Left: Panasonic Lumix GF1 camera and 14-45mm f/3.5-5.6 lens Right: Canon EOS-40D and EF-S 17-85mm f/4-5.6 lens

Several micro four thirds cameras (Panasonic and Olympus) were evaluated as a small, light-weight travel camera to substitute for my Canon EOS-40D when space and weight restrictions during travel would benefit from a smaller camera system. The main objective initially was to compare image quality produced by micro four thirds cameras with that produced by the EOS-40D. After initial tests were run, I concluded that additional tests on dynamic range, resolution and high ISO noise were warranted. The following cameras were used in the tests: Panasonic Lumix G1, GF1 and GF2 Olympus E-PL1 Canon EOS-40D and EOS-5D Mark II

Initial tests indicated that at low ISO settings (ISO 100-400), photographs produced by the Panasonic and Olympus micro four thirds were closely comparable in overall image quality to those produced by the EOS-40D. Images printed at 16 inches width were hard to tell apart between the two camera systems other than color output. Auto white balance and default print color output appeared better in the images produced by the EOS-40D. In some images with wide lighting extremes, the images produced by the Panasonic cameras appeared to have blown out highlights. The images appeared to be overexposed, but it raised the possiblity that the Panasomic cameras had narrower dynamic range than the other cameras.
Images produced at ISO 800 with the Panasonic and Olympus cameras appeared to have high noise. The following additional tests were run:

• Dynamic range using photographic step tablets and Imatest Master software
• Noise calculated from step tablet exposures and from SFRplus with Imatest Master software
• Resolution measured from ISO12233 chart
• MTF50 calculated from slant-edge in ISO12233 chart and from with Imatest Master software

Initial Photographic Test Images

Images of parking garage shot at ISO 100 and f/8 with GF1 camera and 20mm f/1.7 lens, G1 Camera and 14-45mm lens, and EOS-40D camera with 17-85mm lens.	Center of parking garage image

Images of campus dining building shot in a mixture of bright light and shade with a Panasonic Lumix GF1 camera and 20mm f/1.7 lens as well as an EOS-40D camera with EF-S 17-85mm lens. Brick detail in the sunlight in the left of the field is lost in the GF1 image. Is this overexposure by the GF1 or a reflection of narrow dynamic range? Further tests on dynamic range were run.

Dynamic Range and Noise

Methods used with Imatest Master and with Step Tablets

Images of a Kodak Photographic Step Tablet 3 shot with Olympus E-PL1 (upper panel) and Panasonic GF2 (lower panel) camera.

EV units in this dynamic range test are based on a 0.15 density 21-step Kodak step tablet and Imatest (see Imatest reference).	EV units in this dynamic range test are based on a 0.10 density 41-step Stouffer tablet and Imatest (see Imatest reference).
Luminance Channel Noise from Kodak Step Tablet. Micro four thirds cameras had the highest luminance channel noise, especially at high ISO settings. Mean ± SEM (n=3)	Noise measured from the SFRplus with Imatest Master. Mean ± SEM (n=3) Standard error of the mean is less than the size of the symbols.

Resolution and MTF50

Among the micro four third cameras tested, the GF2 and the E-PL1 had the highest MTF50 scores at ISO1600 and 3200. The E-PL1 had the highest resolution (panel to right). The EOS-40D can't be compared directly to the micro four third cameras because tests were run with a different format chip and different lens. The 40D test was run to see how MTF50 was affected by ISO setting. The GF1 has decreased MTF50 at ISO 1600 and 3200. Mean ± SEM (n=3)

High ISO settings decreased resolution in the EOS-40D. The micro four thirds cameras were affected only mildly, if at all. Measurements made off ISO12233chart. Mean ± SEM (n=3). Standard error is less than the size of the symbols.

Conclusions from Tests
After shooting with DSLR cameras for 11 years, I spent a little time evaluating micro four third cameras before adopting them as a light weight travel alternative to the Canon D30, 10D, 20D and 40D cameras I have been using over the years. The small size and light weight of the micro four third camera systems is very attractive. My conclusions based on experience of using the cameras for about 2 months and extensive testing:

• Quality of 16 inch wide prints prepared from raw files is comparable between micro four thirds cameras and an EOS-40D. Greater post-exposure processing may be required to get the color you want in prints with the micro four third cameras.
• The Lumix GF2 had the greatest dynamic range of the cameras tested.
• Among micro four thirds cameras, the Lumix GF2 and Olympus E-PL1 had the highest MTF50 scores across all ISO settings
• High ISO noise is greater in micro four thirds cameras than in Canon EOS cameras that I tested (40D and 5D mark II)
  • The Olympus E-PL1 had the lowest noise of the micro four thirds cameras tested. It also had the highest resolution.

Snap shots with micro four thirds cameras Lenses Used: Panasonic 14-45mm f/4-5.6 Panasonic 14mm f/2.5 Panasonic 7-14mm f/4 Olympus 9-18mm f/4-5.6 Nikkor 50mm f/1.4 AIS

Paynes Praire Panasonic GF1 with Olympus 9-18mm lens @ 9mm

Index

Intially Posted 09 January 2011; revised 17 January 2011