Digital Cameras - Fuji DX-10 Zoom Test Images
|
|
Outdoor portrait:
(272k) Good color and tonal balance, although the
default exposure setting (264k)
was rather dark, as are most other digicams on this shot. (Most digicams
significantly underexpose this high-key subject.) Highlight detail was good
in our main image (272k),
shot at +0.9EV exposure compensation, although best in this
shot (272k), taken at +0.6EV. Detail is very
good (particularly considering the 0.8 megapixel sensor resolution), and
color is exceptional, with the DX-10 handling even the ever-difficult blues
of the model's pants and the blue flowers just right! Color accuracy and
saturation is very good, yet skin tones are very natural and not over-saturated
at all. A very nice performance! |
|
|||||||||||
|
|
Closer portrait:
(276k) This closer shot shows very good detail (surprisingly
good for a sub-megapixel camera), and excellent color. With the subject
coming closer to filling the frame, less exposure compensation was required,
with our main shot (276k)
at +0.6EV, but this one (276k)
at +0.3EV doing a better job of preserving highlight detail. For reference,
here's a shot taken with the default exposure setting
(276k). One thing that this shot does show though,
is the limitation of the wide-angle lens for portrait shooting: Note how
rounded the model's features are. There are some contrast breaks in the
shadows under the model's eyes, but interestingly, not as noticeable as
those we observed with the DX-10's "big brother," the MX-600Z.
Overall a very good performance! |
|||||||||||||
|
|
Indoor portrait,
flash: (280k) We were very impressed with
how well the on-camera flash of the DX-10 did in this shot: The overall
image is just slightly dark (at least for Windows users, less so for Mac
folks), but the color balance between the on-board flash and the strong
incandescent ambient lighting is unusually good. Most digicams we've tested
either produce strong bluish highlights or very yellow shadows on this shot:
The DX-10 is among the best-balanced of the cameras we've tested on this
particular subject. The DX-10 is also relatively unique in its ability to
adjust flash intensity as well as ambient exposure. This
shot (284k) was taken with the flash turned-down
0.6 EV units, while the ambient exposure was boosted by 0.9EV. The result,
while dark, shows very natural lighting and color balance, difficult to
obtain with more conventional on-board flash units. |
|||||||||||||
|
|
Indoor portrait,
no flash: (288k) This shot is a very tough
test of cameras' white-balance circuitry. We frequently find that digicams
do best on this particular shot using their "auto" white balance
setting, as the "incandescent" settings are often more suited
to professional lighting than household lamps. Not so the DX-10: The main
shot here (288k) was taken with the "incandescent"
white-balance setting, and the exposure compensation adjusted up all the
way to +1.5EV. Even the default exposure setting (280k)
in "incandescent" mode produced a good shot, albeit a rather dark
one. There is some remnant yellowish/magenta cast to the image, but we found
that it cleaned up remarkably well in Photoshop, using a simple "auto
levels" adjustment, as shown here
(268k). By contrast, the automatic white balance
of the DX-10 produces a very yellowish shot, as seen here
(284k), with the exposure compensation set to +0.9EV.
|
|||||||||||||
|
|
House shot: (680k) Our
standard "House" poster shot is a great test of resolution and
detail rendering. The DX-10 did fairly well on this shot (although we
blew the framing slightly!), with good color balance, although the exposure
was slightly dark overall, and the "normal" sharpness setting
produced annoying "checkerboard" artifacts on some of the strong
horizontal detail, such as along the ridge line of the roof, and in the
grill at the top of the central gable. For this shot, we provide a full
set of images (4), showing the camera's performance in both compression
modes, at both resolution settings. We also include samples shot with
both auto and daylight white balance settings, although there is really
very little difference between the two, the daylight being only very slightly
cooler in its color balance.
