon & Canon DSLR cameras/lenses are very similar. They are fierce competitors for the same market, and one will do a little innovative leap frog here, then a little catch-up there. One Nikon weakness I see (@6/2009) is the lack of a good-enough mid-priced 400/500mm lens. Canon has a very nice 100-400 with IS, but Nikon's 80-400 VR, although sharp, has very slow autofocus, rendering it less useful for wildlife & sports. If I were buying today, this would be the main reason to go Canon, although the Sigma 150-500 fills the gap pretty nicely. My prediction: Nikon will plug this hole within the next year or two. Nikon's bodies, on the other hand, are generally considered (at the time of this writing) to be slightly better than Canon's, and is why I went with Nikon (D300). However, in another year or two, or at the time of this reading, Canon bodies may be considered better... Regarding Sony, Olympus, Fuji, Pentax, Panasonic, Sigma, Ricoh, Leica, ... My understanding is that their DSLR cameras are good too, however they dont have the same level of support via 3rd party hardware & software, and in the community - more risky.

 

References

• Nikon AutoFocus System Technology
• Diffraction & Photography
• Digital Camera Sensor Sizes
• Norman Koren's MTF (resolution) Tutorial
• Ken Rockwell - A lot of good information, mixed with a lot of over-simplification and consequent seeming contradiction, & opinion - you have to learn how to read Ken Rockwell, but its worth the effort if you're hungry for the truth.
• DOFMaster - Aperture and hyperfocal distance calculators.

 

Final Thoughts

I have composed this document to help myself remember what settings to set, to keep from forgetting anything, and to take/make better pictures - hope it helps you too; and because I am interested in camera technologies. See the document currently titled "Photography Notes" for more general tips.

If you discover an error in this document or feel inspired to enlighten me further, please Contact Me.

 

Sample of pictures I've taken with the D300.

 

The End