More on Factoring In the Crop Factor: Hand Holding
The generic rule of thumb is that you can expect to successfully hand hold your camera down to a shutter speed of one over the focal length. For example, if you are shooting with a 200mm lens, hand held shots should generally be sharp if your shutter speed is at least 1/200 second. The rule was devised back in the film days though, so it's worth taking a look at whether digital has changed things — especially after spending the past two weeks looking at other effects of the digital crop factor.
When you take a picture, the results will be sharp so long as the camera doesn't move very much. Deferring for a moment just how much is too much, the effect is definitely proportional to focal length. A longer lens magnifies more than a shorter one, and the effects of any camera shake will get multiplied along with it. If you've shot with a long telephoto, you know how frustrating it can be to get sharp results, just as you likely have found out how much easier it can be with a wide angle lens if you have one of them in your camera bag.
But as for where specifically the "one over focal length" formula came from, I can find nothing inherent or magical in lens design that would yield this particular result. Clearly, there is some basis to it and longer lenses require faster shutter speeds since they magnify the image proportionally to their focal length. So the basic idea of 1/focal length is correct, but what bugs me is the elegant simplicity of the formula seems to have no real basis in physics. I mean, why not 0.864/focal length, or perhaps 1.127/focal length? How did it work out to precisely "one" divided by focal length? Or is it more likely that, as a rule of thumb, it was always just an approximation to make doing the math in your head easier?
It's undoubtedly not the same for every photographer and circumstance either. Clearly, some people have a steadier hand than others, and some hand-holding techniques inherently yield better results than others. The weight of the camera and lens must play some role too, whether the greater mass produces more stability by having a greater moment of inertia, or whether it produces less stability by inducing muscle strain in the photographer holding it isn't the point. Either way, it will affect your results, and it has nothing to do with focal length. Of course, if you have a Vibration Reduction or Image Stabilization lens, it will produce a significant change to hand hold-ability too.
The effect though is definitely dependent on focal length, even if it must also depend on other factors. All else being equal, the smaller field of view from a longer lens will magnify the appearance of any camera shake we may induce through hand holding. If a given lens sees a particular angle of view measured in degrees, a 0.1 degree angular movement of the camera will result in things shifting by by a certain amount. If the angle of view that lens sees decreases due to zooming it to a longer focal length, that same 0.1 degree camera movement will be a greater portion of the total angle of view, resulting in a more apparent blurring of the image.
Also, whatever the track record for hand-holding a given photographer may have, he would need to be steadier with digital than film when shooting with the same lens. Something that would be rendered acceptably sharp on film may not be acceptably sharp if we substitute a smaller sensor since we would need to use a lower circle of confusion value for digital. Movement induced by the photographer hand holding the camera may be small enough to still produce an acceptably sharp image on film but large enough not to on a digital sensor.
But the image that photographer would end up with when using any given lens on digital will not give them the same image as they would have gotten on film. The crop factor would mean that the digital image would be more tightly cropped. If we change lenses to compensate, we also change focal length, on which the rule of thumb depends. In fact, since we would end up with the same field of view, any camera movement would affect the resulting image by the same amount. A camera shift of a given angle would cause things to move in the frame by the same amount since that angle would translate into the same lateral shift in inches at the subject distance. The field of view of a 300mm lens on a film body would be the same as a 200mm lens at the same subject distance on a 1.5x digital body, and any camera movement would cause the subject to move by the same amount on each. Again in terms of angle of view, that 0.1 degree camera movement I used as an example earlier would be the same proportion of the total angle of view for both the 300mm lens on film as it would be for the 200mm lens on digital. It musts be, since the angles of view for each setup would be the same due to the change in crop factor.
Does it make sense then to replace the simple "one over focal length" with "one over 1.5 time focal length" or something similar? It depends on just how accurate you consider the rule to be in the first place. I am unsure as to whether the increased shutter speed you would need for any given focal length on digital is sufficient to rewrite this well ingrained rule of thumb. Who's to say whether the error in approximation the rule already has due to the various factors discussed here offsets or augments the change that would be needed due to the change in crop factor?
To be safe, it is probably best to factor in the crop factor when making decisions about hand holding. Multiplying the focal length to obtain the "35mm equivalent" before dividing to get the inverse for your maximum acceptable hand holding shutter speed seems prudent even though you likely can go beyond it if you have steady hands and often could shoot with somewhat lower shutter speeds when using film. To be even safer though you will be even more certain of getting sharp results if you try to stay a stop or so faster than the rule would dictate, regardless of whether you use film or digital.
Personally, I shoot virtually everything with the camera mounted on a tripod which is always the best option when possible. That way, I don't need to worry about this whole hand holding thing at all.
Update 12/28/2007 - Now that both Canon and Nikon have "full-frame" digital SLR's, one needs to be careful when discussing crop factor. It's no longer just a film-versus-digital thing since Nikon FX and other full-frame sensors behave the same as film with regard to crop factor effects. Nikon FX has a simple 1.0 crop factor since, relative to 35mm film, it does not crop. Nikon DX still has a 1.5 crop factor, while Canon makes bodies with a variety of crop factors.