Fuji FinePix 4700Fuji packs a 2.4 million pixel "SuperCCD" sensor and 2400 x 1800 images into an ultra- compact digicam!<<Reference: Datasheet :(Previous) | (Next): Print-Friendly Review Version>> FinePix 4700 Sample ImagesReview First Posted: 10/7/2000 |
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! ;) |
|
 |
Outdoor Portrait (1684k)
This is a tough shot for many digicams, due to the extreme tonal range (which is why we set it up this way!). The trick is to hold highlight and shadow detail without producing a "flat" picture with muddy colors and the F4700 performed nicely. We shot this image using the automatic (1706k) and daylight (1684k) white balance settings, both producing very similar results. The auto setting produced a slightly cooler image, leading us to choose the more hue-correct daylight setting for our main series. Color balance looks very good on this shot, judging from the always-difficult blues of the flowers and the model's pants, which the 4700 rendered correctly in color and tone. (Many digicams have a tendency to reproduce these with a purplish hue). The bright colors of the flower are rendered very well, but the skin tone is slightly over-saturated. Overall very good color handling though. Resolution and detail look good, although not as crisp as we've seen (most noticeable in the small green leaves next to the model's shirt and the strands of her hair). Overall, we'd have to say that the resolution here looks very much like a 2+ megapixel CCD that's been enlarged to a 4 megapixel image size: Compared closely with shots of earlier 2 megapixel cameras, the 4700 seems to show about the same level of detail here, but spread across a larger chunk of pixel real estate. The shadow areas preserve a fair amount of detail with a moderate amount of noise. We required a positive exposure compensation of +0.6 EV units on this shot for our main image, which managed to get the best exposure overall without blowing the highlights in the shirt and the white flowers too much. The table below shows the results of a range of exposure settings from zero to +1.5 EV in the daylight white balance setting.
Exposure Compensation Settings:
|
|
||||||||||||||||||||||||
 |
Closer Portrait (1670k)
The F4700 also does a nice job with this "portrait" shot, thanks in part to the camera's 3x zoom lens. (Shorter focal length lenses tend to distort facial features in close-up shots like this and the availability of longer focal lengths is a key feature if you're going to be shooting close-up people shots). Our main shot (1670k) required just one notch of exposure adjustment (+0.3 EV) to brighten the shadow areas while not getting too hot in the highlight areas. Sharpness and detail look good in both the highlight and shadow areas, again showing a moderate amount of noise in the shadows though. The camera does a very good job holding fine detail against a background of significantly different tone and hue (the model's hair against the background), but tends to lose detail in areas of similar tonal and hue (the model's hair, particularly in areas in open shadow). The table below shows the results of a range of exposure compensation settings from zero to +1.5 EV. As with the Outdoor Portrait shot, we chose the daylight white balance setting as it produced the most true color balance.
Exposure Compensation Settings:
|
||||||||||||||||||||||||||
 |
Indoor Portrait, Flash (1625k)
The F4700 does pretty well in this category also, the variable ISO and flash intensity settings giving the camera a little more flexibility. We took our first series of shots in the normal sharpness setting and found the resulting images to be a bit soft (1582k). So, we reshot with the plus (+) sharpness setting and achieved much nicer results (in our opinion anyway, a matter of personal preference). In our first shot, we fired the flash at its normal (1592k) intensity setting, which produced slightly dark results. We then bumped up the flash intensity level to +0.3 EV, producing this (1607k) brighter image. Finally, we increased the intensity level up to +0.6 EV, which resulted in this (1598k) much brighter image. This last image produces nice shadow and highlight areas. The white of the model's shirt and flowers looks correct, as do the other tonal values in the image, despite the slightly warm color cast. For our main shot (1625k), we left the flash at the +0.6 EV level and set the ISO value to 400, which allowed more room light into the image and we felt looked the best overall. The incandescent room lighting left a fair bit of warmth in the image, although the camera's flash seemed to match the color balance of the household incandescent fairly well. (There were no harsh blue highlights as we sometimes see in this shot with other cameras.) We discovered that the remaining color cast cleaned up very quickly in Photoshop(tm), with a simple "auto levels" operation while leaving enough of the warmth of the room light in the image to retain the mood of the image, as shown here (996k).
|
||||||||||||||||||||||||||
|
|
Indoor Portrait, No Flash (1601k)
This shot is 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. For our main EV test, we chose the incandescent (1601k) white balance setting, as it produced the best color balance overall (although still rather warm, in an oddly greenish way). Once again though, we found that a quick "auto levels" in Photoshop cleaned up the image and made it very usable. (We suspect that our favorite image-adjuster PhotoGenetics(tm) would be very effective for developing a "genotype" to automatically correct exposures such as this.) The automatic (1630k) white balance setting produced very warm results (shot at +0.9 EV). We also shot with the camera's 400 (1668k) and 800 (1659k) ISO settings, to check the noise level and exposure. As you'd expect, the images do increase in noise level as the ISO speed increases. We didn't do any formal comparisons, but would say that the noise level at ISO 400 is quite good, and at 800 is probably similar to other "prosumer" digicams at ISO 400 (pretty noisy, but usable if that's the only way you'll get a shot). Both images were shot with the incandescent white balance setting and with a +0.9 EV adjustment. The table below shows our standard series, a range of exposure compensation settings from zero to +1.5 EV in the incandescent white balance setting.
Exposure Compensation Settings:
|
||||||||||||||||||||||||||
 |
House shot (1781k)
NOTE that this is the "new" house shot, a much higher-resolution poster than we first used in our tests. To compare the F4700 with previously tested cameras, here's a shot of the original house poster at the automatic (1717k) white balance setting. For this test, we shot with the automatic (141k), daylight (141k), daylight fluorescent (140k) and warm white fluorescent (140k) settings. Interestingly enough, we found that the warm white fluorescent and the automatic setting produced similar results that were the most accurate. Both the daylight and daylight fluorescent settings were too warm, so we selected the automatic setting for our main shot (1781k). Resolution and detail look rather soft, but still good. (This target is a very strong test of fine detail and resolution on a "natural" image, as compared to the highly structured image of the resolution target. As a result, we compared this image very closely with several from 2 megapixel cameras. The result? We have to say that (in this image at least) the 4700 behaves exactly as a ~2.5 megapixel camera that's been upsampled. There's clearly more detail evident than on 2.1 megapixel cameras we've tested, but equally clearly less detail than that revealed by most 3 megapixel models.) Some noise can be seen in the telltale shingle area and a small halo around the dark and light edges gives minute evidence to the in-camera sharpening. Interestingly, we found we could add Photoshop unsharp masking to the default-sharpened 4700 images and obtain noticeable increases in sharpness without significant artifacts - Try 200% at 0.4-0.5 pixel radius, if you're inclined to experiment.) Overall, the F4700 does a good job with this test, but a good job relative to other ~2.5 megapixel cameras. The table below shows our standard resolution and quality series.
|
||||||||||||||||||||||||||
| Sharpness Series
We also shot with the camera's variable sharpness settings, which do a nice job of not going too far in either direction. The minus setting just softens the image while the plus setting is crisp and clear. Below are the results. Plus Sharpness (1761k) Normal Sharpness (1781k) Minus Sharpness (1724k)
|
|||||||||||||||||||||||||||
 |
Far-Field Test (1740k)
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. We shot this image with several white balance settings, choosing the automatic (142k) setting as the most accurate. Daylight (142k) produced nearly identical results. Both the cloudy (142k) and daylight fluorescent (141k) settings came out very warm, the latter with a slightly pink cast. The warm white (144k) and cool white (141k) fluorescent settings also came out slightly pink and incandescent (141k) produced very blue results (no surprise there). We shot with all of these settings because in the studio, we found that some of the fluorescent settings seemed to be more accurate than the auto, even though we weren't shooting with fluorescent lighting. Regardless, on this shot, we chose automatic white balance for our main shot (1740k). This test is the strongest test of detail of any we do, and the bright white of the central bay window often tricks digicams into losing detail in that area. The F4700 was nearly fooled on this one, as the bay window is slightly blown out but the outlines are still visible. Color balance and saturation look great but a moderate amount of noise shows up in the roof shingles. Resolution looks good but slightly soft, especially in the bricks and tree limbs. Resolution appears a bit less than the best ~2.5 megapixel cameras we've tested, suggesting that the 4700's lens is just a hair soft at infinity. We also shot with the 200 (1742k), 400 (1755k) and 800 (1724k) ISO settings, as a gauge of noise performance at elevated ISO and high scene brightness. No surprise, the noise increased directly with the ISO rating, and as before was fairly manageable at ISO 400, but quite evident at ISO 800. The table below shows our full resolution and quality series.
