Digital Cameras - Fuji MX-2900 Zoom Test Images
(Original test posting: 8/29/99)
|
|
Outdoor Portrait:(829k)
Very good color and tonal range. Well-controlled contrast in this high-contrast
shot preserves both highlight and shadow detail well. The default
shot (829k) is typically dark, due to the light
background. Our main shot (842k)
here was taken with an exposure compensation of +0.6 EV, which brought the
fleshtones to the right brightness level, yet still left detail even in
the model's white shirt. Color balance is very good, with none of the tendency
we've noticed in some cameras to push the particular blues of the model's
pants and the flowers toward purple. We liked the results of the "daylight"
white balance a bit more than the "auto" setting, as shown here(842k),
which we felt was very slightly cool. Resolution and sharpness is good,
but just a shade off from the highest we've seen in other 2 megapixel digicams
we've tested. We observed a tendency we've seen in other digicams that we've
referred to as "tone breaks:" Visible on the shadowed side of
the model's face, minor variations in the subject produce sharp speckles
that look a bit like posterization. Although we didn't try variations of
in-camera sharpening for this shot, we've in the past observed that artifacts
of this sort are often associated with the sharpening function: Drop to
an unsharpened image, relying on post-capture manipulation for sharpening,
and the "breaks" often disappear. For those intersted, the table
below contains links to shots taken with a range of exposure compensations,
from 0 to +1.2 EV.
Exposure Variations:
|
|
|||||||||||||||||
|
Closer Portrait:(865k)
As we often find to be the case, with the model's face filling more of the
screen, the default exposure becomes more accurate. With the MX-2900, the
default exposure (865k)
was actually our choice for the main shot for this test. The zoom lens is
a real help for portrait shots like this, as the longer focal length at
the telephoto end of the zoom doesn't distort facial features (such as the
nose) the way a fixed-focus, wide-angle lens would. Good resolution and
tonal range, although the tendency toward "tone breaks" note above
tends to emphasize wrinkles a bit! As above, the table below has links to
shots taken with a range of exposure compensations, from 0 to +0.9 EV.
Exposure Variations:
|
|||||||||||||||||||
|
Indoor
Portrait, Flash:(849k) The indoor flash pictures we shot with the MX-2900 were a bit on the dark side (probably due to the large white wall behind the model) and had a bit of a warm cast overall, likely because of the heavy incandesdent room lighting. - This shot is generally very difficult for digicams to achieve proper color balance on, due to the huge difference between the color temperature of the flash and the room lighting. To its credit, the MX-2900 produced a very uniform color cast as shown here (849k), with none of the blue highlights we frequently find. As a result, the images "clean up" remarkably well in Photoshop, as shown in this image (684k), which had an "auto levels" operation and a slight gamma boost performed on it. Like other Fuji digicams we've tested, the MX-2900 is unusual in that it allows exposure adjustments on the flash, as well as on the ambient-light exposure. We didn't take any shots with the flash exposure boosted, will try to reshoot before we have to send the camera back... The "Slow Synchro" mode uses slower shutter speeds in conjunction with the flash, allowing more of the ambient light into the scene. This really only works with daylight-balanced ambient lighting though, as you can see here(852k), where the strong yellow cast of the incandescent lighting overpowers the picture. One of the biggest features of the MX-2900 (IOHO) is the presence of the "hot shoe" on the camera's top to accept an external flash unit. You're a little limited in that you only have two apertures available to match the output of the flash unit, but the results produced by a more-powerful external flash unit bounced off the ceiling or walls are exceptional, as seen in this shot (844k).
|
|||||||||||||||||||
|
|
Indoor
Portrait, No Flash:(850k) This shot is a very difficult test of the white-balance algorithms of most cameras, given the strong yellowish cast of the household incandescent lighting used to illuminate it. Still, it's a situation likely to be encountered by users, so we think is a valid test. Like so many cameras we've tested, the MX-2900's automatic white balance system wasn't quite up to the task of getting rid of so much yellow, producing shots with a strong yellow cast, as shown here (850k). The "incandescent" setting fared quite a bit better, producing the result shown here (866k), leaving just enough of the lighting coloration to reflect the way the original scene appeared to our eyes. Both of the shots above were taken with the exposure compensation set to +1.5EV, as much of a boost as the camera provided. The default exposures were quite dark, as seen in these auto (864k) and incandescent(862k) examples. |
|||||||||||||||||||
|
House shot:(876k)
Our standard House poster is one of our strongest tests of detail and resoltion.
The MX-2900 performed exceptionally well on this shot, to our eyes clearly
having the sharpest resolution and capturing the finest detail of any camera
we've tested to date! (September 1999) Our main shot
(876k) has a crispness about it that edges out the
other cameras we've tested to date. Fine details are also finer, again indicating
a higher fundamental resolution. Color balance with the daylight setting
was very good, while the "auto" setting produced this
(159k) rather yellowish low-resolution shot. On the
subject of the low resolution (640x480 pixels) mode, the MX-2900 took very
"clean" pictures that way, something you can't always be guaranteed
in a 2-megapixel digicam. There also seems to be very little noise in the
grey roof, an area where noise is frequently apparent. The table below contains
a full set of resolution/image quality sample shots, including uncompressed
image at both high and low resolution. Overall an exceptional performance
on this test!
We also include for your perusal an uncompressed TIFF-formatted
image. Click here (6,334k!)
to download. (NOTE that this link is NOT an HTML
page, or even a JPEG file, but rather a raw TIFF image: Your browser won't
know what to do with it, but should give you the option to just save the
file to disk. Note too, that this is a huge,
6.3 megabyte file!) |
|||||||||||||||||||
|
Far-Field Test:(881k)
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.
Wow! That's detail! Despite the seasonal variability, this shot is the strongest test of camera resolution of any we do, and the MX-2900 again did exceptionally well with it. Given the changes in the subject from test to test, it's hard to make exact comparisons with it, but it certainly appears that the MX-2900 is noticeably sharper in this test than any other camera we've tested to date (September, 1999). Our main shot (881k) was taken with auto white balance, and the in-camera sharpening set to its default value of 2. Even more remarkable is the result we obtained by turning off the in-camera sharpening (sharpness setting of 0), and then applying unsharp masking in Photoshop(tm) of 200% at a 1.3 pixel radius. Check out that detail! (1056k) Normally this would be too much sharpening on an image like this, but with the in-camera sharpening turned off, the results are extraordinary. (Our heavy-handed post-capture sharpening did boost the shadow noise dramatically, producting the horizontal streaking in the shadows of the trees' foliage.) Color balance here using "auto" mode is very good also. The table below shows all variations of image size and sharpness:
Here's an uncompressed TIFF-formatted version of this image.. Click here (6,496k!) to download. (NOTE that this link is NOT an HTML page, or even a JPEG file, but rather a raw TIFF image: Your browser won't know what to do with it, but should give you the option to just save the file to disk. Note too, that this is a huge, 6.5 megabyte file!) For those interested in the behavior of the in-camera sharpening, this table shows the effects of all five sharpening settings on a high resolution image:
|
|||||||||||||||||||
|
|
Lens Zoom Range (new)
We've received a number of requests from readers to take shots showing the
lens focal length range of those cameras with zoom lenses. Thus, we're happy
to present you here with the following series of shots, showing the field
of view with respectively, the lens at full wide-angle, plus the 0.8x wide-angle
adapter, the lens at full wide-angle, and the lens at full telephoto. (We
didn't bother with the digital zoom settings, as they're just cropping into
the frame, rather than magnifying it.) (Ed. note: Whoops!
Tele shot is badly blurred! (accidental manual focus?) We'll try to reshoot,
but even if we don't, the sizing is accurate. Click on an image to view,
for now don't pay attention to the focus!!)
|
|||||||||||||||||||
|
Musicians Poster:(890k)
While this is just a picture of a poster, the color values for the
various skin tones are pretty representative of the three ethnic groups
represented. Skin tones are tough for digital cameras, both because the
Caucasian skin color is so sensitive to over-saturation, and because all
of the tones are "memory colors:" People are so familiar with
the range of "correct" colors that any deviation is immediately
obvious. The MX-2900 again did very well here, as seen in our main
shot (890k). This image was shot using the "daylight"
white balance because the auto setting produced a very warm-toned photo,
as shown here. (148k) The
daylight version is still slightly warm, but well within acceptable limits,
and easily on a par with other high-end digicams we've tested thus far (September
1999). Detail again is exceptionally good, as can be seen in the fine silver
lines on the Oriental model's robe, and in the hair and flowers of the Caucasian
girl.
As before, we've arranged a full set of resolution/image quality samples in the table below, as well as samples showing the effect of the full range of sharpness settings on a high-resolution image.
