The S1 Pro produced some of the most pleasing, accurate, natural without being dull colours straight-out-of-the-camera I've ever seen from a digital camera (professional or otherwise). FujiFilm definitely have their colour algorithms sorted. Best of all you have control over these algorithms, if you find the colour a little strong then you can shoot with colour set to "Original" which will produce flatter colours perfect for post-processing. White balance is also very good, in general Auto white balance did well in daylight (sunny or shady) with a nice wide selection of preset white balances for other situations and white balance custom preset for more difficult lighting.
Time after time I pulled up new images and was pleasantly surprised at the colour balance of the image, skin tones are rich and natural but not over saturated, primary colours are bright and vivid and pastels are perfectly toned. But don't take it from me, have a look at the samples below (and the 50 sample images in the samples galleries) - note images below have been reduced in size to make viewing them easier, originals are available in the galleries.
Dynamic range simply defines the range of light the camera is able to capture before it either loses detail in darkness (shadows for example) or blows out a highlight (edges of chromed metals are good examples of this). Most consumer digital cameras only have a 8-bit analog to digital converters, plus their CCD's are not built to have a particularly large dynamic range. FujiFilm clearly stated when they announced the SuperCCD that its design specifically enhances dynamic range..
Metering was excellent, on many occasions I simply pointed and shot, only later examining the image and realising how well the S1 Pro had metered the scene, accurate as not to lose shadow detail but also not to blow-out highlights.. It looks to me as though there are two metering systems at work, the first being the standard Nikon meter which calculates the shutter speed and aperture, the second being the processing algorithms which take the data off the CCD and produce the final image which are always able to ensure the complete dynamic range of the CCD is fully expoloited.
Well I'd not argue with that, samples showed a good dynamic range, however it's a shame the camera only has JPEG and an 8-bit TIFF format, if we'd had a RAW format we probably would have had more opportunity to compensate badly exposed images later in software (as you can with Kodak DCS RAW TIFF's). The samples below (attempt to) show the cameras ability to capture both hard highlights (without blowing them out) and detail in shadows.
The last two crops have had their levels pushed upwards
a little (in Photoshop) to show that the camera is still capturing colour
and detail in shadows.. Unfortunately due to the nature of the SuperCCD
and the fact that it's locked at a minimum ISO of 320 there's also pattern
noise combined with this shadow detail.
Again, a fairly good job of capturing the strong highlights of this chrome object without generating the nasty blown-out pixel effect we see in most prosumer digital cameras. Again, the last two crops have had their levels pushed up to show the detail available in what at first appears to be blackness, and again unfortunately this detail has been added to by pattern noise (why didn't FujiFilm provide a lower ISO?).
Resolution & Sharpness
Here's the contentious area... The S1 Pro has a 3.4 megapixel CCD which outputs 3.07 megapixels of data. Sounds simple if you say it quickly enough, however it outputs a 6.1 megapixel image. This means that it's using a special proecssing algorithm to generate pixels which weren't captured in the original frame. So what of it? Well, if you look at an image from the S1 Pro at 100% (1 image pixel = 1 screen pixel) you'll probably not be that impressed, although detailed, the image won't look as sharp as you may expect but then it won't because you're actually looking at the processed image (pixels which weren't captured). Reduce the image down to about 3.5 megapixels 2316 x 1536, or shoot at the built-in lower resolution of 2304 x 1536 and images are much sharper, smoother and altogether better.
So why did FujiFilm choose to process the CCD output at all? Because of the design of their SuperCCD, it has a honeycomb pixel pattern which has to processed to produce a normal square pattern, FujiFilm's processing algorithms generate pixels between existing pixels to create the final image.
So how best to use the camera? This is a tough one... If the camera had a normal 3.5 megapixel CCD you'd shoot all day long at 2304 x 1536 and be happy with your shots, you could print them at 200 dpi and get a nice 11" x 7.7" print which would look good, if you wanted bigger prints then you'd probably use some photo package to interpolate the image up..
And there you have it, FujiFilm are saying "Here's a big 6 megapixel image, if you want to print it then use it as is and you'll be able to print up to 15" x 10" at 200 dpi and it'll look good, if you want to use it on the web then just reduce the image or shoot in a lower resolution".
Here's how I'd use the S1 Pro.. Shooting for a job where the output was high quality print I'd shoot at 3040 x 2016 (6.1 megapixel, 2.5 MB per image), shooting for a job where the output was the web, lower resolution print or for personal use I'd shoot at 2304 x 1536 (3.5 megapixels, 1.4 MB per image) - remember at that lower resolution you're not loosing any detail as the imager only has 3.07 megapixels. (For the samples galleries attached to this review all images were shot at full resolution 3040 x 2016).
It's enough to confuse anyone, I'm sure that wasn't the intention, but the S1 Pro is still being promoted as a 6.1 megapixel digital SLR (I've seen the ads) and that's what the average buyer will see, which is a bit of a shame really because he's likely to get it home and be disappointed with the 6.1 megapixel output - when the truth is that the S1 Pro is in fact a very good 3 megapixel digital SLR (and that puts directly up against Canon's EOS-D30 and Nikon's D1 for resolution).
Just to throw another spanner in the works.. Don't forget that all other digital cameras currently use the Bayer pattern CCD, this means that each pixel captures just one colour (red, green or blue), colours from neighboring pixels are then use to calculate the full 24-bit colour for that pixel.
To attempt to demonstrate what I'm talking about the images below are shown as both 6.1 megapixel and 3.5 megapixel images, 100% crops taken from both sizes are also shown.
(I think FujiFilm were just trying to make life harder for us reviewers).