Size/Resolution Variations,
Daylight WB:
This test also provided a significant revelation about the source of the "checkerboard" artifacts we noted above: We've seen these in other cameras from time to time, but in past tests couldn't find any way to avoid them if they appeared. In the case of the DX-10 though, they seem to be almost entirely caused by the camera's sharpening algorithm. When we ran a series of images (see table below) varying the "Sharpness" setting, the artifacts were very aggravated by the "Hard" setting, prominent in the "Normal" setting, but almost completely absent in the "Soft" setting! - This means that you can dramatically improve the DX-10's picture quality (at least for objects with strong horizontal lines in them) by setting the sharpness to "Soft" in the Setup menu, and applying sharpening after the fact in an image-manipulation program. This is very good news indeed, and removes what we had seen as a significant image-quality problem with the camera! |
|||||||||||||
|
|
Far-Field shot: (276k)
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.
|
|||||||||||||
|
|
"Musicians" poster: (276k)
Very good color, good detail - really a very good performance on
this image! While the DX-10's resolution isn't up to that of the higher-priced
1-to-2 megapixel digicams, it does a surprisingly good job of capturing
detail, and the color is very good, with decent saturation, and natural
skin tones. We show only "auto" white balance samples here,
as we found virtually no difference between auto and daylight settings
with this image. The tables below shows samples of images shot at all
resolution/compression settings, as well as the range of sharpness settings,
taken at high resolution.
|
|||||||||||||
 |
Macro shot: (284k)
- Despite its somewhat wide-angle lens, the DX-10 did pretty well on the
macro test: The one penalty was a rather short working distance of 3.9 inches,
which makes it a bit trickier to get adequate light onto the subject. At
the minimum focusing distance, the DX-10 captured an area of 2.5 x 3.3 inches
(6.3x8.4 cm). With the flash throttled-down by the -0.6 EV maximum, it produced
this (288k) surprisingly
well-exposed shot. - While the illumination is a bit uneven, due to the
very short working distance, it appears the flash is very usable for macro
shooting. |
|||||||||||||
 |
"Davebox" test target: (280k) This test proved a little more challenging to the DX-10, in that its default exposure produced rather dark images. We re-shot with the exposure compensation adjusted upward by +0.6EV, and obtained the results shown in the main shot (272k), and in the table below. The images have a very slightly overall yellow cast, that cleans up well in Photoshop, or by boosting the blue channel a bit in Fuji's EZTouch program. Other than that, color rendition is very good, with well-saturated primaries, yet delicate handling of the pastels on the Q60 target. Overall, a surprisingly good performance for an inexpensive digicam!
|
|||||||||||||
|
Low-Light Tests (NEW!)
|
|||||||||||||
 |
ISO 12233 ("WG-18") resolution target:
(276k) (Technoids only) The DX-10 showed a visual
resolution of about 575 lines/picture height in both the horizontal and
vertical directions on this test. In some of the tests outlined above,
we felt that the "Soft" sharpness setting produced somewhat
superior results to those obtained in "Normal" mode, due to
the elimination of checkerboard patterns. The resolution target revealed
though, that the Soft setting in fact did reduce the resolution markedly,
particularly in the horizontal direction. This was surprising, because
we didn't feel we saw as much of a decrease in resolution with our other
tests when working in Soft mode. Bottom line? We'd probably shoot in Soft
mode and sharpen after the fact most of the time. If we wanted an image
to print at a larger size, and it had no strong horizontally-oriented
elements in it, we'd probably go with the Normal sharpness setting. The
tables below show the resolution results with all combinations of image
size and compression setting, as well as with the three different sharpness
settings at large size.
Sharpness Variations
|
|||||||||||||
 |
Viewfinder accuracy/flash uniformity target:
We griped somewhat about the DX-10's optical viewfinder in the main review,
the viewfinder accuracy test shows what we were talking about: We normally
try to align the viewfinder accuracy target using the edges of the viewfinder
frame. In the case of the DX-10, we had to really squint and move our eyes
around to see the edge of the frame, and as a result ended up seeing a lot
larger swatch of the target than the viewfinder is probably intended to
reveal. The net of this was that the DX-10's optical viewfinder (300k) was
very atypical in our tests, showing MORE of the target than the CCD ultimately
captured (about 110%, as near as we can tell), rather than less. We re-shot
the test, taking a more "casual" approach to the boundaries of
the viewfinder (as a typical user would), with this
(144k) result, and found a much-improved result,
with the finder showing 96% of the final image area. This is very good,
but we still dislike the uncertainty as to just what's being captured...
The LCD finder (296k) was
much better-behaved, showing 89% of the final image area, about typical
for LCD viewfinders we've tested. We also use this test to measure geometric
distortion of cameras' lenses, and the DX-10 does quite well in this regard,
showing only 0.6% barrel distortion. |
Follow Imaging Resource: