Very mixed feelings. I copped an assignment from a Brit hi-tech magazine to cover the Foveon introduction back in 2002. The Foveon hype-machine was working overtime. Getting straight answers was like interviewing slick politicians who make statements not related to the question and act like the question was answered. If their hype was to be believed, we all would have been shooting with Foveon sensors for more than the past half-dozen years—all the major camera companies were beating on their door, or so it was hinted.
In theory, it should have meant the end of the Beyer mosaic that we all use. In practice it was underwhelming, odd colour, clipped colours, short exposures only, no end of problems. Under a narrow range of circumstances it was capable of outstanding image quality, showing its potential but seldom achieving it.
I tried to get an answer about whether the sensor would scale. It was 3.34MP and by 2002, 5-6MP cameras were sprouting in profusion. The answer I got is that "our pixels are the best in the industry..." stuff like that, which did not answer the question. Eventually, someone in marketing got the bright idea to use the sensor pixel count rather than the resolution of the image they produced. Since there were three layers of sensors at any given site, they multiplied by three and suddenly had an industry leading 10MP camera—that still delivered 3.34MP images.
They are still at it, claiming it is a 46MP camera, chasing the medium format cameras with its APS-C sized sensor. If they actually have achieved the potential of the Foveon design, it could very well be worth the asking price. Every pixel with a Beyer mosaic sensor is interpolated with the pixels around it. Anti-aliasing filters are required, which reduce sharpness more. There is no need to use an anti-aliasing filter and the Foveon image does not need to be de-mosaiced. Even at an honest 15MP, it could well be printable at sizes and qualities comparable to medium format sensors. We will see when the reviewers have tested it. A jump from the 4.7MP in the SD15 to an honest 15MP is a huge jump—specially considering how slowly it developed over the past decade or better part thereof.
Sigma lenses tend to be optically very good, though QA issues exist and mounts can leave a bit to be desired. I don't for a moment think that this camera will make or break Sigma. On the other hand, if they actually do have the sensor's problem sorted out, their hype about other manufacturers beating on their door in 2002 just might happen a decade later. It is a brilliant design in theory, but on the other hand, the Beyer mosaic, with all its theoretical weaknesses certainly does deliver the goods.
