Olympus E-20Olympus updates their bargain-priced Pro SLR with a 5 megapixel sensor and improved electronics
<<Reference: Datasheet :(Previous) | (Next): Print-Friendly Review Version>>
E20 Sample ImagesReview First Posted: 11/28/2001
|We've begun including links in our reviews to a Thumber-generated index page for our test shots. The Thumber data includes a host of information on the images, including shutter speed, ISO setting, compression setting, etc. Rather than clutter the page below with *all* that detail, we're posting the Thumber index so only those interested in the information need wade through it!|
Portrait: (2192 k)
The extreme tonal range of this image makes this a tough shot for many digicams, which is precisely why I set it up this way, and why I don't use fill-flash on it. The object is to hold highlight and shadow detail without producing a "flat" picture with muddy colors, and the Olympus E-20N handled the challenge pretty well (though it required a contrast adjustment to get the best midtones while holding highlight detail). In addition to snapping sample images with the camera's Auto (2328 k) and Manual (2328 k) white balance settings, I also shot with the 4,500 (2372 k), 5,500 (2393 k), and 6,500 (2330 k) Kelvin temperature settings, ultimately choosing the 5,500 K setting as the most accurate. The Manual and 6,500 K settings both resulted in warm images, though with different color casts, while the 4,500 K setting produced a cooler image. Auto white balance produced nearly accurate color, but was slightly warm.
The shot at right required no exposure adjustment, though I did lower the contrast, which brought out more midtone detail. Skin tones look about right, but the blue flowers are slightly dark with a purplish tint (this is a difficult blue for many digicams). The overall color seems just a bit muted, but seems pretty correct in the hues, leading me to wonder if perhaps the E20 uses something other than the typical sRGB color space. - The slight undersaturation is good news for pros who'd want to work in a managed color environment, since it means the camera will be capable of capturing a wider than average color gamut. No promises (things are more than usually hectic as I write this), but I hope to revisit these images after having generated an ICC color profile for the E20, to see how it corrects the color. (Given the understated color saturation we see here though, I'm expecting great things from this...)
Resolution looks great, with a lot of fine detail visible throughout the frame. Shadow detail also looks good, with low noise.
To view an exposure series from zero to +0.7 EV, see files E20OUTLCWB55P0.HTM
through E20OUTLCWB55P2.HTM on our thumbnail page.
Closer Portrait: (2256 k)
Results are similar to the wider shot above, and the 4x zoom lens helps prevent any distortion of the model's features. Color is again just slightly understated, but quite accurate. Excellent detail in the Marti's face and haireven the tiniest hairs and lashes are visible. Details are clear and fairly sharp in the shadows, although there's a tad more noise in the red and blue channels in the shadow areas that I'd like to see. In-camera sharpening is just about right, bringing out detail, without creating obvious halos. (The camera appeared to be focusing toward the front of Marti's face, and depth of field on the E-20 is a bit shallower than many digicams, due to the larger sensor size. The result is that some of her hair is a bit softer, but I think that's more a function of depth of field, rather than a lack of sharpness.) The main shot was taken without any exposure adjustment, though I again shot with a low contrast setting.
See files E20FACLCP0.HTM through E20FACLCP1.HTM on the thumbnail
page to view the exposure series (not really a series, I guess,
just two shots, zero to +0.3 EV).
Indoor Portrait, Flash: (2368 k)
Dim intensity with the normal exposure, and a yellowish color cast.
The E-20N's flash is rather dim at its normal intensity (2580 k) setting, though the lighting is very even. I increased the flash intensity to +1.3 EV (2368 k), which brightened the image considerably, producing better color as well. The background household incandescent lighting produces a yellowish cast in both images, stronger in the unadjusted shot, but overall color balance is very nice, with just enough of a tint from the household lighting to suggest the warmth of the original scene without leaving an objectionable color cast.
Finally, I attached a stronger, external flash unit to the E-20N (nothing fancy, just my SunPak cheapie), which produced in the best image overall, as I could bounce the light from the ceiling, with some front-lit diffusion, producing a very even wash of light over the subject. The result was better color and a much brighter image. (This underscores why I think provision for an external flash should be considered a mandatory feature for any "enthusiast" or pro camera.)
See files E20INF1P0.HTM and E20INF1P4.HTM on the thumbnail
Indoor Portrait, No Flash: (2585 k)
Accurate color with the Manual white balance setting.
This shot is always a very tough test of a camera's white balance capability, given the strong, yellowish color cast of the household incandescent bulbs used for the lighting, and the E-20N's Manual (2461 k) white balance performed quite well (Slightly greenish, but not bad at all). The Auto (2455 k) white balance resulted in a very warm image, with a strong orange tint. Because the camera offers a range of Kelvin temperature settings in place of an Incandescent white balance mode, I shot with the 3,000 K (2509 k) setting, which resulted in a slight warm cast, though not as strong as with the Auto setting. (Illustrating why I think Kelvin white balance settings should really extend down to 2400 or 2500K.) I chose the Manual setting for the main selection, with an exposure compensation adjustment of +1.0 EV (2585 k). (Click here (2441 k) to see an example with no exposure compensation.) Skin tones look good (though slightly pale), and the blue flowers are dark and purplish.
The E-20 has ISO options ranging from 80 to 320, as shown in the series in the table below. As usual, image noise increases sharply as the ISO is raised. (Interestingly, in these shots, the red channel shows the lowest noise, while the blue and even green are much higher.) You'll have to make up your own mind about acceptable noise levels, but even the ISO 320 doesn't look too bad to my eye. That's pretty relative though - Compared to other pro cameras (with much larger sensors) like the Nikon D1X, the noise levels here are quite high.
House Shot: (3534 k)
Excellent resolution, with good color.
I shot this target with the Auto (3518 k),
Daylight (3523 k), and Manual
(3534 k) white balance settings, choosing the Manual setting as the
most accurate. Surprisingly, the Auto setting produced a very warm image,
with a strong yellow cast in the white trim, while the Daylight setting
produced nearly accurate results. The Manual setting resulted in a slightly
cool image, but I felt that the white value of the house trim was more
true. Excellent resolution, with strong detail in the fine foliage surrounding
the house. Details are just slightly soft throughout the frame (seemingly
from Oly's understated in-camera sharpening), but the overall image
Far-Field Test (3527 k)
This image is shot at infinity to test far-field lens performance. NOTE that this image cannot be directly compared to the other "house" shot, which is a poster, shot in the studio. The rendering of detail in the poster will be very different than in this shot, and color values (and even the presence or absence of leaves on the trees!) will vary in this subject as the seasons progress. In general though, you can evaluate detail in the bricks, shingles and window detail, and in the tree branches against the sky. Compression artifacts are most likely to show in the trim along the edge of the roof, in the bricks, or in the relatively "flat" areas in the windows.
This is my ultimate "resolution shot," given the infinite range of detail in a natural scene like this. The E-20N captures great detail throughout the frame, with good definition in the tree limbs above the house and in the shrubbery in front. Details initially look slightly soft (Olympus low-key in-camera sharpening), but what's there is very finely & delicately rendered. Comparing this to another top 5 megapixel camera, the Sony F707 (with Zeiss optics), the detail here looks much more delicate and precise, to my eye, at least. (Here's a link to same shot taken with the Sony model, check the detail in the bush just to the right of the central bay window, to make the comparison.) The bright sunlight tricks the E-20N slightly, as the camera loses all but the starkest details of the white bay window area, even in the main shot above, which has the contrast dialed down somewhat. The shadow area above the front door shows stronger detail, however, with a well-defined brick pattern above the door. I shot with a low contrast setting, as the camera's default exposure tends to have too much contrast, and boosted the exposure compensation to +0.3 EV. The table below shows our standard resolution and quality series, followed by ISO, Contrast, and Sharpness series.
Lens Zoom Range
A slightly longer-than-average 4x zoom range.
