Color (3.68)
We test color accuracy by photographing an industry standard GretagMacbeth ColorChecker test chart, and running the images through Imatest. The ColorChecker is made up of 24 tiles of different colors from around the color spectrum. Many of the colors are what you would find in common scenes: grass greens, sky blues, and flesh tones. We determined the FE-250’s color accuracy by comparing the actual color of the chart with the color the camera reproduced. This is shown in the image below. The outside square is the color the camera reproduced, the inside square is the color of the chart corrected for luminance, and the tiny rectangle is the ideal color of the chart.
As you can see, almost every color tile is significantly different than what the camera reproduced. The following graph shows this in a more discrete way, by plotting each of the ideal chart colors (the squares), and each of the corresponding colors the camera reproduced (the circles). The graph’s background represents the entire color spectrum, and shows how the camera’s colors have strayed from the ideal chart colors. Saturation increases as colors get further from the center of the graph. The line connecting the squares and the circles represents the color error. The longer the line is, the more inaccurate the color.
This is an extremely poor color score, and is an example of what happens when white balance can’t be adjusted. Because the white balance is so off, every color the camera shoots will be shifted. As you can see, the blues have turned purple, the greens have turned yellow, and the yellows have lost saturation.
White Balance (4.49)
Auto (8.99)
Preset (0.00)
In addition to testing color accuracy, we test white balance accuracy by shooting the ColorChecker under different types of light. Normally we test both the auto white balance setting as well as the preset white balance settings under four different types of light: flash, fluorescent, outdoor cloudy, and tungsten. However, the FE-250 doesn’t have white balance presets, so we could only test the auto settings. Additionally, the FE-250 has no auto focus assist light, which prevented us from testing the white balance using flash. We typically test flash white balance without additional light shining on the ColorChecker so it doesn’t throw off the white balance. The absence of an auto focus assist light prevented us from focusing on the chart.
|
Auto WB - Cloudy Illumination
|
|
Auto WB - Fluorescent Illumination
|
|
Auto WB - Tungsten Illumination
|
The FE-250 had mixed results under the three different types of light. Under outdoor cloudy light it was very accurate, under fluorescent it was mediocre, and under tungsten it was terrible. The real problem is that if your images are coming out very strange colors, there is nothing you can do about it, because there is no way to change the white balance. This is readily apparent in our still life images further down the page. Notice that a couple of the images have a yellow cast to them. This is because the camera would sometimes change the white balance when the ISO was changed, and would stay that way no matter how many photos were taken. Get used to these sorts of “fun” surprises with the Olympus FE-250.
Still Life
Click to view the high-resolution images.
|
Still Life Scene
|
|
|
|
|
ISO 64
|
ISO 64
|
|
|
|
|
ISO 100
|
ISO 100
|
|
|
|
|
ISO 200
|
ISO 200
|
|
|
|
|
ISO 400
|
ISO 400
|
|
|
|
|
ISO 800
|
ISO 800
|
|
|
|
|
ISO 1600
|
ISO 1600
|
|
|
|
|
ISO 3200
|
ISO 3200
|
Resolution (4.94)
We tested the resolution of the 8-megapixel Olympus FE-250 by photographing an industry standard resolution chart under bright studio lights. We varied aperture, focal length, and exposure settings in order to find the camera’s “sweet spot”, i.e. the settings that produced the sharpest image. We quantify resolution in terms of Line Widths per Picture Height (LW/PH), which indicates the number of alternating black and white lines that could fit across the picture frame before becoming blurred by the camera.
Click to view the high-resolution image
Running the images through Imatest, we found that the FE-250 was sharpest at ISO 64, f/4.7, and a 22mm focal length. The camera resolved 1715 LW/PH horizontally with 11.6 percent oversharpening, and 1556 LW/PH with 0.76 percent oversharpening vertically.
These are not very impressive scores, but it is commendable that the camera uses an appropriate amount of sharpening to make the photos as sharp as possible out of the camera without causing undue image artifacting. However, you won’t want to make large prints of images taken with the FE-250.
Dynamic Range (5.87)
To test dynamic range - the range of tones a camera can reproduce - we photographed a backlit Stouffer test chart. The Stouffer chart consists of a row of rectangles that vary gradually in tone, from bright white to dark black. The more rectangles the camera can discriminate, the better the dynamic range. We shot this chart with the FE-250 at every full resolution ISO setting, in order to see its performance over the entire ISO range. The graph below shows our results, with dynamic range measured in Exposure Value (EV).
The dynamic range at ISO 64 is solid, at about 6.5 EV. It falls quickly from there, but then unexpectedly rises from ISO 800 to ISO 1600. Dynamic range is closely linked with noise levels, and the reason the dynamic range rises at ISO 1600 is because there is less noise at ISO 1600. But as we explain below, this drop is noise is caused by noise reduction that also smoothes over detail. In other words, just because dynamic range appears to rise at ISO 1600, it doesn’t mean your photos will look better than they do at ISO 800.