|
||||||||||||||||||||||||||
| Sharpness Series
As with the House poster, we shot with the camera's variable sharpness settings, with the results below. Plus Sharpening (1762k) Normal Sharpening (1744k) Minus Sharpening (1734k)
|
|||||||||||||||||||||||||||
|
|
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 at full wide angle, 3x telephoto and 3x telephoto plus 3.75x digital telephoto. (Note that these shots were all taken at the lowest resolution of 640x480. Digital telephoto on the 4700 is only available proportionate to the image size selected (which is really as it should be, in our opinion.) Thus, the 3.75x digital tele option is only available at the 640x480 image size, a 1.88x digital tele is available at 1280x960, and no digital tele is available at the maximum 2400x1800 size.
|
||||||||||||||||||||||||||
 |
Musicians Poster (1744k)
Again, we shot samples of this using a variety of white balance settings. We found the auto (139k) and daylight (139k) white balance settings nearly identical and a little too warm. The cool white fluorescent (138k) setting produced bluish results and the daylight fluorescent (139k) setting was too warm. Thus, we chose the warm white fluorescent (139k) setting because it produced the best skin tones and overall color balance (we're still a little surprised at this outcome because we aren't shooting with fluorescent lights in the studio, but rather a carefully daylight-balanced source). Color saturation is good, as the model's blue robe looks just about right. Resolution and detail again look a little soft, judging by the subtle detail of the bird’s wings and the tiny silver threads on the Oriental model’s robe which aren't as crisp as we've seen (this does get a little better with the plus sharpness setting, see below). Again, we'd have to say that the 4700's resolution is that of a good ~2.5 megapixel camera (IOHO). Below is our normal resolution and quality series.
|
||||||||||||||||||||||||||
| Sharpness Series
We again shot with the camera's variable sharpness settings, with the results below. Plus Sharpening (1729k) Normal Sharpening (1744k) Minus Sharpening (1703k)
|
|||||||||||||||||||||||||||
 |
Macro Shot (1687k)
The F4700 does well in the macro category, capturing a minimum area of 2.56 x 1.92 inches (65.05 x 48.79 mm). Nice detail and color, although the larger file size makes the results look soft on-screen: Printed out, they again look like a 2.5 megapixel camera. The flash (1721k) does a pretty good job of throttling down up close as well, although the shiny coin proves to be a little tricky.
|
||||||||||||||||||||||||||
 |
"Davebox" Test Target (1633k)
The F4700 does a good job in this category as well. We shot with the daylight (140k), auto (138k) and warm white fluorescent (144k) white balance settings, again choosing warm white fluorescent as the most accurate based on the white color block and small resolution target (daylight and auto both produced very similar, warm results). The color blocks on the left all look relatively vibrant and accurate, although the cyan, magenta and yellow appear just a tad weak. The red and magenta color blocks on the middle, horizontal color chart also look nice, as the F4700 clearly distinguishes between the two (many digicams have trouble here and try to blend the colors into one). The subtle tonal variations in the Q60 chart are also distinguishable, with the "B" range in the pastels completely visible. Detail in the shadow area of the briquettes is a little limited and a moderate level of noise is present, but not too bad overall. Because its large areas of flat color and tone are well-suited to observing image noise, we shot the Davebox with the 200 (1632k), 400 (1647k) and 800 (1639k) ISO settings. The results were consistent with our other ISO tests, with the noise level increasing in steady progression from 200 through 800. Again, our assessment was that the noise level at ISO 400 is very tolerable, but that at ISO 800 would be objectionable for most uses. Below is our standard resolution and quality series.
