DxOMark review for APS-C camera 2010 : who takes the lead ?

Monday January 24 2011

Sensor Recommendations
Introduction | Use Cases: K5, D7000 and A580 take the lead. | SNR: K5, D7000 and A580 at the top | Dynamic Range: K5 rules! | Case Study: D90 vs D7000 vs D700: | Conclusion

Sony alpha 580 vs Sony alpha 55 vs Canon EOS 60D vs Pentax K5 vs Nikon D7000: K5, D7000 and A580 come first

The five bodies selected for this review represent the latest developments in APS-C sensors. They all possess high-definition figures, with 16 megapixels packed on the sensor and even 18 for the EOS 60D. These cameras also have another point in common: they all display very wide ISO latitudes, ranging from ISO 100 to 12800, 25600, and even 51200 for the Pentax K5. On the spec sheets at least, these figures are impressive, and suggest avery good behavior in low-light environments, perhaps better than for some “ancient” full-frame bodies.

First we will analyze these cameras’ DxO Use Case scores to see which tendencies they reveal. We will then dig deeper in our examination of the SNR and dynamic range metrics that are crucial to this study. And since these APS-C sensors are new, it is legitimate to ask how they compare with their predecessors… and with full-frame sensors.

a55: the audacious choice of a semi-transparent mirror

If the sensors all look roughly similar in their specifications, the Alpha 55 uses a radically different technology than the others bodies with its implementation of a semi-transparent mirror. The technology itself is not new: film cameras such as the Canon Pellix used this kind of mirror back in 1965.

The principle is simple: when light strikes the semi-transparent mirror, 70% manages to go through it, while the remaining 30% is deflected on the AF detection array. There are several advantages for the Alpha 55: the mirror does not move anymore, allowing much faster burst speeds (the Alpha 55 is actually the fastest one by far among the five cameras, reaching an impressive 10 fps); this technology also allows direct and constant live-view operation. While helpful for video operation, the bottom line for photography is that the amount of light hitting the sensor is reduced compared to a camera using a conventional mirror. This can pose problems, especially under low-light conditions, and it is probably the cause of a higher noise level. —This is, by the way, the main reason why the semi-transparent mirror bodies did not manage to seduce photographers in the past. Is it worth improving one function (the video) if it endangers another crucial one (the photography) ?