Low Light (4.17)
We dimmed the studio lights to test the FE-250’s performance in low light. We shot the ColorChecker at light levels of 60, 30, 15, and 5 lux, which correspond to a softly lit living room (60 lux), down to very low light that tests the limits of the camera. Admirably, the FE-250 could expose properly all the way down to 5 lux at ISO 1600. The problem came, as usual, with the color accuracy. The mean color error rose to an astronomical 33.4 at 5 lux with 130 percent saturation, which visually translated to a very strong yellow-pinkish cast over the images. This was undoubtedly enhanced by the camera’s trouble with white balance. Noise levels stayed low, but not without a loss of detail due to the heavy noise reduction described in the noise section.
|
Low Light Tests
|
 |
 |
|
60 Lux
|
30 Lux
|
 |
|
|
15 Lux
|
5 Lux
|
We also test long exposures in low light, but the FE-250 only takes exposures longer than half a second in certain scene modes, and then only up to 4 seconds. To standardize our testing and scoring, we always test long exposures at ISO 400, and none of the scene modes on the FE-250 will shoot at ISO 400. However, for ½ second exposures at ISO 400, the camera was very noisy, and had enormous color error and oversaturation. Overall, the FE-250 scored quite poorly in low light.
Noise – Auto ISO (1.48)
We shot our test chart under bright studio lights with the FE-250 set to auto ISO to see how it handled noise. The camera chose ISO 160, a seemingly reasonable choice because it shouldn’t produce too much noise. However, this was not quite the case. Even at ISO 160, the FE-250 still had significant noise levels - 1.59 percent of the image was victim to it.
Noise – Manual ISO (5.26)
Under the same studio lighting, we also shot our test chart at every full resolution ISO setting on the FE-250, so that we could see how it handled noise over the entire ISO range. The following graph shows our results, where noise is measured by the percentage of the image it covered.
Noise levels are nice and low at ISO 64, and then rise sharply up to ISO 800. To keep noise levels down, use this camera at ISO 64 whenever possible. Interestingly, noise decreases significantly from ISO 800 to 1600. This strongly suggests high sensitivity noise reduction being applied automatically inside the camera. The problem with noise reduction is that while it can remove a lot of splotchy noise, it also smoothes over a lot of detail, resulting in soft-looking images. Without noise reduction, the FE-250 would probably have extremely high levels of noise at high ISO settings. These are common issues for cameras that try to cram too many pixels onto their sensors; more pixels mean smaller pixels, and smaller pixels mean more noise.
Speed/Timing – All speed tests were conducted using a 256MB Olympus XD Picture Card.
Startup to First Shot (7.4)
The FE-250 took 2.6 seconds to start up and take a picture. This is quite a delay if you’re trying to capture an unexpected moment.
Shot-to-Shot (1.5)
The FE-250 will only shoot continuously with image size set as large as SQ1, which is a resolution of 2048 x 1536. This means the files are only 15 percent of the size of full resolution shots. At this resolution, the camera took 12 shots in 2.2 seconds, which translates to approximately one shot every 0.2 seconds.
Shutter-Shot (9.0)
With the shutter held down halfway and pre-focused, the camera take shots instantly. Without being pre-focused, it takes 0.7 seconds to take a shot.
Processing (2.6)
The FE-250 takes a very long 3.7 seconds to process one 2.75MB shot at ISO 64. Get used to waiting for that little red light to stop blinking.
Video Performance (3.13)
Bright Indoor Light – 3000 lux
We shot video of our ColorChecker lit evenly by our studio lights at 3000 lux. As in the still tests, the color accuracy and saturation were very poor, with a mean color error of 25.8, and saturation of 125.1 percent, as you can see in the color error diagram below.
Low Light – 30 lux
In low light, the results weren’t any better. Mean color error was 29.9, and saturation at 133.6. The good news is that noise levels stayed quite low, at only 1.3 percent. To put it simply, color accuracy on the FE-250 is terrible in both still images and video, and the lack of white balance control means you have to take what it gives you.
Resolution
We also tested video resolution by recording videos of our resolution test chart. While the FE-250’s video color was poor, its resolution was much better. On the standard 640 x 480 frame size, the camera captured 262 LW/PH horizontally with -11.3 percent undersharpening, and 453 LW/PH vertically with -0.5 percent undersharpening. These values may seem very low compared to the still image test, but that’s because video is recorded at a much lower resolution. The FE-250 scored about average compared to other point and shoots.
(100 percent crops)
Outdoor Motion
To get a look at how the FE-250 handled motion in video, we took it outside to shoot the hustle and bustle of city streets. The camera handled motion quite well, with only a little jerkiness caused by moving objects as they left the frame. The overall contrast and color looked quite good as well. The big problem with the FE-250’s outdoor video was that it kept shifting its exposure, meaning the screen would often flash bright or dark, even when there were no moving objects in the frame. This was extremely distracting, and negated the other good aspects of video performance.
