Nikon Coolpix 990 Review
Overall Image Quality - The Praise
What can I say? Nikon have done it again. The 990 produces excellent quality images time after time. Probably its best assets are the great metering system (you can't knock that Matrix metering) and white balance / colour algorithms. The camera still has its share of problems (detailed below) but they're easily overlooked when it simply does what you need it to in the field, take great pictures. Sharpness is good, detail definition is also very good.
There are enough manual controls to feed even the most addictive shutterbug, enough overrides to bring the camera down to its basic levels for the purists, you can disable almost every internal effect (knock off sharpening, no image after effects, manual white balance, manual exposure, TIFF format). Or you can use the internal systems to get a picture out that's great first time and will need very little or no "digital darkroom" correction.
The great thing about the 990 is you can just pick it up and shoot and not worry that the shot won't come out just right... Then as you get more confident you can start to explore the HUGE range of manual features and get a bit more creative.
Nikon have done a lot more than just popping a larger CCD in the camera, they've truly improved on what was an extremely popular (cult like?) digital camera to produce an extremely fine digital camera.
As with all my reviews opinions expressed are my own, the advice would be to download the samples and view them for yourselves.
Purple Fringing (Chromatic Aberrations)
One thing I picked up on in my 950 review was purple fringing near the edges of the frame in shots which contained a lot of contrast. And the same problem exists with the 990. The general consensus of opinion is that this is caused by chromatic aberration, the interaction of the lens with the CCD at wide angles (I may also theorize that some CCD blooming is also involved here). The good news I suppose should be that it's no worse than the 950 which is actually a fair achievement when you consider that the pixel count has increase and thus there are more pixels to define the aberrations. In general these effects can be cleaned up with third party software and aren't visible in the majority of images.
The second image here is our new chromatic aberrations test shot. It's a sheet of black card with a test pattern cut into it which produces very bright regions, against a window. Camera set to full wide angle and deliberately overexposed 3 stops to make the aberrations more visible.
Barrel and Pincushion Distortion
The lens system in the 990 appears to be pretty much unchanged from the 950 (give or take a few millimeters) and thus still suffers from a similar amount of barrel (a spherizing of the image at wide angles) and pincushion distortion (a pinching of the image at telephoto) these effects can be removed afterwards with third party software and techniques.
Apologies for the slightly amateur looking target in this test, but it's adequate to calculate the amount of distortion encountered at full wide and full tele. Distortion calculated as the amount of distortion to the horizontal line (from left or right to its center) as a percentage of image height.
|Full Wide Angle, 1.1% Barrel Distortion||Full Telephoto, 0.9% Pincushion Distortion|
I've received several requests from readers to do a quick comparison between the amount of noise generated by the 990 for longer exposures compared to the 950, so here it is. The first two shots were taken in a dark room with the lens cap on. Eight seconds is a long exposure (for a digital camera) and both cameras produced what to the visible eye is a black frame, using Photoshop to increase the brightness 10 fold we can see that the 990 does indeed produce a little more low level noise than the 950, but these tests should be taken with a pinch of salt, they're pushing the cameras to their absolute extremes considering that they're performing fairly well.
Crops taken from the center 240 x 180 of each image. Shutter Priority mode was used to lock the exposure at 8 seconds, on the 990 the sensitivity was fixed at ISO 100, on the 950 you can't manually alter the sensitivity in shutter priority mode I measured it from the EXIF headers as ISO 112.
|Nikon Coolpix 990, 8 seconds, 10x brighter||Nikon Coolpix 950, 8 seconds, 10x brighter|
The next shot was taken in a darkened room, measured light was 1.8 EV (flash meter). The 950 under exposed the shot because of a higher aperture and lower ISO. From a noise point of view there is some visible on the 990 shot but it's very very slight, there are also visible stuck pixels from both cameras, more visible on the 950 shot.
|Nikon Coolpix 990, 8 seconds, F4.3, ISO 100|
|Nikon Coolpix 950, 8 seconds, F5.6, ISO 80|
The following shot was taken from a tripod with an EV compensation of +0.3 EV. The image on the right has had the noise removed using a dark frame technique. Put simply take a full black frame (lens cap on) at the same length exposure, in Photoshop paste this frame as a new layer and give it a slight Gaussian Blur (0.3 pixels), then change the Layer Options to "Difference" and voila the noise will disappear (well not quite, it'll leave tiny holes but be less visible).
|Nikon Coolpix 990, 8 seconds||Fixed in Photoshop using a black frame|