Pentax K10D Digital Camera Review$799.00
Read an unbiased, expert review of the Pentax K10D digital camera.
Testing / Performance
*Color accuracy has a huge impact on the quality of photographs, and good accuracy allows more flexibility when post-processing photos. We test this by photographing an industry-standard GretagMacbeth ColorChecker test chart, which consists of 24 color tiles from around the color spectrum. We determine color accuracy by comparing the colors the camera reproduces with the known colors of the test chart. The image below shows how well the Pentax K10D’s colors match up to the ideal colors of the ColorChecker chart. The outside squares show the colors the K10D reproduces, the inner squares show the ideal colors of the ColorChecker corrected for the luminance, and the inner rectangles show the ColorChecker colors under a perfectly even exposure.
Comparing the outer squares with the inner squares shows a number of the color tiles match up quite well. However, several colors differ significantly, especially the blues, yellows, and oranges. This information is shown in a more quantitative way in the graph below. The ideal colors are represented by squares and the Pentax K10D’s colors by circles. The lengths of the lines connecting the squares and circles show the amount of color error for each tile.
The graph shows many of the color tiles are quite accurate, though there is significant drift in yellows and blues. This is actually quite common in digital cameras, since shifting blues toward purple and yellows toward green can enhance certain photos, such as landscapes with blue skies and green foliage. However, the amount the blues are shifted in the K10D is a bit worrisome; it can make certain blue skies look purple. Overall, the color accuracy of the K10D is decent, but not as spot-on as we’ve seen from competing DSLR manufacturers, such as Canon and Olympus.
We test resolution performance by photographing an industry-standard resolution test chart and varying focal length, aperture, and shutter speed. We run the photos through Imatest, which determines resolution in terms of line widths per picture height (lw/ph), the number of equally-spaced, alternating black-and-white lines that can fit across the image frame before becoming blurred. Imatest also evaluates the amount of sharpening that has been applied to an image by the camera. Too much oversharpening, especially in a DSLR, is undesirable because it destroys image information. On the other hand, too much undersharpening is also undesired because it doesn’t maximize the resolution of the camera and may require post-processing. We tested the K10D using both the SMC Pentax DA 18-55mm F3.5-5.6 AL kit lens and a SMC Pentax DA 50-135mm F2.8ED [IF] SDM lens. Because of the number of substantially different RAW image processing software options available, we test resolution with cameras set to highest quality JPEG so we can compare cameras to each other in a more standardized, accurate way.
The K10D proves sharpest using the 50-135mm f/2.8 lens set to ISO 100, f/6.3, and a focal length of 70mm. The 10-megapixel camera resolves 1343 lw/ph horizontally with 15.5 percent undersharpening, and 1613 lw/ph vertically with 12.8 percent undersharpening. What jumps out about these numbers is the significant amount of undersharpening applied. In fact, most settings produce images with even more undersharpening, with both the wide-angle and telephoto lenses. The amount of undersharpening means the K10D does not produce photos right out of the camera as sharply as it could. Users who want to maximize resolution must sharpen their photos through post-processing. That said, it is good the camera doesn’t oversharpen, which causes ugly image artifacts. The shots are sharp from edge to edge and show little to no sign of "ghosting" or jaggedness. Overall, however, the K10D’s resolution performance leaves something to be desired.
Noise – Manual ISO***(10.13)*
Image "noise" refers to the grainy or splotchy patterns that appear in digital images, especially at higher ISO speeds. Unlike film grain, digital image noise is almost always ugly and unwanted, as it obscures fine image detail. It is an unavoidable side effect of the digital electronics, similar to static in TVs or the background hiss in stereo systems. Cramming more megapixels onto camera sensors intrinsically increases noise levels because the pixels must be made smaller to fit on the same sized sensors. Smaller pixels have worse signal-to-noise ratios than larger pixels. One of the major engineering challenges in the digital camera industry is to make cameras with more megapixels and better resolution but still keeping noise levels low.
We test noise levels by photographing our test chart under bright, even studio lights at all of a camera's ISO speeds. We run the photos through Imatest, which measures noise levels in terms of the percentage of image detail it drowns out. The K10D does a very good job keeping noise levels low up to ISO 640. Above ISO 640, noise levels are higher. The noise itself is quite typical of digital cameras; small splotches of colors and fine gray grains. The color noise is especially distracting in high ISO images. Yet admirably, the camera keeps images quite sharp at high ISO speeds, meaning it doesn’t attempt to smooth over image noise significantly. We performed the test with the Noise Reduction setting in the Custom menu set to both On and Off. There was no difference between the two settings, which is good because the Noise Reduction is only supposed to kick in for long exposures. Overall, the K10D does a solid job handling noise levels, but not quite as well as some of its competitors. It is slightly noisier than the 6-megapixel Pentax K100D.
Noise – Auto ISO*(6.62) *
We also test noise levels with cameras set to Auto ISO, under the same bright studio lights as the Manual ISO noise test described above. The K10D chose ISO 200, producing only a small amount of noise. You can trust this camera to keep ISO settings reasonably low when set to Auto ISO.
**Still Life Sequences
***Click to view the high resolution images
White Balance* (3.93)
*Color accuracy means nothing if a camera cannot white balance correctly. Every light source produces a different color cast, and cameras must be able to adjust accordingly. We test white balance accuracy by photographing the ColorChecker test chart under four types of light: flash, fluorescent, outdoor shade, and tungsten. We test the Auto white balance setting, as well as the appropriate white balance presets found in the Function menu. The K10D also offers a Manual white balance setting.
*Set to Auto white balance, the K10D is mediocre using the flash, poor under white fluorescent light and tungsten light, and abysmal in outdoor shade. Normally we would suggest staying away from the Auto setting by using the presets, but unfortunately the presets aren’t much better.
*Set to the appropriate white balance presets, the K10D is accurate under white fluorescent light (with the Fluorescent "W" setting), but poor under flash, outdoor shade, and tungsten light. This creates obvious color casts, such as the deep blue cast to photos taken under tungsten light. The only ways to get good white balance accuracy with the K10D is to manually white balance with a white card or shoot in RAW and set the white point on your computer after you shoot.
We’ve seen how the K10D handles colors and noise levels in bright light; now let’s take a look at its performance in less-than-ideal shooting conditions. We test low light accuracy by photographing the ColorChecker chart at light levels of 60, 30, 15, and 5 lux. Sixty lux corresponds roughly to the amount of light fill a room lit softly by two table lamps, 30 lux approximates a room lit by a single 40-watt bulb, 15 lux is about as bright as a room lit only by a large television screen, and 5 lux is very low light that tests the limits of a sensor. All shots are taken at ISO 1600.
Color accuracy suffers significantly in low light, reaching a mean color error of 14.8 at 5 lux, compared to 7.8 in bright light. Colors are oversaturated, as well. At ISO 1600 noise levels are not too high, but are visible enough to notice even in scaled-down versions of the photos. Overall, the K10D tests well in our low light level test.
We also test the quality of images taken at long exposures, this time at ISO 400. The K10D can take exposures as long as 30 seconds. The camera has a bit of trouble exposing properly in long exposures, but keeps noise levels very low. Color accuracy did suffer a bit, however. We tested long exposures with the Noise Reduction setting in the Custom menu both on and off, though you can see from the graph that it doesn’t do much to the images. In fact, the photos with Noise Reduction on tend to have slightly more noise than with the setting on Auto.
**Dynamic Range ***(7.60)*
Dynamic range is an important image quality factor that describes the full tonal range of a camera. This is particularly helpful in high contrast scenes, where detail must be shown in bright highlights as well as dark shadows. Some examples where dynamic range is very important include wedding photography (white dress and black tux) and outdoor landscape or portrait photography in bright sunlight (bright highlights, dark shadows). We test dynamic range by photographing a backlit Stouffer test chart at all ISO speeds. The Stouffer chart consists of a long row of gray rectangles varying in tone from brightest white to darkest black. The more rectangles a camera can discern, the better its dynamic range.
The K10D has excellent dynamic range at ISO 100, and keeps it relatively high up to approximately ISO 800, where it starts to degrade significantly. Dynamic range is tied closely to noise levels, as noise is often the limiting factor when trying to show fine image detail at high ISO speeds. The K10D has worse dynamic range than the K100D, its less-expensive brethren, but beats out the more-expensive Olympus EVOLT E-510.
Speed/Timing – All speed tests were conducted using a Kingston Ultimate 120X 2GB SD Card, with the camera set to highest resolution and best quality JPEGs, unless otherwise noted.
*Startup to First Shot (9.7)
*The K10D takes only 0.3 seconds to turn on and fire its first shot. This is excellent.
In Continuous shooting mode, the camera fires every 0.3 seconds for approximately 50 shots, and then slows down a little, firing shots intermittently every 0.3 or 0.6 seconds.
*The K10D has no measurable lag, prefocused or not. The camera has an incredibly fast autofocus system.
It takes the K10D 1.5 seconds to process one 2.2 MB JPEG taken at ISO 200.