Sony SLR sensor ranks below Nikon, above Canon
DxO Labs added scores for three Sony SLRs to its site for image sensor tests. (Click to enlarge.)
(Credit: DxO Labs)Three midrange Sony SLRs now are included in DxO Labs' measurements of image sensor performance, and the Alpha A700 proves to be reasonably competitive.
Sony's A700, which costs about $1,100 with an 18-70mm lens, has a score of 66.3 on the test, which calculates how well the sensor handles color, a range brightness and darkness, and low-light shooting. That puts it behind the top-scoring camera with a comparably sized sensor, the Nikon D90, almost ties it with the Pentax K10D and Nikon D300, and gives it a a few points' lead over Canon's 40D and 50D.
Meanwhile, the A200 scores 62.9 and the A300 an even 64, according to the DxOMark Sensor test results that were updated Tuesday. A five-point difference makes a difference of about 1/3 stop in exposure, DxO says, meaning that a higher-scoring camera can attain the same raw image quality as a rival even though the higher-scoring camera is using a faster exposure or higher ISO.
DxO Labs, a French company, makes a business of measuring camera image quality, developing technology for image-processing hardware and software, and selling software to convert the raw files produced by higher-end cameras into less flexible but more convenient formats such as JPEG. The DxOMark score measures sensor performance based on the raw file, a foundation for overall image quality but only a facet of a camera's overall performance.
Almost all cameras must 'demosaic' data from the image sensor, which records light in a checkerboard pattern of red, green, and blue, to produce a JPEG image.
(Credit: DxO Labs)The company postponed scoring the A700 because until the newer version 4.0 of its firmware, the camera performed noise-reduction processing on the green light captured by the sensor before generating the raw file. DxO frowns upon cameras reducing noise before the raw file is produced, in part because it misses out on steady improvements in software such as Photoshop that can convert raw images.
DxO's scores haven't been met with universal acclaim. In response to some criticisms and in an attempt to dispel some confusion, DxO has published a boiled-down DxOMark Sensor explanation.
For example, the company has this to say about sensor resolution, an issue that arose when comparing medium-format cameras with large, high-resolution sensors to high-end SLRs: "DxOMark Sensor and resolution are two independent metrics of sensor performance. As a consequence, just because camera A has more pixels than camera B (and thus sees more details) does not mean that its raw DxOMark Sensor score will be better. So before comparing cameras with DxOMark Sensor, it is important to first determine the resolution you are looking for."
The company also offers a more detailed DxOMark frequently asked questions page.
Stephen Shankland writes about a wide range of technology and products, but has a particular focus on browsers and digital photography. He joined CNET News in 1998 and since then also has covered Google, Yahoo, servers, supercomputing, Linux and open-source software, and science. E-mail Stephen, or follow him on Twitter at http://www.twitter.com/stshank. 



I own Sony DSLRs and love them. In many instances Sony makes (or made) Nikon's sensors...so...please explain the comparison?
The overall ranking of all cameras can be found here: http://www.dxomark.com/index.php/eng/DxOMark-Sensor
Panasonic also makes sensors; Olympus uses them for its SLRs.
Nikon has confirmed that Sony does the fabrication of the CMOS imaging array in the D3X. As far as I know the only D-SLR Nikon fabricated their own imaging arrays for was the D2H series of cameras.
CMOS is popular with camera manufacturers for asevera lreasons, not the least of which are lower manufacturing costs and power consumption.
So, whenever I compare sensors, I always click on "print" in order to get a better view of how printouts would compare. After all, any camera above 3mpx will give me good, sharp images on my 24" iMac..
"The differences in performance between the cameras in this comparison are only marginal at low sensitivities (only the E-420 is visibly softer than its counterparts) but from ISO400 upwards the different processing strategies become more apparent. At ISO 400 chroma and luminance noise starts to become visible in the Sony crops although at the same time the camera manages to maintain relatively good detail.
At very high sensitivities the noise reduction is much more heavy-handed resulting in visible smearing of fine detail without really eliminating the noise. You end up with a large amount of fairly unpleasantly intrusive blurred chroma noise blotches. In this comparison the EOS1000D delivers arguably the best balance between image detail and noise reduction."
I note that this generation of Sony cameras is a huge improvement over its predecessors, which were nothing short of horrible in the noise department, and had been getting worse with each generation as resolutions rose. Despite the relative improvement they are well behind the leaders in a lot of respects.
It used to be that Canon was the clear leader in low noise sensors at higher ISOs, but once Nikon decided to switch to CMOS, and design their own sensors rather than sourcing them from someone else (Sony, I think, but I can't remember for sure), they immediately leapfrogged.
I'll consider Sony an also-ran until they get noise under control.
jim frost
jimf@frostbytes.com
http://tinyurl.com/aotlvv
The signal-to-noise comparison looks pretty competitive with Sony A200 vs. Canon XSi (450D) vs. Nikon D60, too:
http://tinyurl.com/apn3j2
I can't comment on what methodology differences lead DxO Labs and DP Review to differ on high-ISO noise control, but there are some data points, for what it's worth.
Fact is...it just doesn't matter, as we're all evangelists of our brands.
Yes its nice that Sony, after buying Minolta, has come out with a powerful line of cameras (ones that are appealing to me) but there are so many good cameras out there, and focusing on just three companies makes no sense
How about a comparo between the CMOS and CCD cameras? Or just looking at the price range of each, utilizing this data (The A700, at $1,100 is close to the D90, but above the now-defunct K10 and $500 less than the D300, and I think the K10 is the only one using a CCD sensor)
- by jmcoa March 12, 2009 4:48 PM PDT
- I too am sick of brand defending or trashing. I use Nikon (film for many years, digital) and Sony (my main DSLR and lenses) and Canon (little P&S sucker). I have also owned every major brand over the years. Leica, Kodak, Pentax, Chinon!, Agfa (4x5 camera), Fuji/Zeiss/Schneider lenses, Sigma, you name it. I think the only camera I did not like was an overpriced Pentax 35mm point and shoot that made dark edged images.
- Like this Reply to this comment
-
(16 Comments)What really matters is your creativity during the photographic process (that's art) and, these days, during post processing (art and some technology).