Sony Alpha A550

Color (12.68)

The Sony A550 was reasonably accurate in reproducing color hues, but routinely produced oversaturated images in every available color mode, which brought down the overall score in this section. We test color accuracy by shooting the X-Rite ColorChecker chart under bright 3000 lux studio illumination in all of the camera’s color modes (in this case, these are called Creative Styles, and there are six in all). We use Imatest software to analyze the captured color compared to the known values of the industry-standard chart, and derive a score based on the camera’s most accurate color mode. For the A550, this was the Portrait mode, though Standard mode produced results that were very close to these values. More on how we test color.

In all of the color modes, light skin tones were reproduced very accurately, though dark skin didn’t fare as well. Purple and green were also handled well in Portrait and Standard modes. However, blues and reds were seriously off the target values. As you’d expect, the Landscape mode pushes blues and greens into unrealistically ‘pretty’ tones, Sunset pumps up the reds and yellows. Vivid mode doesn’t change color values much (except for blue tones), but pumps saturation up to over 125%, and Landscape and Sunset modes are about the same when it comes to flagrant oversaturation, As for Standard and Portrait modes, they each come in at around 110% saturation. It’s interesting that, when we tested the Sony A900 some months ago, we found similar color reproduction patterns with these same Creative Styles, but that camera also included a Neutral mode that was far more accurate. Presumably, someone at Sony decided that for this more consumer-oriented SLR, the less sophisticated buyer would welcome more pumped-up colors.

Camera Color Comparisons					Expand
	Ideal	Sony Alpha A550	Canon EOS Rebel T1i	Nikon D5000	Olympus E-30	Panasonic Lumix DMC-GF1
Dark Skin
Light Skin
Blue Sky
Foliage
Blue Flower
Bluish Green
	Ideal	Sony Alpha A550	Canon EOS Rebel T1i	Nikon D5000	Olympus E-30	Panasonic Lumix DMC-GF1
Orange
Purplish Blue
Moderate Red
Purple
Yellow Green
Orange Yellow
	Ideal	Sony Alpha A550	Canon EOS Rebel T1i	Nikon D5000	Olympus E-30	Panasonic Lumix DMC-GF1
Blue
Green
Red
Yellow
Magenta
Cyan

NOTE: Because of the way computer monitors reproduce colors, the images above do not exactly match the originals found on the chart or in the captured images. The chart should be used to judge the relative color shift, not the absolute captured colors.

The Sony A550 and Panasonic GF1 each produced so-so color reproduction performance, with the GF1 actually less accurate when it came to color values but without the Sony oversaturation problem. As shown here, the Canon, Nikon and Olympus all delivered significantly more accurate color.

Color Modes (4.00)

The A550 offers six Creative Style settings (the A900 has thirteen). These settings affect color reproduction in addition to contrast, saturation and sharpness (these last three can be adjusted, as explained in detail the Picture Effects section of this review). Shown below are sample swatches taken from our test shots of the X-Rite ColorChecker chart in each Creative Style mode (except black and white). The ideal chart value is shown in the leftmost column. The color names are those used by X-Rite.

Color Mode Comparisons						Expand
	Ideal	Portrait	Standard	Vivid	Landscape	Sunset
Dark Skin
Light Skin
Blue Sky
Foliage
Blue Flower
Bluish Green
	Ideal	Portrait	Standard	Vivid	Landscape	Sunset
Orange
Purplish Blue
Moderate Red
Purple
Yellow Green
Orange Yellow
	Ideal	Portrait	Standard	Vivid	Landscape	Sunset
Blue
Green
Red
Yellow
Magenta
Cyan

NOTE: Because of the way computer monitors reproduce colors, the images above do not exactly match the originals found on the chart or in the captured images. The chart should be used to judge the relative color shift, not the absolute captured colors.

White Balance (8.38)

The A550 did reasonably well in our automatic white balance testing, but taking a custom WB reading didn’t produce the level of precise color reproduction we expect from custom white balance systems, leading to a mediocre overall score. We test white balance accuracy by shooting the ColorChecker chart inside the X-Rite Judge II lightbox, which produces consistent color temperatures for incandescent, compact fluorescent and daylight illumination. More on how we test color.

Automatic White Balance (11.09)

The automatic WB system was particularly adept at handling often-tricky fluorescent lighting, and did almost as well in daylight. Only incandescent illumination, the tungsten-bulb light source that so often produces overly warm, orange-hued photos, tripped up the A550.

Custom White Balance (5.67)

The color errors we measured after taking a custom white balance reading aren’t extreme, but they are larger than we like to see in an SLR, where accuracy standards in this mode are demanding.

The level of color error during daylight shooting is unlikely to be noticeable to the naked eye in your photos. With the noteworthy exception of the Panasonic GF1, none of the comparison cameras handled incandescent lighting particularly well. The A550 has the honor of being less bad than the others here. Fluorescent lighting can be tough on an automatic white balance system, but the A550 handled the challenge well, particularly compared to the Nikon D5000 and Olympus E-30.

The overall white balance score for the Sony A550 is a bit below the competition, owing largely to the mediocre custom white balance results.

White Balance Options (7.50)

In addition the Auto WB mode there are six white balance presets: Daylight, Shade, Cloudy, Incandescent, Fluorescent, and Flash.

Each of these white balance presets can be adjusted, with three available steps toward a more reddish tone and three toward a more bluish tone. However, since you can’t see the effects of this tweaking while you’re changing the setting, it’s not likely to get much use.

Taking a manual exposure reading is simple enough, once you figure out the slightly obscure series of controller presses required to actually take the reading. Two additional white balance adjustment methods are available; enter a color temperature value in degrees Kelvin, or choose a virtual Color Compensation filter, selecting one of nine values for either green or magenta. Again, since there’s no visible feedback until you’ve actually taken and reviewed your photo, we don’t expect these options to be very popular.

White balance bracketing is also available, as a drive mode setting. A single image is shot and stored in three separate JPEG files, each with an adjusted white balance value. The user can choose between having a large (20 mired) or small (10 mired) adjustment value applied.

Long Exposure (9.30)

Our long exposure test weighs two component factors, image noise and color accuracy. The Sony A550 did well when it came to noise and less well on color accuracy (which makes sense, since even under ideal conditions its color reproduction isn’t great). Overall, this produced a decent but not exceptional score.

The long exposure test procedure involves shooting the X-Rite ColorChecker chart under subdued lighting, at shutter speeds ranging from 1 second to 30 seconds, with any long exposure noise reduction system set both on and off. The resulting test images are run through Imatest to determine image noise levels and color accuracy. More on how we test long exposure.

Color error remained nearly constant, at a relatively high level, across all tested shutter speeds, and saturation ran about 108% across the board. Long exposure noise reduction had no effect on the results.

The image noise levels didn’t vary much with shutter speed either, remaining at under 1% throughout. The long exposure noise reduction system had little effect here either, which is a result we often see — in fact, it’s not uncommon for long exposure noise reduction to make matters worse. The concept is to have the camera take a second exposure with the shutter closed, analyze the noisy areas in this dark exposure and then mathematically remove these spots from the original shot. However, since the overwhelming majority of image noise is random, this approach rarely produces much improvement and, what’s more, doubles the time before you can take another shot.

The two stragglers in our comparison group, the Olympus and Panasonic, have something in common: they both use smaller Four Thirds / Micro Four Thirds sensors, which are subject to higher image noise than the larger APS-C format sensors found in the other three cameras. The Sony didn’t match the performance of the Canon Rebel T1i or Nikon D5000, but it’s not far behind.