Which digital cameras are least sensitive to camera shake?

[JDArt]

I've personally noticed that some cameras seem more prone to "camera shake" than others.

Investigating the topic further, I read with interest this account of how "camera shake" is intensified on the Nikon D2x:

=======
"How could the D2X's smaller pixels make it more sensitive to camera shake than cameras with larger pixels?

Look at these diagrams and you'll see that the same amount of movement more than doubles the amount of pixels that are influenced by camera shake when comparing the D2X to a typical six megapixel camera.

I've always had a pretty steady shooting hand, and haven't run into much of a problem with this. There have been some pictures I've taken, however, that at first glance looked a bit soft & out of focus, but on closer inspection turned out to be slightly motion blurred. (It's hard to tell the difference between sharply focused pictures with a soft /slight motion blur and pictures that are truly slightly out of focus.)







To overcome the problem, some D2X shooters double the traditional shutter speed rule of thumb (a minimum of 1/ lens focal length shutter speed). In other words, if you're shooting at a 200mm focal length and the old rule of thumb equaled 1/200th of a second, do what's necessary to double the shutter speed to 1/400th in order to overcome the greater camera shake & motion blur sensitivity of the D2X's smaller pixels. You can easily control this by observing the shutter speed in your viewfinder and making the necessary adjustments to Vari-Program, Aperture or ISO settings to increase shutter speeds, or by shooting Shutter Priority and presetting the shutter speed in accordance with your focal length.


http://www.lonestardigital.com/D2X.htm

============

I find the author's solution to the problem less than satisfying, as it's often impossible when handholding a camera to get shutter speeds twice that of what you'd normally want -- unless you increase ISO to ridiculous levels.

So my interest is in finding out which digital cameras on the market are least likely to magnify camera shake.

Does anyone know? I would suspect those large sensor, large pixel cameras like the 1DS (original)?

[dc2]

I noticed a huge difference with shake when I jumped to the 1Ds2 from a 1od and after reading an EOS mag it comfirmed your findings here .
Canons answer was higher speeds / ISO / triopod as you say less than satifying .
If anything though it dose force you to get better at holding the camera more steady and certainly a reason I much prefer IS lenses now .

[dougjgreen]

The ones mounted on a tripod are least susceptible to shake

[ohenry]

Originally Posted by dougjgreen

The ones mounted on a tripod are least susceptible to shake

[dc2]

Originally Posted by ohenry

OH god I need a holiday

[deluco]

...as well as the ones on which you use IS lenses.

[Mr. Pickles]

Yeah the shake is worse, but is it really? I mean the amount it shook is the same. The number of pixels is greater. Thus at 100% magnification, you will see more "shake", right? But if you have twice the pixels, then view it at 1/2 the magnification.

I think if you shoot both cameras (10-plus and 6-8mps) and print them both at 8x10 or whatever, the bigger pixel count (smaller micron pixel) will still look better, as in less shake.

If you print the 6mp image at 8x10, and the 10-plus one at 13x19, then they are probably the same in shakiness.

If viewed at 100%, yeah the higher pixel count will show more shake...but it will show more noise too. If you base the images on these things, you wouldn't buy it.

[scoundrel1728]

The amount of motion blur you will get in an image will depend on the distance the image moves on the imager while the shutter is open.

The greater the (35 mm equivalent) focal length, the greater the susceptibility to camera shake.

For a given final image size, the greater the resolution standards, the greater the susceptibility to image shake: it will take less movement to degrade an otherwise high-resolution photo noticeably than one where the blur would be masked by degradation from other sources.

The greater the final image size, given a fixed viewer-to-image distance, the greater the susceptibility to image shake. This final cause is why image shake is more noticeable on higher-megapixel imagers when displayed on monitors: the on-screen images from these imagers are blown up bigger, specifically by the ratio of the pixel pitch on the imager to the (logical) pixel pitch on the monitor.

And, of course, the steadier the camera is held, the less the image will move. A tripod-mounted camera will have a steadier image than a handheld one and one held by steady hands will be steadier than one held by less steady ones. There may also be subtle differences in how steady a given individual will hold the camera and lens depending on the camera's shape, overall mass, location of its center of mass, and the moments of inertia.

The idea behind lens-based image stabilization is to move certain lens elements just enough to cancel the movement on the imager. Imager-based image stabilization - or anti-shake, as Konica-Minolta prefers to call it - moves the imager around to track the image motion, effectively cancelling the relative motion.

The imager-based anti-shake mechanism, which effectively gives every lens mounted on it anti-shake capability, is the main reason I am leaning so strongly to the Konica-Minolta 5D despite the imminent takeover of its camera division by Sony. Phil Askey is of the opinion that moving the imager may be less effective than moving some of the lens elements. He may be right; nevertheless, the difference isn't obvious and buying the option once rather than with every lens that might need it should save quite a bit of money for someone who would be doing a fair amount of telephoto photography.

[scoundrel1728]

The 1/focal length rule of thumb for shutter speed is based upon a specific criterion for image sharpness and (maximum limiting) angular velocity for camera movement. As long as the unsteadiness of the photographer's muscles is the main factor for camera movement, the bandwidth of the velocity will be limited. For shutter speeds much higher than that bandwidth limit, the image will be traveling at nearly constant velocity while the shutter is open, hence the amount of blur will be largely proportional to the length of time the shutter is open. At slower shutter speeds or where vibration is transmitted to the camera from other sources - mirror recoil, for example, or shooting from a vibrating structure like an airplane - this rule of thumb will break down.

This brings up the question: Will the vibrations of mirror recoil and the internal vibrations thus set up die away before the camera's shutter opens? If not, how much will they contribute to camera movement?

[dc2]

This brings up the question: Will the vibrations of mirror recoil and the internal vibrations thus set up die away before the camera's shutter opens? If not, how much will they contribute to camera movement?

I have thought the mirror seems to be an old pice of engineering for modern DSLR'S
I would hope it will be replaced soon . How exactly I am not sure