|
||||||||||||||||||||||||||
| Sharpness Series
Again, we shot a series of sharpness settings and were pleased with the results. Plus Sharpening (1610k) Normal Sharpening (1633k) Minus Sharpening (1594k)
|
|||||||||||||||||||||||||||
|
|
Low-Light Tests Our low-light tests are always an interesting experience, as they often hold surprises in how the cameras perform. In the case of the 4700, the surprise was not that it did quite well, but that it did so in a way that we really wouldn't have anticipated. It turns out that in "normal" exposure modes, the longest shutter time it will allow is 1/4 of a second. As a result, low-light performance in all but "Night" modes was very modest, even with the very high ISO 800 light-sensitivity setting. Minimum usable light level was about 4 footcandles at ISO 200, 2 footcandles at ISO 400, and 1 footcandle at ISO 800. (Actually, these numbers are all a little bit generous: A more conservative set of ratings would be 8, 4, and 2 footcandles respectively.) As we'd have expected from our other tests, the ISO 800 shots were very noisy. What a surprise then, to find how well the camera did at only ISO 200, in "Night" mode, with exposures extending to nearly 3 seconds! There was very little normal image noise, although we saw a fair number of "stuck" pixels as we got out to the 3 second exposure time. (A note to the reader: Check out Mike Chaney's excellent Qimage software program. It's a great tool for dealing with the "stuck pixel" problem in digicam time exposures, among its many other capabilities!) What's more, a little of our typical Photoshop "levels" fiddling can take a 4700 image shot under conditions as dark as 1/4 footcandle (~3 lux) and produce a *very* acceptable image from it, as seen here (1207k). The table below lists links to shots taken with the 4700 at a wide range of light levels and ISO settings, as well as in "Night" mode. Really though, Night mode wins hands down: Don't even waste your time with anything else unless the light level is at typical residential levels or above. (By way of reference, a typical city street scene under average street lighting is about 1 footcandle (11 lux) of illumination, while the rather bright residential interior of our Indoor Portrait test is around 32 footcandles (~350 lux).)
|
||||||||||||||||||||||||||
|
|
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 model will have similar tests available). Fuji rates the F4700's flash out to a maximum of 13.1 feet (4 m) in wide angle and out to 9.8 feet (3 m) in telephoto under the normal intensity setting. We shot this test using the 200, 400 and 800 ISO settings and found that a noticeable loss of brightness began around 10 feet at the ISO 200 setting, agreeing with Fuji's own rating (we shoot this test with the lens set to the telephoto position). With both the 400 and 800 settings, we found the flash to be effective out to 13 or 14 feet, with increased brightness levels from 10 feet on. (Based on the ISO 200 performance, the flash range at ISO 400 should be right about 14 feet, while at ISO 800, it should be about 20 feet.) Below are test shots taken at each ISO setting, and ranges from eight to 14 feet.
|
||||||||||||||||||||||||||
 |
ISO-12233 (WG-18) Resolution Test The FinePix 4700 Zoom did well in this test, but as expected from our other shots, gave a resolution performance very characteristic of a ~2.5 megapixel camera, not a 4+ megapixel one. We scrutinized the test images closely, and "called" the resolution at ~850 lines per picture height horizontally, and 800 vertically. As predicted based on its photosite layout, Fuji's SuperCCD indeed performs better when resolving elements with strong horizontal and vertical spatial frequency components than elements having a 45 degree orientation. (Check the diagonal resolution element at upper left on the target.) Resolution on this target is indeed higher than other 2.1 to 2.5 megapixel cameras we've tested in the past, but not dramatically so. There are also some image artifacts at a 90 degree orientation to the main axis of the test elements that aren't present in most other cameras. While these will only appear in images of subjects with very high-frequency, regularly-spaced horizontal or vertical elements, we did feel we need to call attention to them. The bottom line? An excellent 2.4 megapixel performance (at the top of the 2.1-2.5 megapixel field), packaged in a larger file size.
|
||||||||||||||||||||||||||
 |
Viewfinder Accuracy/Flash Uniformity
We found the F4700's optical viewfinder to be rather tight, showing only about 76 percent of the final image at wide angle (294k) and about 77 percent at the telephoto (302k) setting. (We're changing our former nomenclature: Formerly we would have called this accuracy error "loose", but now describe the action of the viewfinder on the scene, as opposed to the result in the final file.) Images framed using the optical viewfinder tend to show quite a bit more of the subject in the final image than what you saw in the viewfinder frame, and significantly shift the final image towards the upper left corner. The LCD monitor was much more accurate showing about 96 percent accuracy at wide angle (292k) and about 95 percent accuracy at telephoto (287k). To give you an idea of what we're looking for here, we generally like to see the LCD monitor as close to 100 percent accuracy as possible. Distortion on the F4700 was somewhat high, as we picked up a 1.6 percent barrel distortion (very noticeable) at the wide angle end and about an 0.3 percent barrel distortion at the telephoto end. Chromatic aberration is present but low, we caught about two pixels' worth at wide angle and telephoto. (This distortion is visible as a very slight colored fringe around objects at the edges of the field of view on the resolution target). By contrast though, chromatic aberration was quite low, with only one pixel of color appearing around resolution target elements in the extreme corners at the wide angle setting, and none at all in telephoto mode. Flash uniformity is pretty good overall, with some falloff in the corners at the wide angle setting, none at all in telephoto mode. |
Follow Imaging Resource: