Olympus EVOLT E-510

Color (10.22)
To test the color accuracy of the Olympus E-510, we photographed an industry standard GretagMacbeth Colorchecker test chart, which consists of 24 strategically picked color tiles. The colors correspond to commonly photographed colors such as blue skies, green grass, and flesh tones, as well as a sample of other colors from around the color spectrum. We ran these photographs through Imatest, which compared the actual color of the tiles with the colors the camera produced. In the image below, the outside square of each tile corresponds to the color the camera produced, the inside square is the ideal color of the chart corrected for luminance, and the small rectangle is the ideal color of the chart.

The camera produced the most accurate colors when 1/3 of a stop underexposed, which is the reason why the small color rectangles look lighter than the colors the camera produced. With luminance taken into account, most colors look very accurate. The following graph shows precisely how far off the camera’s colors are from those on the test chart. The squares indicate the ideal, while the circles represent the colors the camera reproduced. The line between them illustrates the degree of error.

With an uncorrected mean color error of only 5.87, the E-510 scored very well. The colors to have significant drift are mostly the blues, which are often shifted on purpose to enhance skies. However, the saturation level of 96.46 percent means it is also fairly undersaturated.

White Balance (6.31)
Auto (6.00)
We tested white balance accuracy by photographing the ColorChecker in four different types of light: flash, fluorescent, outdoor cloudy, and tungsten. The E-510's automatic white balance was not very good. Flash accuracy using the Auto setting was solid, and under cloudy light it was decent, but under fluorescent and tungsten it was miserable. The images below show what kind of color cast your photos will have in these different situations.


Preset (6.62)
The E-510's white balance presets faired a little better than the Auto setting, but not by much. The cloudy preset was very accurate, but fluorescent and tungsten were as miserable as using auto. Under tungsten light, the Auto setting will give you a very strong yellow cast, while the preset will turn everything blue. Considering the difficulties the camera had manually white balancing in low light, users should bless the camera gods that the E-510 can shoot in RAW, so that you can adjust the white balance later.

Shade WB - Shade Illumination

Fluorescent WB - Fluorscent Illumination

Tungsten WB - Tungsten Illumination

Still Life Sequences
Click on the thumbnails below to view the full resolution files.

Still Life Scene
ISO 100	ISO 100
ISO 200	ISO 200
ISO 400	ISO 400
ISO 800	ISO 800
ISO 1600	ISO 1600

Resolution (8.76)
Packing 10 megapixels onto its small Four Thirds sensor, the Olympus E-510 entices consumers with higher resolution than ever before. To test the E-510's performance we shot an industry standard resolution chart at varied apertures, shutter speeds, and focal lengths, to see where the camera was sharpest. We did this for both kit lenses, the 14-42 mm wide angle, and the 40-150 mm telephoto.

We ran the images through Imatest, which calculated a value for resolution, measured in Line Widths per Picture Height (LW/PH). This value describes how many theoretical alternating black and white lines could fit across the image frame before becoming blurred. The E-510 was sharpest using the 40-150 mm telephoto lens, at an ISO of 100, an aperture of f/9.0, and a focal length of 40 mm.

Click to view the high-resolution image

The 10-megapixel E-510 scored well, capturing 1646 lw/ph horizonally with -14 percent undersharpening, and 1589 vertically with -15.5 percent undersharpening. Undersharpening is advantageous for DSLRs, because it enables users to post-process the files to their specifications and control artifacts that emerge from oversharpening. However, the E-510 pretty severely undersharpened the images. We would have liked to see a little more sharpening applied, particularly since much of the camera's targeted demographic often favor usable JPEGs right out of the camera.  

Interested in which of the two kit lenses is better to use for medium focal lengths? The answer is clear: the telephoto lens produced images with much more information and fine detail, scoring far better than the wide angle images shot in the same setup. Use the telephoto kit lens as much as you can with this camera!

Noise – Auto ISO (4.45)
We photographed a GretagMacbeth color chart at auto ISO, and calculated the noise levels with Imatest. Pointed at our brightly-lit chart, the E-510 selected ISO 400 for the shot. This is an unusually high ISO to use under our bright studio lights, and risks ending up with an unnecessarily noisy image. However, with the Noise Filter set to Standard, the noise levels stayed quite low. Though this was a much different story with the Noise Filter turned off. The Noise Filter smoothes over noise at high sensitivities, resulting in parts of the image looking fuzzy. The tradeoff here is clear, turn on the Noise Filter to remove ugly noise, but sacrifice sharpness.

Noise – Manual ISO (11.73)
We also shot our test chart at all ISO sensitivities, to see how the noise levels varied over the whole range. The graph below shows the amount of noise measured by the percent of the image it obscured.

With the Noise Filter set to Standard or High, noise levels stayed very low and rose slowly at higher ISO settings. However, with the Noise Filter off, the noise levels were quite high, with a large jump from ISO 200 to 400. Using the Noise Filter sacrifices sharpness for noise removal, as described above.

Dynamic Range (6.91)
Dynamic range is a measure of how many tonal gradations a camera can reproduce from pure black to white. We measure dynamic range by photographing an industry standard backlit Stouffer test chart. The chart consists of a row of rectangles that are all slightly different shades of gray and range from brightest white to darkest black. We run the images through Imatest imaging software to measure the amount of the chart the camera can expose while retaining detail. The more rectangles the camera can reproduce, the better the dynamic range. We shoot the Stouffer chart at every ISO setting, and for the Olympus E-510, we tested three of the Noise Filter settings: Off, Standard, and High.

The graph above shows the dynamic range plotted at all ISO sensitivities. Noise levels have a strong impact on dynamic range, and as you can see, more noise reduction yielded higher dynamic range. The E-510’s dynamic range levels at low ISO settings are not very good, but the camera does manage to keep them from dropping very much at higher ISO settings. Unfortunately, if the Noise Filter smoothes over too much detail for your liking, turning it off will yield very poor dynamic range. This is due to the very high noise levels with the Noise Filter off.

Low Light (9.31)
To test the E-510’s performance in low light, we dimmed the studio lights and tested color and noise performance at 60, 30, 15, and 5 lux. This corresponds to a room softly lit with two lamps (60 lux), down to very low light that would cause you to squint (5 lux). We always shoot low light levels at the highest full resolution ISO sensitivity the camera offers, which in the case of the E-510 was ISO 1600, using the Noise Filter set to Standard (the default).

Low Light Tests
60 Lux	30 Lux
15 Lux	5 Lux

The Noise Filter kept the noise levels quite low, but the main issue here was the difficulty getting the E-510 to manually white balance correctly. The camera had enormous difficulty manually white balancing in low light, and this often resulted in a green cast in the photos. Because of this problem, the E-510 had trouble reproducing accurate colors in low light, having a mean color error of 10.2 at 5 lux.

We also took a look at low light performance from another angle, long exposures. For this test, we shot the ColorChecker at shutter speeds of one to 30 seconds at ISO 400. The graph below shows the noise levels at each shutter speed, with both Noise Reduction on and off.

Noise levels are admirably low, and rise slowly but steadily with increasing exposure length. The Noise Reduction appears to kick in with 20-second exposures, where the two lines on the graph start to separate. Color accuracy was maintained well in long exposures, ranging from a mean color error of about 8 at one second exposures, to about 9.8 at exposures more than 20 seconds. Noise levels also stayed quite low, but it is very important to note that with the Noise Reduction set to Off, there was an abundance of "hot" pixels throughout the images. This caused lots of little points of bright white to be speckled throughout images. These hot pixels were removed with the Noise Reduction turned On, so the lesson here is to always shoot long exposures on the E-510 with Noise Reduction on.

Speed/Timing - All speed tests were conducted using a SanDisk Ultra II 2.0GB Compact Flash card.
Startup to First Shot (8.8)
The Olympus E-510 took 1.2 seconds to start up and snap the first shot. While this may not seem terribly slow to users upgrading to a DSLR from a point-and-shoot camera, most DSLRs now keep it under 1/2 second.

Shot-to-Shot (9.7)
In Burst mode, the E-510 took 11 shots 0.3 seconds apart, and after the eleventh shot slowed down to shooting one shot every second. This 11 shot quick burst took a total of 3.5 seconds.

Shutter-to-Shot (9.0)
With the shutter held halfway down and prefocused, the camera took the shot instantly. Without being prefocused, it took 0.2 seconds to take the shot.

Processing (8.0)
The E-510 takes approximately one second to process a 7MB 3648 x 2736 JPEG image shot at ISO 100. However, while the camera is processing, the playback cannot be viewed.