Home >> Photography >> Digital

Pixel-peeping the GF1; GH-1; D2X; 7D; 5D2

ISO 100

Panasonic GF1	Panasonic GH-1
Nikon D2X	Canon 7D
Canon 5D2

ISO 200

GF1	GH-1
D2X	7D
5D2

ISO 400

GF1	GH-1
D2X	7D
5D2

ISO 800

GF1
D2X	7D
5D2

ISO 1600

GF1	GH-1
D2X	7D
5D2

ISO 3200

GF1	GH-1
D2X	7D
5D2

Any surprises? I was initially surprised that from ISO 800 onwards, the GH-1 clearly does better than the GF1, but then I remembered that the GF1 uses the same sensor as the older G1 while the GH-1 uses a newer sensor (with probably some processing enhancements).

As expexted the full frame Canon 5D2 maintains superiority throughout the comparison and really shows why those that need quality in low and/or questionable quality light really need a full frame sensor to eke out the best quality.

To my eyes, the DX format 7D maintains a one-stop advantage over the M4/3 sensor while the full frame 5D2 maintains a one-stop advantage over the 7D. If directly comparing the GF1 to the 5D2, I would suggest a three-stop advantage.

With the GH-1, I would be okay using it at up to ISO 1600, but the GF1 is quite questionable at this ISO speed.

Looking at the ISO 3200 images, I can see why in the beginning, subjectively, I was not that happy with the 7D's quality. Now I can see why, as the 5D2 is clearly better. While the 7D is clearly better than the M4/3 cameras, it still has a bit of a "crunchy" quality to it that will limit how large you can print.

Do consider that these results are with some modicum of technical parameters to provide a stable platform and consistent lighting. Real world results are likely to be worse than this if handholding the cameras in poor lighting conditions.

Technical
All cameras mounted to a Gitzo 1345 tripod. Two-second self-timer mode used for all cameras to avoid me disturbing the cameras right before shutter release. Mirror lockup was not used with the Canon or Nikon SLRs because the Panasonic cameras do not have this feature (and have no need for it due to not having a mirror). You can quibble about the Canons and Nikon being disadvantaged slightly, but this is a high ISO noise test, not a resolution test.

Cameras set to manual mode using a reference exposure of 1 second at f4 and ISO 100. Files imported into Lightroom 2.5 at default settings with only white balance correction and cropping done before exporting the files as maximum quality JPEGs tagged with the sRGB color space. A WhiBal card was included in each shot to allow for white balance correction in Lightroom.

The Canon 7D and 5D2 used the Canon 50mm f1.8 lens. The Nikon D2X used the Sigma 30mm f1.4 lens. The Panasonic cameras used the Panasonic 20mm f1.7 lens. The tripod was moved forward or backwards as needed to keep the framing and field of view consistent for each camera. The size difference you see is the relative difference in resolution for each camera.

White Balance and Metering

White balance set with the WhiBal card	Nikon D2X Auto WB and default exposure
GF1 Auto WB and default exposure	GH-1 Auto WB and default exposure
Canon 7D Auto WB and default exposure	Canon 5D2 Auto WB and default exposure

The first shot shows what the proper white balance should be after clicking on the white section of the WhiBal card. The other shots show how far the cameras deviate from proper white balance in Auto White Balance mode. The photos were taken in my rec room with ceiling pot lights as the only light source.

The reference exposure of 1 second @ f4 with ISO 100 was set with the Nikon D2X. The Canon 7D gave the same exposure and the GF1 is about 1/3 of a stop under the reference exposure. The GH-1 is a half-stop under as is the 5D2, but the 5D2 looks subjectively darker and the worse looking of the group.

The difference in framing with the D2X and the 5D2 is due to their higher placement on top of the tripod head due to the D2X's integral grip and the 5D2 having its accessory grip attached.

Subjectively, the GF1 offers the most pleasing combination of exposure and white balance, but as you can see, it's quite a bit warmer than what the true white balance is. Personally, I don't find it objectionable at all compared to the yellowish WB from the D2X and the orangish WB from the Canons. However, this is only one type of WB/lighting test and the cameras will perform differently in other settings such as outdoors or with alternative lighting.

Correspondence & About this website     Copyright © 1998-2010 Edwin Leong