I've had many requests from readers to take shots showing the lens focal length range of those cameras with zoom lenses, so now routinely include the following series of shots, showing the field of view with the lens at full wide angle, as well as at full 4x telephoto. The E-20N's lens covers a range equivalent to a 35-140mm zoom on a 35mm film camera. Following are the results at each zoom setting.
Musicians Poster (3747 k)
Nearly accurate color, with great resolution.
For this test, I shot with the Auto (3754
k), Daylight (3747 k) and Manual
(3733 k) white balance settings, resulting in a slight dilemma over
which setting to choose for our main selection. None of the shots were
spot-on for color: The Manual white balance setting produced a more
accurate color balance in the blue background, but the skin tones were
a bit pale. The Daylight image has a warmer color balance, but the background
has a strong reddish tint. (The Auto setting was much too warm to even
be considered.) In the end, I chose the Daylight setting, despite the
slightly warm cast, as I felt the skin tones were the most accurate.
Resolution is very high, with outstanding detail in the embroidery of
the blue robe, as well as in the beaded necklaces and musical instruments.
(Really, the E-20 outstrips the capabilities of the poster that's the
Macro Shot (3700 k)
Not the smallest macro area I've seen, but great resolution, color, and detail.
The E-20N performs well in the macro category, though I expected a
slightly smaller minimum area than the 2.92 x 2.19 inches (74.29 x 55.72
millimeters) that the camera captured. Still, the camera does a great
job here, with very high resolution. Details are clear and sharp in
the printing of the dollar bill as well as the coins, though a limited
depth of field softens the details of the brooch. I did notice slight
corner softness from the lens. (A common problem in tight macros like
this, unless you have a dedicated fixed focal-length macro lens to work
with.) Color looks about right, though just a bit warm. The E-20N's
flash (3808 k) throttled down fairly well
for the macro area, though it remained just a bit bright.
"Davebox" Test Target (3838 k)
Saturation a little low, but good color accuracy. (Good for color-managed workflows?)
I shot samples of this target with the Auto (3622 k), Daylight (3615 k), and Manual (3838 k) white balance settings, noticing the most accurate color with the Manual white balance. Daylight and Auto both resulted in warm images with yellowish color casts. Exposure looks good, as the E-20N picks up the subtle tonal distribution of the Q60 chart. The large color blocks are nearly accurate, though saturation is a bit low. The shadow area of the charcoal briquettes shows a lot of detail, with fairly low noise. Following are Contrast and ISO series. (Again, I personally think the default contrast of the E-20N is a bit high, although I shot all these with no contrast adjustment applied.)
Excellent performance, with very low noise.
The E-20N's full manual exposure control and maximum (timed) shutter speed of 60 seconds resulted in excellent low-light performance. (Bulb exposures can run as long as 8 minutes, although we'd expect even the low noise levels of the E-20 to result in unacceptable image noise with that long an exposure.) The camera captured bright, clear images all the way down to the lower limit of our test setup, 1/16 foot-candle (0.67 lux) at all three ISO settings (even ISO 80). Color balance was slightly warm with the Auto white balance setting, at all the light levels I tested, but didn't stray notably at the very low levels, as many cameras do. Noise levels remained low throughout the series, even at the ISO 320 setting. I shot sample images without the camera's Noise Reduction system at the 1/16 foot-candle light level, and frankly noticed only subtle differences. (Click here to see the shots without Noise Reduction for the 80 (3477 k), 160 (3759 k), and 320 (3651 k) ISO settings.) The table below shows the best exposure I was able to obtain for each of a range of illumination levels, at each of the ISO settings. Images in this table (like all of my sample photos) are untouched, exactly as they came from the camera.
REALLY Low-Light Test
I've gotten a lot of interest from readers lately about very long time exposures with the latest digicams. Advances in sensor technology and noise reduction algorithms mean that time exposures of more than a minute in length are actually feasible with many cameras.
In bulb exposure mode, the E-20N can get out to some really long exposures, and Oly's use of the die-cast camera body and frame as a heat sink for the CCD helps keep noise to a minimum. The noise reduction feature is also very effective at eliminating "hot" pixels from the final image.
