kwalsh
Offline
• • •
Upload & Sell: Off
|
I've watched Foveon for a lot of years, even way back before the stacked pixel design when they were using three monochrome CMOS sensors and a dichroic prism in a studio camera that used a laptop as a viewfinder. It has been a compelling approach from the start and it has been frustrating to watch it never really reach prime time.
In my humble analysis there are four long present problems that stacked upon each other have culminated to make the SD1 as irrelevant a camera as it is.
- First, and probably the most minor on their own at this point, are the sensor limitations. Finally it appears the sensor has both VGA and CDS so the high-ISO performance while not spectacular is also not abysmal like the earlier sensors. That leaves the delta-E limitation which is a fundamental limit of the stacked design but is probably not relevant to a lot of photographers. On its own I think the sensor is now very compelling for some genres. I list it mostly because it dove-tails into the next point and historically it may have had a lot to do with why X3 wasn't adopted way back when by other manufacturers.
- Second, is SPP. Because of the stacked pixel design and its extremely challenging color processing most all of the normal RAW workflows are not an option. The spectral filters on Bayer make for relatively straight forward color, the broad diffusion profiles of the X3 require very specialized processing to get color even close to correct. Because of this you are forced into SPP and SPP is a piece of junk compared to what else is out there. I know SPP can output quality images, the problem is no other RAW processor really can. In order for the camera to be successful it has to be easy for people to migrate to and SPP is a huge barrier to this. But really, part of that barrier is the sensor itself - it is the sensor that prevents easy support in other converters. Bit of a chicken and egg problem - not relevant until better RAW support, no point in the significant investment of RAW support until relevant.
- Third, is Sigma camera design. The cameras themselves are awful. If they didn't have an X3 sensor in them no one would purchase them at any price. Without X3 I doubt the SD1 would sell for $500. Sigma designs horrific cameras and nothing is worse than their firmware. The SD1 time to write files and buffer management is insanely bad. Sure, part of the problem is that the sensor produces a lot of data, but that isn't the real issue - lots of cameras with higher frame rates and deep buffers both write faster and behave while writing. Many of their cameras actually crash - who else have ever used a camera where the firmware regularly crashes?
- Fourth is the management. It is easy to throw stones at management. Foveon got trapped in an awkward spot before their acquisition and ended up down a financially fruitless camera phone path. Sigma didn't move on a new sensor for a long time. Whatever, hard from the outside to really know what was up but the repeated schedule slips and announcements 18 months before product availability clearly point to something wrong at the top levels. The real problem though is the SD1 price. And this is clearly, confirmed by multiple sources, pure management idiocy. There was no reason it couldn't be a $2000 body, costs were very similar to previous bodies but an executive in Japan had a "brilliant" pricing idea.
Sigh. In summary, yes stacked pixel technology has some compromises but in some cases it might be worth the compromises. Unfortunately there are other forces that are far, far, far more detrimental than just the sensor that make the SD1 a non-starter. The various sensor quibbles are the very least of what is keeping the X3 as a fan-boy niche.
Ken
|
