GE A830

Color (5.71)
If you were to take the same photo of your family in front of a landscape with several different cameras, each camera would reproduce the colors of the landscape and your family’s faces differently. Some cameras will saturate blues and greens, making landscapes more brilliant, but less realistic. Some cameras will adjust reds and yellows, making some faces more enlivened, and others appear flushed. We test color accuracy by photographing an industry standard GretagMacbeth ColorChecker test chart, and comparing the known colors of the chart with the colors the camera reproduces. The ColorChecker consists of 24 colored tiles, each representing a commonly photographed color or an example from around the color spectrum.

The image below shows how the GE A830’s colors compare to the actual colors of the test chart. The outside squares show the colors the camera reproduced, the inside squares show the actual colors of the test chart corrected for luminance, and the inner rectangles show the actual colors of the test chart under a perfectly even exposure.

If you look at the outer squares compared to the inner squares, you can see that almost every color is off. The yellows look green, the blues look purple, and almost every color appears duller than it should. The graph below shows this information in another way. The background of the graph shows the entire color spectrum, and the ideal colors of the test chart are shown as squares, while the camera’s colors are shown as circles. The length of the lines connecting the squares and circles shows the color error.

The graph confirms the significant color error seen in the image of the ColorChecker. Many colors drift away from their ideal values, especially yellows, greens, and blues. These colors will greatly affect landscape photos, making foliage look bluish, and skies turn purple. In addition to the color shifts, almost every color is significantly undersaturated, which will make your photos appear less vibrant.

The A830 has another issue with color that is rarely seen in production-level digital cameras. As you can clearly see in the still life images below, half of the image frame is a different color tint than the other half. This becomes more obvious when ISO sensitivity is increased. It is almost as if half of the frame was not white balanced, while the other half was. This is a huge problem, and one that GE needs to address. It makes shooting at ISO sensitivities above 400 almost impossible.

Resolution (8.10)
We test resolution by photographing an industry standard resolution test chart at varied focal lengths and exposure settings. We then run the images through Imatest, which determines resolution in terms of line widths per picture height (lw/ph), which correspond to the number of equally spaced alternating black and white lines the camera can discern before blurring them. The 8-megapixel A830 captured its sharpest images at ISO 80, f/4.8, and a focal length of 18mm. At these settings, resolution was impressive, recording 1880 lw/ph horizontally with 2 percent oversharpening, and 1633 lw/ph vertically with 13.7 percent undersharpening. The vertical undersharpening is a bit of a concern however, meaning some high contrast lines will look soft. Also, the A830 was very inconsistent. The resolution at slightly different exposure settings was very different from the resolution recorded above, so you can only expect this camera to occasionally produce really sharp photos.

Click to view the high-resolution image

Additionally, the A830 has significant chromatic aberration ("color fringing") and blurring at the edges of its photos. You may get some sharp photos of your subjects in the center of the frame, but the edges will be blurry and artificially-colored.

Noise – Manual ISO (2.08)
Like all electronic devices, digital cameras produce signal noise. The noise is created inside a camera’s sensor, and has nothing to do with the scene being photographed. Image noise appears as speckled dots or splotches scattered randomly throughout a photo. Sometimes the noise is monochromatic, and sometimes it is colored – and almost always unwanted. We test noise levels by photographing our test chart under even studio lab lighting at different ISO sensitivities.

The graph above shows the noise levels, measured in the percentage of the image the noise drowns out, at each ISO setting. Noise intrinsically escalates as ISO is increased, because boosting the ISO amplifies the signal. In the GE A830, noise levels increase up to ISO 400, but then drop way off at ISO 800. As evident on closer inspection of the photos, this decrease in noise is caused by in-camera noise reduction. This noise reduction smoothes over the noise, making it less grainy, but drastically reduces image detail. This noise reduction does not make photos look better, and is an attempt to hide an incredibly noisy camera sensor. Click on the still life images below to see the effects of noise and smoothing.

Noise – Auto ISO (0.94)
We also evaluate the noise levels of cameras set to Auto ISO. Under bright studio lights, the A830 fired at ISO 100, with 2 percent of the image lost to noise. This is an unacceptably high amount of noise in such bright light, and results in a very poor auto noise score.

White Balance (7.03)
Accurate white balance is very important to proper color reproduction in photos. Poor white balance can result in color casts to photos -- sometimes obvious, sometimes subtle. If you’ve owned a camera before, you’ve probably noticed that some of your indoor photos looked extraordinarily yellow. This is a result of inaccurate white balance.

We test white balance by photographing the ColorChecker test chart under four different types of lighting: flash, fluorescent, outdoor cloudy, and tungsten. We test both the Auto white balance setting and the appropriate white balance presets, found in the Function menu.

Auto (8.65)
White balance accuracy using the Auto setting is mixed. Using the flash, white balance is extremely accurate, which is great because there is no flash preset. Under outdoor shade lighting the A830 is also fairly accurate, but under fluorescent and tungsten light the accuracy is very poor.


Preset (5.41)
Under outdoor shade light, the preset is as accurate as the Auto setting. However, under fluorescent and tungsten light the accuracy is very poor, though slightly better than the Auto setting. Overall, it makes more sense to keep the camera on the Auto setting, since worrying about setting the presets is not worth the tiny advantage it gives in only a couple lighting situations.

Fluorescent WB - Fluorescent Illumination

Shade WB - Shade Illumination

Tungsten WB - Tungsten Illumination

Still Life Sequences
Click to view the high-resolution image.

