Canon 5D mk II vs Panasonic GF1 (RAW): Pt. 2 - Blue Sky & Sunshine

Its been quite a long time since the last time I posted any comparison shots. Since then I've bitten the bullet and "upgraded" all my camera equipment and have a new 5D MkII with the Canon 17-40 f4, Canon 50mm f1.8, 85mm f1.8 and Tamron 70-200mm f2.8 lenses.

The aim of these posts isn't really to determine which is better - they're 2 very different cameras, but more to give an impression of how much you gain or loose by switching from one to the other.

Click on the title to see the full blog post with comparisons!

The Cameras:

Camera:	Canon 5D Mark II	Panasonic DMC GF1
Lens	Canon EF 17-40mm f/4L USM	Panasonic 20mm f/1.7 Aspheric
Weight (total)	1310g	385g
Price	$3228 (body + lens)	$689 (kit)
Raw file size	29.2 MB	13.9 MB
Aperture	f11	f5.6
Shutter speed	1/160 seconds	1/1250 seconds
ISO rating	100	100
White Balance	Auto	Auto
Focus	Auto	Manual

Notes about Shooting and Processing

Both images are taken hand-held with camera default settings. I didn't attempt to equallise the exposure by shooting in full manual mode, instead chose aperture priority with full frame equivalent apertures - ie. f5.6 on the GF1 and f11 on the 5DII. The resulting shutterspeed is vastly different. That's due to 3 factors: 1- the 2 stops difference in aperture (ie. the GF1's shutterspeed should be 4 times as fast), 2- The base ISO setting on the GF1: ISO is really iso 125, 3- the exposure settings on the GF1 are designed to protect highlights so the image was underexposed by about 2/3 stop.

In post processing my aim was to get the images to look as close as possible from the RAW files - that meant: cropping the GF1 to 3:2 (a more "fair" approach would have been to shoot the 5D at slightly wider and crop that to 4:3), adjusting the white balance and colors slightly and probably most significantly increasing the exposure by about 2/3 stop. The 5D II's image has had distortion correction applied and chromatic aberration removal applied (which the happens with GF1 files automaticaly). In the end there's still a slight color difference, but given that we're dealing with 2 raw images, both have plenty of room to adjust.

Observations:

Sharpness:

At a pixel level, both images are incredibly sharp. The 5DII has a lot more pixels, and its showing that resolution. This is particularly evident in the second crop where you can see a wire mesh in the 5D image that's almost invisible in the GF1 image. This mesh is about 2 pizels wide on the 21MP image, so I'm surprised that you can even see it on the GF1 (but you have to look very closely). At the extreme corners teh GF1 image is marginally sharper, though given that the extreme corners show different things in this aspect ratio, I haven't included these crops.

Noise

The GF1 shows a little bit more noise - in part that's due to the way the files have been treated - but in a situation like this if you push the exposure, you also risk blowing the highlighs due to the slightly lower dynamic range. To see the noise you need to look at the sky at 100% or peek into the shadows. Noise at these levels wouldn't show up in print.

Dynamic Range

To me Dynamic range isn't really a problem in this image. The ability to pull back the highlights means the 5D image results in less noise. the other advantage is that in post processing its easier to get the bright areas such as the gold leaf looking natural with smoother highlight transitions.

Stars

The first crop shows the difference caused by the design of the aperture blades in each lens. The 20mm lens has a "traditional" 7-bladed aperture design, while the 17-40mm also has 7 blades, but these are circular in design which means that the highlights don't diffract quite as evenly or uniformly.

Fairness

This isn't intended to be a fair test. In hindsight I'd probably do this with a slightly wider 5D II image to get the same field of view on a 4/3 axis, but the size differences are really quite small either way.