Sunday, November 28, 2010

G6 vs G12: Noise


As noted previously, there seems to be little difference between the in-camera JPEGs produced by the G6 and G12 when the images are seen on screen (without pixel peeking) or viewed as an 8x10 print.  But those images were made at the lowest ISO settings.  What about comparing higher ISO results?

Fortunately, my low light test scene had remained relatively untouched since the comparisons between the G9 and G12 at high ISO settings.  The above shot was taken with the G6 at ISO 400, f4 and 0.6 seconds.  (Apparently about 2 stops less available light from the previous tests -- don't know why.)  Not too bad at first glance (click for enlarged view). I also re-tested the G12 under the same lighting conditions.

Next, we yield again to temptation and take a peek at the pixels.  The assembled comparison below shows various 100% crops from both the G6 and G12 (click for enlarged view).

Based on pixel peeping, one might say that the G6 is not usable at ISO 400 but as seen in the opening image, small image sizes could still be OK.  It appears to me that the G12 is about 2 stops better than the G6 with respect to noise.  That is, I'd prefer to use the G12 at ISO 1600 than to use the G6 at ISO 400.

Of course, the G6 has a one stop aperture advantage over the G12 but then the G12 has image stabilization for which is claimed a two stop improvement in shutter speed.

My conclusion is that, in spite of having fewer pixels, the G6 is "noisier" than the G12.  Whether the G12 noise reduction is due to improvements in sensor technology or noise reduction processing in-camera, I couldn't say (probably both).  Obviously, there is more to noise than pixel count or pixel density.

No comments: