FujiFilm FinePix 4700Z Review
Well. Fuji made a lot of claims about their new SuperCCD's, the main one being that a standard SuperCCD with a 2 megapixel CCD could produce an image as good as a standard CCD with 1.6 times the number of pixels. And thus we discovered that indeed the "4.3 M" (as the label on the front of the camera proclaims) does not in fact stand for 4.3 million pixels on the CCD but rather that the camera produces a 4.3 million pixel image from the 2.4 million pixel SuperCCD.
Because Fuji's SuperCCD isn't orientated squarely (its pixels are aligned in a diagonal pattern) it's not as straightforward to generate the pixel for pixel image directly from the CCD, the pixels on your monitor for instance are squarely orientated, as are the pixels of ink from a printer. Thus the Fuji algorithm manages to generate 4.3 million pixels from just 2.4 million pixels of "information".
This always troubled me. For some time Agfa sold a range of cameras with interpolation, images larger in pixel count than the CCD, these cameras didn't have many fans. The golden rules has always been you can't generate information you didn't capture. Imagine shooting a very fine wire mesh fence. If one of the wires falls between a pixel then you simply don't have the information to create that wire in the final image, it doesn't matter how clever the algorithm it just won't know that wire was ever there.
And here is the problem with the 4700Z. Yes, the images are big, printed at 300 dpi they'd come out 8" x 6" which is big. But going by what I've found and the test images below show this image would be no better (and in some cases worse) than the identical shot taken with a 2.1 megapixel camera interpolated in a standard manner using something like PhotoShop (bicubic interpolation) or other interpolation algorithms.
Not only that but the 4700Z actually performed worse than my standard test camera, a Nikon Coolpix 950 and significantly worse than a 3.3 megapixel Canon S20. Going by the pixel count labels on the front of the camera your average consumer would indeed be confused. UPDATE: I've had it confirmed by Fuji UK that models in the UK will not have the 4.3 label on the front of the camera and that there will be full clarification (as much as is possible) of how the camera generates 4.3 megapixel images from the 2.4 megapixel CCD (thanks to Fuji UK for that).
Reducing images down to monitor resolutions does give good results (as it does with most 2+ megapixel cameras), and I've no arguments with the cameras white balance and colour dynamics, both of which are excellent. If you're buying a camera for images for the web then the 4700Z would indeed be a good choice. To see that just have a look at the images in the samples gallery, they look excellent at monitor resolutions, to get up close you can click on any sample for the untouched original.
Come on Fuji, you've let down an excellently designed and built with average image quality.
Compared to the Powershot S20 & Nikon Coolpix 950
The following scenes were shot with each camera from the same tripod, same lighting within minutes of each other. Cameras were set to automatic white balance, JPEG compression set to the best setting.
In the first set of samples below each 200% crop was taken out of unaltered images, remembering that each of these cameras has different pixel counts you need to look at the image quality and amount of definition rather than size of details.
- Canon EOS M58.8%
- Panasonic G85/G803.3%
- Panasonic FZ2500/FZ20001.9%
- Panasonic LX10/LX151.2%
- Panasonic GH5 development3.6%
- Sony a99 II15.9%
- Nikon KeyMission 170 and 801.0%
- Fujifilm GFX 50S development28.3%
- Olympus E-M1 II development18.7%
- Olympus E-PL80.1%
- Olympus 25mm F1.2 Pro1.5%
- Olympus 12-100mm F4 IS Pro1.9%
- Olympus 30mm F3.5 Macro0.1%
- Sigma 85mm F1.4 Art3.6%
- Sigma 12-24mm F4 Art2.6%
- Sigma 500mm F4 DG OS HSM Sport2.4%
- YI M12.2%
- GoPro Hero50.8%
- GoPro Karma drone2.2%