Here's an uncompressed TIFF-formatted version of this
image. Click here (6,344k!)
to download. (NOTE that this link is NOT an HTML
page, or even a JPEG file, but rather a raw TIFF image: Your browser won't
know what to do with it, but should give you the option to just save the
file to disk. Note too, that this is a huge,
6.3 megabyte file!)
|
|||||||||||||||||||
 |
Macro Shot:(855k)
The MX-2900's macro function does pretty well, although it's not in the "microscopic" range of some current digicams. At closest approach, it captures an area of 2.6 x 3.9 inches (6.7 x 9.8cm), as shown here (855k). The flash throttles-down quite well for closeups, in fact somewhat underexposing this shot (847k), due to the strong reflection from the silver dollar. As usual, the "digital telephoto" mode doesn't add any detail, simply cropping into the central portion of the image. For normal shots, we tend to pooh-pooh digital tele, but macro shooting may be an exception: For shots going onto the web, letting the camera do the cropping for you ahead of time makes sense, so the effect really could amount to the 2.5x "magnification" it's touted as. The shots taken here show the results with the digital tele set to 1.25x (651k) and 2.5x (150k). |
|||||||||||||||||||
 |
"Davebox"
Test Target:(853k) The MX-2900 did reasonably well in this test, but it's color performance was hurt a bit by an overall yellowish cast that we've observed with other Fuji cameras in the past. This cast appeared in both the main shot (853k), taken with auto white balance, and in this sample low-res one (143k), taken with the daylight setting. (In this test, the two white balance settings produced almost identical results. We've commented before on these overall color casts, and don't understand why the manufacturers can't eliminate them. Certainly, they're easy enough to deal with after the fact, using programs like Adobe Photoshop, or amazingly useful PhotoGenetics , by Q-Research, which we've reviewed, elsewhere on this site. (For reference, here's a low-res version of the image that we cleaned-up in Photoshop (128k) with just an "auto levels" command.) Tonal range is about typical of the 2-megapixel cameras we've tested, although shadow noise is a bit higher than we've seen with some units. The camera struggles with the very difficult red/magenta separation in the small horizontal color-separation target, a common failing among the cameras we've tested. Overall, a good performance, but one that could be improved by eliminating the yellowish overall tint. As before, we have a full resolution/image-quality series in the table below:
Here's an uncompressed TIFF-formatted version of this image. Click here (6,344k!) to download. (NOTE that this link is NOT an HTML page, or even a JPEG file, but rather a raw TIFF image: Your browser won't know what to do with it, but should give you the option to just save the file to disk. Note too, that this is a huge, 6.3 megabyte file!)
|
|||||||||||||||||||
|
Low-Light Tests: Our low-light test results of the MX-2900 were a bit of a mixed bag, with an overall excellent performance marred by image noise at the longest exposure times. We found the minimum usable lighting (with post-processing of the image) was about 6EV units by our traditional reckoning; more properly 0.5 footcandles, or 5.5 lux. (That's *really* dim!) The positive points about the MX-2900's low-light behavior are that it appears to focus well down to extremely low light levels, and perhaps even more significantly, it maintains a true color balance, with none of the odd tonal inversion of certain dark-colored objects we've seen in some other high-end cameras (which we're tentatively attributing to infrared behavior.) It's minimum usable light level is limited by a maximum exposure time of only 3 seconds, which was probably chosen by Fuji as being the longest shutter time possible before noise became *really* objectionable. - We're perhaps pushing it a bit with our rating of 0.5 footcandles, as the resulting image is somewhat dim, but it's still within the range we'd consider "usable." The table below shows the best exposure we were able to obtain for each of a range of illumination levels. Images in this table are untouched, exactly as they came from the camera.
Cleaning up the noise: We commented on the noise in the MX-2900's lowest-light images, but were somewhat intrigued in that it looked rather different than noise we've seen in other high-end digicams. Rather than the "blotchy", large-featured noise pattterns of other cameras, those of the MX-2900 are characterized by very fine "grain", mostly horizontal stripes of differing brightness. After some experimenting, we found that it was actually very easy t o clean up the images in Adobe Photoshop, producing truly excellent results even at the 0.5 footcandle minimum lighting level we tested. Check out this pair of pictures: This one (928k) is the result of a simple "Auto Levels" operation - the image is much brighter, color is really excellent(!), but the noise is quite prominent. Going one step further though, this shot (692k) shows the effect of performing a 1-pixel "median filter" operation. WOW! A lot of the fine-grained noise is, detail is still pretty good (better than most digicams we've tested at this low a light level), and the color is much better than we're accustomed to seeing in low-light digicam images. Bottom line, if you're willing to spend a little time in Photoshop with your photos, the MX-2900 can deliver exceptional low-light imagery.
|
|||||||||||||||||||
|
|
Flash Range Test (New):
(This test was added in August 1999, so cameras tested before that time
won't have comparison pictures available. As we go forward though, all the
new models will have similar tests available.)
