No, I'm not into S&M, though sometimes achieving best shot discipline seems like it.
I keep writing about "shot discipline," so perhaps it's time that I laid out what I mean by that phrase. Another word I use to talk about the same thing is "optimization." If you optimize your equipment, your procedures, your settings, and your data collection, you end up with pixels at the other end that are the best that they can be. So just what do I mean when I use these terms?
It starts before you photograph. Here are the five big items to consider:
- You start with clean equipment. A clean sensor, clean lenses, clean carrying cases, clean contacts, clean viewfinder eyepiece with the diopter set right for your eyesight, clean LCD, and a clean (empty and known good) card. Every one of these things can cause you either to get sub-optimal data (dirty sensor, dirty lens), or even miss a shot (dirty contacts, full or failing card). While you're at it, a full battery with clean contacts is a good thing, too. I have recommendations about these things in my books, but the basics--with one exception--boil down to you don't put the horse away wet. What I mean by that is that at the end of a day (or sometimes session) of shooting, you spend ten minutes making sure that everything is returned to clean. The exception is the sensor. I don't get anal about sensor cleaning, as doing it often in the field sometimes just exposes the sensor chamber to even more dust that eventually attaches itself to the sensor. I clean the sensor once just before going out on a big shoot, then only after that when I notice that I have dust spots that are objectionable. But if you were going for truly optimal, you'd clean the sensor just before each day's shoot (not at the end of the previous day, as the camera is likely still warm and you want any dust in the chamber to settle before cleaning).
- You use proper equipment. That means that there's a hood on your (clean) lens (and sometimes an additional "flag" to shade the shade). That you use a sturdy tripod that has been chosen correctly for the mass you're putting on it. That you remove all "poor joints" that have any give to them. With my current support equipment, it is so good that the primary source of vibration in my system is Nikon's terrible tripod collars, which have to be really over tightened past what you consider normal to take out as much give as possible. With the long lenses (e.g. 400mm f/2.8), using the Really Right Stuff brace kit takes the tripod collar out of the equation, but with some of the shorter lenses (70-200mm, 200mm f/2, 70-180mm) you really have to work at removing the "collar slop" Nikon has engineered in. I should be able to tap the very front of your lens and see no movement whatsoever in the camera body. None. Nada. Zero. If you see any, you're not locked down and you wasted a lot of money on the rest of your support system. Put another way: your support is only as good as the weakest link. You have to make sure that your weakest link is well beyond the point where it can cause any vibration or movement issues.
- You get the settings right. Here's where most photographers start to make many mistakes, and most of those show up in the data you store on your card. If you're using the histogram to judge exposure and are shooting raw, there's only one way to get truly useful information: UniWB and a linear Custom Curve. Anything else and you will likely set exposures slightly lower than you could, which means less dynamic range in your data than you could have recorded. Some also believe that using a 30M or similar filter with UniWB is the right thing to do outdoors, as it balances the color response of all channels better. I'm not 100% convinced on this one, as now you have to worry about filter quality and you're letting a little less light back to the sensor for that benefit of channel balancing. Few folk are shooting at a high enough level to get tangible benefit from the 30M with UniWB, is my guess. With the D300 and D3x, 14-bit shooting is definitely something the raw shooter needs to consider if they're trying to extract every last bit of detail out of their camera. JPEG shooters don't have a lot of choice: besides the right exposure they need to get white balance and all camera settings dead on. Anything less and you have suboptimal data compared to what the camera could capture. Many of the complaints I've seen about noise from various cameras comes back to this: wrong settings shooting JPEG. If you're off by even a couple hundred Kelvin in incandescent light, when you rebalance to neutral in post processing, you're swinging a channel that didn't get a lot of light in the first place (blue), and when you do you often get a noise buildup. Need I mention JPEG Optimal Quality and Fine settings?
- You thought seriously about the lens. I'm not going to say primes are better than zooms, as in many cases that's no longer true. But Lens A may be better than Lens B at the same settings. That's why in my full lens rationalization article I speak about which lens is best at each focal length. At 35mm, for instance, I don't want to be using the 17-35mm if I care about ultimate image quality and happen to have one of several better lenses in my bag (the 24-70mm at 35mm, for example). Knowing which of your lenses gives you the best results at any given focal length is something every serious photographer needs to know. Failure just on that one thing can lead to plenty of suboptimal data: even something like having to use vignette correction tends to mean that the corners of your image will be a bit noisier than the rest. But corner performance in general is one thing that you need to seriously look at. The 70-300mm is better in the corners than the more expensive 70-200mm at equivalent apertures on an FX body, for example. If you don't know that, you can't optimize getting the best possible pixels out of your equipment. But "getting the lens right" is more than just choosing the right one. Have you considered diffraction, bokeh, or even something simple like the ability to fine tune focus? In order: you shouldn't choose apertures that produce serious diffraction in your pixel data unless you absolutely have no other option; when you're using wide apertures and are thowing the background out of focus you need to consider what that out of focus area will look like with your lens choices; and many of those older MF lenses have focus rings that go 180 degrees or more to get from near focus to infinity, which means that you can manually tweak to very high precision that you can't get with the recent AF lenses that have focus rings that do the same in 45 degrees or less. And did you use Live View to check focus (if appropriate)? No? Tsk, tsk. If you're using long lenses with autofocus, I should also point out that you need to have done enough AF Fine Tune testing to know where the camera needs to be set (D300 and higher). That might be different for near objects and far objects, by the way.
- You sweat the details on the actual shot. Did you just use Live View? Give the camera a short rest so that it cools back down. Indeed, have you just left the camera out in the hot sun for three hours while shooting? The rule of thumb is that noise increases at the sensor in visible increments every 5 degrees F. Hmm. Death Valley in the summer is sounding like a serious potential noise-fest. What else? Did you turn VR off if you're on a tripod or shooting above 1/500? If you are shooting on a tripod, are you using mirror lockup with a cable release? Did you weight down your tripod? Did you even test to see if your tripod was transmitting vibrations?
Assuming you've done all the above, congratulations, you've got pretty good shot discipline and you've probably optimized the data you collected. Now all you have to do is optimally process and output your optimal data and you're the next Ansel Adams.