Sony Cybershot DSC-H9, Pentax K10D, Panasonic DMC-FX30 and Konica-Minolta Dimage A2
 Image stabilization allows shooting at shutter-speeds slower than normally expected. The expected rule is to shoot at 1 over the focal-length of the lens or faster to get a sharp image. Note that for DSLR cameras, the focal-length multiplier must be applied here too. For more details see Digital Camera Values - Episode 6. Remember, ISO can be increased to allow faster shutter-speeds too, up to the point where images become unacceptably noisy. Unlike stabilization, increasing the ISO sensitivity has an effect on moving subjects.
There are many implementations of image stabilization. True image stabilization is either sensor-based, called the sensor-shift method, or lens-based. Manufacturers now claim between 2 and 4.5 stops of advantage standard hand-holding. Here we put to the test 4 cameras to see how these claims hold up. Two of these cameras use a sensor-shift method: the Konica-Minolta Dimage A2 incorporates the original Anti-Shake system and the Pentax K10D uses Pentax's Shake-Reduction. One the lens-based side, we have the recent Sony Cybershot DSC-H9 which incorporates Sony's Super-Steady-Shot and the Panasonic DMC-FX30 which incorporates Panasonic's Mega Image Optical Stabilization system.
Warning: Stabilization has become such a hot feature that many manufacturers claim to offer stabilization even though their cameras do not compensate for any movements. Instead these cameras simply boost ISO sensitivity when motion is detected. Obviously, this has the undesirable effect of reducing image quality. |
3-Stops
|
Our experiment started at only 2 stops over the standard rule-of-thumb but every one of these cameras gave such good results that we can safely claim that they all provide at least 2 stops of extra hand-holding ability. Note that the rule-of-thumb itself is not scientific and results can greatly vary depending on the photographer's steadiness, his posture, his breathing and environmental conditions.
Since night-photography is one of my favorite subjects, I have a lot of practice holding a camera steady using a solid posture and slow breathing. It is importance is to brace the camera against yourself. This puts the Panasonic FX30 at a slight disadvantage since it does not have a viewfinder and therefore must be hold away from your body.
Below are one representative sample from each of these cameras. Each sample was taken hand-held at about 31mm focal-length which normally requires a shutter-speed of 1/30s. About 15 shots were taken from each camera and one representing the most consistent results was chosen. For the 3-stop advantage samples, all samples were taken at 1/4s. The importance was to obtain the right shutter-speed and so the exposure was allowed vary between the different cameras. Specifically, the FX30's lack of manual controls required moving lights until it chose the desired exposure.
Konica-Minolta Dimage A2
Panasonic Lumix DMC-FX30
Pentax K10D
Sony Cybershot DSC-H9
At 3-stops, results are excellent for all 4 cameras. It is noticeable that these cameras do not all produce the same results. But since these are 100% crops from digital cameras with 7+ megapixels, any of these images could be used for a medium-size print and still look sharp. In this round, the best results came from the Sony H9, followed closely by the Konica-Minolta A2, then followed by the Pentax K10D and finally the Panasonic FX30.
|
4-Stops
Now over to the 4-stop tests. The setup remains the same, with lenses around 31mm, but the shutter-speed required was 1/2s. Dimming the lights was required for the FX30, as well as changing the slow-shutter limit in the setup menu since it defaults to 1/4s. Here we are at the limit of what Pentax claims is possible and just beyond what is claimed from the other manufacturers. This is good because it signifies that their claims were reasonably accurate.
Konica-Minolta Dimage A2
Panasonic Lumix DMC-FX30
Pentax K10D
Sony Cybershot DSC-H9
We just lost Panasonic here. The FX30 was unable to produce a single sharp image at 1/2s. This may have to do with its lack of viewfinder, but that is a limitation inherent to this camera. Among the others, the KM A2 and Sony H9 still produced very usable results at 4-stops. The Pentax K10D still did OK but the image definitely shows motion blur. It may be possible to get a small print out of that.
|
5-Stops
|
Now, at 5-stops beyond hand-holding, we are clearly exaggerating. This is beyond the claims of all digital camera manufacturers. However, when testing it is always interesting to know what happens when a system reaches its limit. Here we have the same setup again, but the shutter-speed have dropped to 1 second. Getting the results below require holding your breath after exhaling.
Konica-Minolta Dimage A2
Pentax K10D
Sony Cybershot DSC-H9
With such slow shutter-speed, every camera produces a blurry image. Still, the Konica-Minolta Dimage A2 produced the least blurry image, followed by the Sony Cybershot DSC-H9 and then the Pentax K10D. The Sony H9 shows strong purple-fringing here but that results from the camera capturing a moving image rather than from the lens.
|
This image stabilization shootout shows that true stabilization is very effective at obtaining acceptably sharp images with slower shutter-speeds and that the claims by these manufacturers are reasonably accurate. Every camera tested here produced good results at up to 3 stops over the standard rule-of-thumb. It is important to know that these results are very sensitive to their specific setup and may vary greatly. What this demonstrates is that these results are achievable, not that they can be obtained in all conditions. As a matter of fact, when we did the Sony Alpha A100 vs the Pentax K10D article, the Pentax K10D did not perform quite as well. The main differences were a longer focal-length and a heavier lens.
|
![[Valid RSS]](images/valid-rss.png "Validate my RSS feed"){kind=small}