In response to several reader requests, we've begun testing the usable working range of the on-board flash units of digicams. We don't have a real formal procedure for this, but mainly want to determine whether the manufacturer's ratings are reasonable or not. In the case of the MX-2900, we're happy to report that the flash unit appears to be quite conservatively rated: Fuji rates it to only 8.2 feet (2.5 m) in telephoto mode (11.5 feet or 3.5 m in wide-angle mode), as shown in this shot (858k). As you can see from the shots in the table below though, light falloff beyond this distance is quite gradual, and we felt the flash was at least usable out to 11 feet or so in tele mode. Light Falloff (dist from object)
|
|||||||||||||||||||
 |
ISO-12233 (WG-18) Resolution
Test: This test is always interesting, as a way to pick apart "in
the laboratory" what the camera is doing resolution-wise. In the case
of the MX-2900, we remarked earlier by what we felt to be its exceptional
resolution and detail-capture ability, noted mostly in the House and Far-Field
tests. The results on the resolution target itself are a little more equivocal.
The vertical resolution (horizontally-oriented target elements) is at first
glance extraordinary, appearing to reach out to 900 lines per picture height
or beyond. This is an "obviously impossible" value, since the
camera only has 1200 pixels vertically. Closer inspection reveals that there
is subtle aliasing, beginning at about 600-650 lines per picture height,
as you'd expect. The aliasing is so well controlled though, that our eyes
tend to ignore it, seeing instead smooth lines all the way to 900 line resolutions.
Pressed for a judgement, we'd have to peg the "visual" resolution
at somewhere on the order of 700-750 vertically. In the horizontal direction
(vertical chart elements), the performance is more modest/typical, with
aliasing becoming evident at 600 lines/picture height, and a visual resolution
of somewhere in the vicinity of 600-650 l/ph. Overall, the MX-2900 did very
well on the resolution target, although its performance was as clearly superior
as we felt we saw in the "natural" subjects. As usual, the tables
below show the full range of resolution/image quality options, as well as
a series showing the effects of in-camera sharpening.
Resolution/Image Quality Series Wide Angle
Here's an uncompressed TIFF-formatted wide-angle shot. Click here (6,344k!) to download. (NOTE that this link is NOT an HTML page, or even a JPEG file, but rather a raw TIFF image: Your browser won't know what to do with it, but should give you the option to just save the file to disk. Note too, that this is a huge, 6.3 megabyte file
Here's an uncompressed TIFF-formatted telephot shot. Click here (6,344k!) to download. (NOTE that this link is NOT an HTML page, or even a JPEG file, but rather a raw TIFF image: Your browser won't know what to do with it, but should give you the option to just save the file to disk. Note too, that this is a huge, 6.3 megabyte file
|
|||||||||||||||||||
 |
Viewfinder Accuracy/Flash Uniformity:
The optical viewfinder in the MX-2900 varied in its accuracy as the lens was zoomed. At the wide-angle end (227k) of the lens' range, it was more accurate than most, showing about 90% of the final image area. At the telephoto end (227k) though, it was much less accurate, showing only 76.5% of the final area. The telephoto performance is somewhat worse than average, but we feel that the variation as a function of focal length is a bigger problem, since it will be hard to know just how much to compensate for the viewfinder at any given lens setting. By contrast, the LCD viewfinder was 100% accurate at both wide(227k) and telephoto (227k) settings. (Surprising as it may seem, many LCD viewfinders show less than the full image area.) We've recently begun testing cameras for optical distortions, such as
barrel/pincushion distortion and chromatic aberration. The MX-2900 did
pretty well on the geometric distortion front, with 0.5% barrel distortion
at the wide-angle end of the lens' range, and 0.06% (effectively zero)
at the telephoto end. Chromatic aberration was very low, with only the
slightest fringe of color on test elements at the extreme corners of the
resolution target. |
Follow Imaging Resource: