Digital Cameras - Kodak DC265 Test Images
The press of several coinciding deadlines meant that it took us a bit longer than expected to get the full analysis of these images posted, but here they are at last! We probably shot more images with the DC265 than with any camera to date, providing full resolution/compression series of many of the shots, because we've had a lot of questions about this aspect of the '265.
As always, we encourage readers to download the images for their own personal use, to examine them more closely, and output them on their own color printer.
|
|
Outdoor portrait: (402k) Excellent detail, really excellent color: Kodak's always had good color in their digital cameras, but their latest products (the DC265, DC240, and even the low-end DC200) have taken the color-handling another step forward. The skin tones here are just a bit ruddy, but everything else is spot-on, including the color of the wall the model is standing against. Detail is excellent, although Kodak uses a little less in-camera sharpening than some manufacturers, resulting in a slightly "softer" image. (Or, maybe the sharpening effect is just spread over a slightly larger area.) We found that a tight unsharp mask operator in Photoshop brought out even more detail, albeit at the expense of increasing the noise somewhat. Almost unique in our experience, the default exposure (402k) turned out to be the right one on this image. (Most digicams are fooled by the overall brightness of this subject, and underexpose as a result. The DC265 seemed to have no trouble though, and got it exactly right on the first "go". (For comparison, here are images shot at +0.5 EV (418k) and +1.0 EV (410k) respectively.) We noted that Kodak has tweaked the color relative to the DC260: Part of the process of getting brighter colors seemed to be an increase in contrast, with the result that detail is lost in the extreme highlights. The overall exposure is superb though.
|
|
||||||||||||||||||||||||
|
|
Closer portrait: (406k) Again, excellent detail and color. Curiously, despite the perfect exposure on the more-difficult shot above, the DC265 underexposed this shot slightly with the default exposure setting, as seen here (384k). A half-stop positive exposure compensation produced a fine result for our main shot here (406k), while a full EV's worth of compensation lost too much in the highlights, in our view, as seen here (417k). (Although, we felt the skin tones benefited from the full f-stop of compensation) For reference, here's a version shot with the "Super" resolution setting (649k). Detail and resolution on this shot are exceptional, in our view competing in the 2 megapixel class, as of this writing (April, 1999).
|
||||||||||||||||||||||||||
|
|
Indoor
portrait, flash: (395k) This indoor shot,
taken with the on-board flash was exposed very well, and showed excellent
color. Enough of the room lighting was left in the final image to preserve
much of the original feeling of the scene, and the flash illumination
is well-balanced with it. There is a somewhat bluish cast to the white
areas shadowed from the room lighting (household incandescent), but the
effect isn't too severe. (This is a common problem for cameras on this
shot, due to the drastically different color balance of the flash and
the room lighting. The DC265 did about as well as most in this respect.)
Ah, but the real treat with the DC265 on indoor shots comes when you use
an external flash, and do things like bounce it off the ceiling,
as we did here (363k).
Wow! Once you've enjoyed the sort of flexibility an external flash provides,
you'll never want to go back. - The 1 f-stop resolution of the DC265's
manual aperture setting for external flash use is more precise than some
cameras we've tested, and it's really handy when you need to tweak the
exposure for external flash units. - If you'll compare this shot with
some of the other indoor flash images in the Comparometer, you'll see
just how big a deal external flash is, and how valuable it is as a feature
in digicams. - This is probably one of the best-exposed, most natural
flash exposures we've taken with a digicam to date. (April, 1999) |
||||||||||||||||||||||||||
|
|
Indoor
portrait, no flash: (429k) The DC265 had
more trouble with this tough shot, where the strong yellow cast of the
household incandescent lighting poses a severe challenge for any camera's
white-balance circuitry. In the case of the DC265, it actually did quite
well (better than most), producing a reasonably neutral color for the
white wall behind the model, but left a rather reddish cast in the model's
skin tones that we found hard to eradicate even with Photoshop. (Actually,
we eventually did manage to get a well-balanced image: The trick was to
use the "Selective Color" control in Photoshop, reducing the
level of magenta in the reds and magentas.) We chose the +0.5EV shot
here (429k) as
the main one for this category, feeling it offered the best tonal balance
and overall exposure. The default exposure (442k)
was somewhat dark, while a compensation of +1.0EV
(416k) lost some highlight detail. As it turned
out, the "incandescent" manual white balance setting (413k)
produced a less well-corrected image than did the automatic setting.
(Actually, we've frequently found this to be the case, as the "incandescent"
setting of many digicams is calibrated to professional tungsten lighting,
which has a much higher "color temperature" than household bulbs.) |
||||||||||||||||||||||||||
|
|
House shot:
(1118k!) This image is always a strong test of
fine-detail handling by the various digicams we evaluate. The DC265 did
very well here, clearly among the best we've tested to date (April, 1999):
Not quite to the level of the just-now-appearing 2 megapixel cameras,
but much closer than many people would probably expect. As has become
our custom, we shot a full series of images, exercising all resolution/quality
settings. (This exercise greatly aided by the "resolution series"
Digita script that comes bundled with the camera!) The results appear
in the table below.
|
||||||||||||||||||||||||||
|
|
Far-Field shot: (941k) 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. Despite the seasonal variability, this shot is the strongest test of camera resolution of any we do, and the DC265 again did well here, clearly at the top of the sub-2Megapixel cameras, and noticeably better than the earlier DC260. (If you're comparing with the DC260, check the greatly increased detail in the brickwork.) As before, we have a full series of resolution/image quality selections presented in the table below.
|
||||||||||||||||||||||||||
|
|
"Musicians" poster: (868k)
Very good resolution and color from the DC265 in this shot. We
were a bit on the fence as to which white balance setting to select
here as the best representation of the original: The auto white balance
setting (which we ended up choosing as the main
image (868k) here) was a little cool in its hues, while
the "daylight" setting produced a rather yellowish cast as
shown in this VGA-level image (215k).