The shot below was captured in my backyard on a moonlit night, with an exposure time of 122 seconds at f/3.6 and ISO 80. (Click either thumbnail to see the original image, exactly as it came from the camera - note that it's a 3.6MB download though.) This is the same general shot I've used occasionally in the past for very long time exposures. It will vary enormously from camera to camera, with phases of the moon, etc. The point though isn't to show a real qnantitative comparison, but rather to get some idea of what the camera's noise level looks like on very long exposures. In this case, the moon was brighter than it often is, so I stopped the lens down to f/3.6, to get a longer exposure without blowing out a lot of the image. The backyard is also different here, as we've added a fence around it. (The dog toys on the cluttered lawn are clues as to why the fence was needed.;)
Another uncontrolled variable in these shots will be camera temperature. This one was taken in winter, albeit during a relative heat wave. Air and camera temperature were a relatively cool 50-55 degrees F or so. In the summertime, you'd see a lot more noise here, with the camera temperature more like 80-90 or so.
Overall, this is pretty clean. As noted above, there's essentially no "hot" pixels, even with a two minute exposure. (Note though, that this could easily change at higher camera temperatures.) Random image noise is likewise pretty low, although some cameras have done slightly better under similar conditions. (The Nikon 5000 did slightly better on this same shot, taken the same night this one was.) Still, this was a pretty impressive performance.
A note on this shot
Flash Range Test
Olympus' estimations of flash range are accurate at the normal intensity level.
Olympus rates the E-20N's flash as effective to about 18.4 feet (5.6 meters) at wide angle, and about 12.5 feet (3.8 meters) at telephoto (at the normal intensity setting). In my testing, the E-20N's flash maintained about the same level of brightness as far as 12 feet from the target, and decreased very slightly with each additional foot of distance. The flash was still usable at 14 feet from the target, but slightly dim. Below is the flash range series, with distances from eight to 14 feet from the target.
ISO-12233 (WG-18) Resolution Test (3686 k)
The E-20N performed very well on our "laboratory" resolution test chart. It started showing artifacts in the test patterns at resolutions as low as 1,000 - 1,100 lines per picture height, in both horizontal and vertical directions. I found "strong detail" out to at least 1,200 - 1,300 lines. "Extinction" of the target patterns occurred at about 1,600 lines.
Geometric distortion on the E-20N is higher than I'd like to see at the wide-angle end of the lens' range, where I measured approximately 0.93 percent barrel distortion. The telephoto end fared much better, as I found only a pixel or less of pincushion distortion there. Chromatic aberration is about average among cameras I've tested, which is to say it's more than I'd like, but about typical of the range of even fairly high-end cameras. (Making the obvious comparison to the Sony F707, it's about the same, with both cameras showing more chromatic aberration than the Minolta Dimage 7. - And of course, all these cameras show more C.A. than the high-end professional lenses on Nikon and Canon Pro SLRs. - Perhaps not surprising, given that one of those lenses can cost as much as the entire E-20N.) (For those not familiar with the term, Chromatic Aberration can be seen as a slight colored fringe around the objects at the edges of the field of view on the resolution target.)
Resolution Series, Wide Angle
Sharpness Series, Wide Angle
Resolution Series, Telephoto
Viewfinder Accuracy/Flash Uniformity
An accurate optical viewfinder, so-so LCD.
The E-20N's optical viewfinder showed pretty good accuracy, though it was just slightly tight. I measured approximately 93 percent frame coverage at wide angle, and about 94 percent at telephoto, both good results. The LCD monitor was somewhat less accurate, showing approximately 90 percent of the frame at wide angle, and about 92 percent at telephoto. I'm accustomed to seeing better accuracy in LCD monitors, so the E-20N has some room for improvement there. The optical viewfinder was much more accurate than usual though, and I recommend using it for more accurate framing (the LCD monitor also shifted the image toward the bottom of the frame, making exact framing difficult). Flash distribution at wide angle shows moderate to severe falloff at the corners and edges of the frame, with a bright spot in the center. At telephoto, flash distribution is more even.