Still Life Scene
ISO 80	ISO 80
ISO 100	ISO 100
ISO 200	ISO 200
ISO 400	ISO 400
ISO 800	ISO 800
ISO 1600	ISO 1600

Low Light (6.83)
We’ve seen how the A830 performs under bright studio lighting, now let’s see how the image quality holds up when the lights are dimmed. To test low light performance, we photograph the ColorChecker test chart at 60, 30, 15, and 5 lux and examine the noise levels and color accuracy. 60 lux corresponds is about the amount of light needed to comfortably read a book, 30 lux approximates the light from a single 40 watt bulb, 15 lux is approximately how bright a room gets when illuminated by a television, and 5 lux is as dark as a room lit by a child’s night light. All shots were taken at ISO 1600.

Low Light Tests
60 Lux	30 Lux
15 Lux	5 Lux

As you can see in the images above, the A830 cannot properly expose at 5 lux, showing that the camera has clear limits. At 60, 30, and 15 lux the color accuracy was very poor, and noise levels quite high. Colors in low light are very oversaturated, and stray far from their ideal values.

We also test long exposure performance. The A830 has a Slow Shutter option at the bottom of the Photo menu. In low light shots in general, the camera had an extraordinarily tough time autofocusing. The A830 can only autofocus correctly about 40 percent of the time when using slow shutter speeds. The Slow Shutter option is a fun addition, but there is also no light meter available in it, so the user has no idea how well exposed the subject is going to be. In long exposures, color accuracy is poor and noise levels are high. There will be very few instances where the A830’s Slow Shutter option will be useful.

Dynamic Range (4.12)
Our dynamic range test evaluates the range of tonal values a camera can discern. Good dynamic range is very helpful for maintaining details in scenes with very bright and dark areas. For instance, a camera with good dynamic range will show detail in both the brightly illuminated side of the building as well as in the dark shadows. A camera with poor dynamic range will blow out the illuminated side and miss the details in the darker areas.

We test dynamic range by photographing a backlit Stouffer test chart, which consists of a row of rectangles, each a slightly different shade of gray. The rectangles range from bright white on one side to dark black on the other. The more rectangles the camera can discern, the better its dynamic range.

At low ISO sensitivities, the A830 has less dynamic range than other budget cameras. As you can see in the graph, dynamic range decreases from ISO 80 to 200, but then increases at ISO 400 and again at ISO 800. This is the result of the noise reduction, as we discussed in the noise section above. Dynamic range is limited by noise levels, and smoothing the noise can help dynamic range slightly, though not enough to make a visible difference. Even with the noise reduction boost, the A830 had one of the worst dynamic range scores so far this year.

Speed/Timing – All speed tests were conducted using a Kingston Ultimate 120X 2GB SD Card
Startup-Shot (6.0)
The A830 takes 4.0 seconds to startup and take its first shot. After pressing the shutter, there is an odd 1-second pause before the screen turns on. This startup time is very slow, and could be very frustrating if you are trying to capture an unexpected shot.

Shot-to-Shot (4.4)
Like other GE cameras, the Burst mode in the A830 is quite confusing. There is a Continuous Shot menu in the Photo menu, but there is also a Continuous Shot mode available by pressing the drive button three times. The camera can be set to take continuous shots by accessing the Continuous mode through the drive button, even when the Continuous Shot menu is set to Off. When set this way, the camera takes four shots, each 0.4 seconds apart, then waits 2.5 seconds and shoots photos continuously 1.1 seconds apart until the card is filled. With the Continuous Shot mode set to 5 Shot, the camera takes 5 shots in 3 seconds, and when set to 5 Last, it takes shots every 1.3 seconds, but only records the last 5. In 5 Last mode, the camera has problems recording all the photos and often says "PICTURE ERROR" for at least 1 photo in every 5-shot burst. Also, the camera can only shoot continuously in Full Quality mode, not Best Quality.

Shutter-Shot (8.8)
The A830 takes 0.1 seconds to fire a shot when the shutter is held halfway down and prefocused, and 1.25 seconds when not prefocused. Try to prefocus your shots if you are waiting for a perfect moment to come along.

Processing (2.4)
A photo taken at 8 megapixels, Best Quality, ISO 80, and 5.5 MB in size takes 3.8 seconds to process. This is quite slow compared to other compact digital cameras.

Video (1.54)
Bright Indoor Light – 3000 lux
We tested the color accuracy of the A830 in Video mode by recording video of our color charts under bright studio lights using Auto white balance. As you can see below, the color error and saturation are horrendous, but this is actually common for digital camera video under tungsten light using the Auto white balance setting. Noise, however, stayed nice and low.


Low Light – 30 lux
We dim the studio lights to test the video performance in low light. Here the A830’s color accuracy was more accurate, though still far from ideal. Noise levels were higher than in bright light, and noise is certainly visible in the footage.


Resolution
We shoot footage of our resolution test chart to test the sharpness of the video. The A830 recorded 190 lw/ph horizontally with 24.5 percent undersharpening, and 845 lw/ph vertically with 77.1 percent oversharpening. These sharpening levels are outrageous, and cause abundant image artifacting. White "ghosting" lines that shouldn’t even be there are clearly visible next to the dark edges on the resolution chart. These ridiculous low scores are also due in part to problems the camera has focusing while shooting video.


Outdoor Motion
We evaluate how cameras capture motion outdoors by bringing them down to the street to shoot footage of moving cars and pedestrians. This is where the problems with the A830’s video are most obvious. Motion is very jerky, there is abundant moiré, the exposure flashes and changes on a whim, highlights are blown out and smear across the frame, and the whole footage looks washed out, undersaturated, and very purple. Video footage with the A830 is altogether embarrassing, so don’t buy this camera expecting anything from the Video mode.