Detail is VERY good, as seen in the subtle silver lines on the shoulder
of the Oriental model's robe. As before, we've included all resolution/image
quality combinations, this time with the added variable of automatic
vs. daylight white balance settings.
|
||||||||||||||||||||||||||
 |
Macro shot: (521k) The DC265 improved somewhat on the macro performance of the DC260, offering a minimum focusing distance of only 8 inches, as compared to the 12 inch minimum of the earlier product. At the 8 inch minimum distance (521k), with the lens set to the telephoto position, the total area covered is a relatively small 2.0 x 3.0 inches (51 x 76 mm). Using the digital telephoto (328k) halves these dimensions to 1.0 x 1.5 inches (25 x 38 mm), but with half the inherent resolution. The same shot, captured with +6 diopters of close-up adapter lenses attached to an Xtend-A-Lens adapter (561k)(see the main review) produces a minimum area of 1.3 x 2.0 inches (33 x 50 mm), but with the full resolution the camera is capable of. (The existence of the Xtend-A-Lens adapter greatly increases the utility of the DC265 for close-up photography. Note though, that the autofocus only works to 8 inches, so using auxiliary close-up lenses will return you to the days of measuring lens-to-subject distances manually.) (Note: As we were "going to press" with this review, we noticed on the Kodak web site that the "official" minimum focus distance is stated as 12 inches. - We'll try to find out what the real story is on this, and update the review as we do.) Meanwhile, here's a shot taken at 12 inches (521k), and another at the same distance, using the on-board flash (521k).
|
||||||||||||||||||||||||||
 |
"Davebox" test target:
(567k) Wow! Along with others in this generation
of Kodak cameras (the DC240 and DC200), clearly some of the best color
we've seen to date! The very difficult red/magenta color separation
is handled exceptionally well, all colors are true and well-saturated,
yet the delicate pastels of column "B" of the Q60 target at
bottom center are clearly visible. Likewise, shadow detail is about
as good as it gets, while highlights are well-preserved as well. Overall,
a very impressive performance, and a new level for the trademark "Kodak
color". In this shot, the automatic white balance setting produced
the best results, with the "daylight" setting leaving a slightly
yellow cast. As with the Musicians shot above, we've included a full
set of resolution/image quality samples here, for both the automatic
and daylight white-balance settings.
|
||||||||||||||||||||||||||
|
Low-Light Tests (NEW!) The DC265 performed well in the low-light tests, producing a very usable image at a light level as low as EV5 (447K), with a 16-second time exposure, and a somewhat usable one at EV4 (467K). This is really quite dark, albeit a good bit lighter than the camera's official specs would indicate as the lower light limit (about -0.5EV, according to our calculations.) The autoexposure system worked well down to about EV7 (357K). Using the autoexposure system, we found that increasing the manual exposure compensation adjustment produced little effect at light levels below about EV8. We also observed an odd artifact in the images shot with long time-exposures, in which a ghostly texture appears overlaid on the images in the upper right-hand corner. Interestingly, we observed almost no "stuck pixels" in the final images, although the "QuickView" instant-review images on the LCD panel seemed to show some. From the presence of the odd texture, and the almost complete absence of stuck pixels, we surmise that the DC265 does a fair bit of post-processing on the images to hide the effects of any off-spec sensor elements. (Just to prove that what we were seeing as an artifact in the upper left-hand corner of the images didn't in fact result from anything in our studio, we shot this image (322K) in complete darkness, and the pattern is clearly evident. All this said though, the DC265's low-light capability is very good indeed, exceeding that of most units currently on the market. (April, 1999)
|
||||||||||||||||||||||||||
 |
ISO 12233 ("WG-18") resolution
target: (599k) (Technoids only) The DC265 did well in this test, producing a visual resolution of about 700 lines per picture height (l/ph) in both the vertical and horizontal directions at the wide-angle end of its lens' range, and slightly less at the telephoto end. In the tables below, you'll find full resolution/image quality matrices for both telephoto and wide-angle settings of the lens.
|
||||||||||||||||||||||||||
 |
Viewfinder accuracy/flash uniformity target: The optical viewfinder in the DC265 is more accurate than most, at a bit over 90% coverage of the final image area, and the LCD finder is DEAD accurate, showing 100% of the final image area. The optical finder is almost perfectly centered vertically at the telephoto end of the lens' range, and off by about 3-5% to the left horizontally. At the wide-angle end (or perhaps more accurately, at the shorter lens-to-subject distance necessitated by our test of the wide-angle setting), the area captured moves more to the left and down, by about 5% and 2% respectively, relative to the image position in telephoto mode. Overall though, the optical viewfinder is more accurate than most. The one gripe we did have about the optical finder though, is that it showed an image tilted about 2 degrees clockwise, relative to what the CCD itself saw. We repeated our tests several times, but the discrepancy persisted. We'd previously seen this problem in one other camera (from a different manufacturer), also with an early preproduction model. We suspect that the cause of the problem in our eval unit of the DC265 is the same, a misaligned assembly jig used for the hand-built prototypes. We seriously doubt that this will be a problem in the final production models. Flash uniformity was good in wide-angle mode, with only relatively slight falloff at the edges of the frame. Uniformity was excellent in telephoto mode. The table below holds links to our viewfinder/flash tests, for both wide-angle and telephoto lens settings.
|
||||||||||||||||||||||||||
Follow Imaging Resource: