Sony Alpha A900

Sharpness

To test image sharpness, we shoot our test chart at three zoom settings (widest, longest zoom and mid-range), at three focal lengths for each distance. We found that, outfitted with the 24-105mm lens used in our testing, the Sony A900 produced a very good image sharpness result overall, significantly higher than either of its full-frame rivals (the Canon 5D Mark II and Nikon D700). Only the Nikon D90 fared better in our sharpness testing, and not by much of a margin.

At 24mm, the widest lens setting, the image is very sharp in the center, with a maximum lw/ph measurement of 1718 horizontally and 2011 vertically, when shot at f/9. At this zoom length, performance remains strong from corner to corner. More on how we test sharpness.

Image Sharpness and Chromatic Aberration

Focal Length:

24.0mm

70.0mm

105.0mm

Click blue squares to change sample crops below
f/3.5 f/9 f/22
f/4.5 f/10 f/29
f/4.5 f/11 f/29

Image Stabilization (4.71)

The A900 did a decent job of compensating for handshake; we found that with low levels of hand shake, images with the SteadyShot stabilization on were significantly sharper than with it off. However, this was reversed with larger amounts of hand shake; the system made the images slightly less sharp. Most SLRs try and deal with the way your hands shake by moving a small part of the lens, but the A900 takes a different approach; the image sensor is on a moving platform. When the gyroscopes in the camera detect that it is moving because your hands are shaking, this small platform moves to compensate. We test how well a camera deals with hand shake by putting it on a motion control platform, shaking the camera and taking a large number of photos with image stabilization on and off. We also test by moving the camera in both a horizontal (left to right) and vertical (up and down) direction, with two different levels of shake; one low and one high. We then analyze these photos to determine how sharp they are; the better the job the camera does of compensating for the camera movement, the sharper the photos will be. For more details on how this test works, see our How We Test article.

Our first test is for a low level of hand shake, such as when you are trying to hold the camera steady with two hands, or braced against a wall. In this situation, the camera only moves a small amount, but there is still more than enough shake to ruin an otherwise perfectly good picture. In this case, the SteadyShot feature of the A900 led to sharper images at all of the shutter speeds that we test at, ranging from 1/500 of a second right down to 1/8 of a second. The only exceptions here were when the camera was moving vertically (up and down); at 1/500 of a second and 1/60 of a second, the SteadyShot system made things slightly worse. But overall, it did a very good job of compensating for the simulated hand shake in this test.

Our next test uses a higher level of shake; about the level you would get if you were trying to take a photo while walking, or while taking a candid shot one-handed. The A900 didn’t do as well here: the larger amount of shake seemed to be more than the SteadyShot system could cope with when the camera was moving horizontally (left to right). On most of the shutter speeds the images taken with SteadyShot on were less sharp than with it turned off, as you can see in the examples below. There was some improvement in sharpness with vertical (up-down) movement, though, but we feel that the horizontal correction is more important,since handshake is more of a left-to-right movement than up-down.

The bottom line here is that we would recommend that you leave SteadyShot turned on in any situation where hand shake is likely to be an issue, as it does a good job of making shots significantly sharper across the range of shutter speeds with small amounts of hand shake. However, like all image stabilization systems, it can’t work miracles, and in situations with larger amounts of shake, it sometimes made things a little worse. So, turn it off if you are running around, but turn it on if you are standing still to take shots.

The A900 didn’t have the problems that the Canon 5D Mark II had with stabilization results worse at some shutter speeds; it made improvements across the entire range. But it’s no slam dunk; we found that the image stabilization of the Nikon D700 produced sharper images at the critical shutter speeds of 1/60 and 1/125 of a second. However, in overall performance, there was not that much difference between the A900 and D700. You should remember, however, that the performance of the 5D Mark II and D700 is dependent on the lens, as the image stabilization on these cameras happens in the lens itself. This means that if you were to use either camera with a different lens, you would probably get very different results.

It is also worth remembering that the two different types of image stabilization have their pros and cons. The sensor approach used by the A900 means that the lens does not require any complex built-in mechanisms, and you’ll get the benefit of image stabilization whatever type of lens you use. This could also mean that the lenses are cheaper as they are easier to make. However, the lens-based image stabilization that cameras like the Canon 5D Mark II and the D700 use means that you can upgrade the image stabilization whenever you buy a new lens, and the lens manufacturers are constantly improving the technology that they use.

Below are stills from some of our test photos taken with the Sony A900, showing an average image from the horizontal test. The target is a slanted line.

Image Stabilization Comparison Table			Expand
	Low Shake IS Off	Low Shake IS On	High Shake IS Off	High Shake IS On
1/500
1/250
1/125
1/60
1/30
1/15
1/8