Okay, it's time to get a little technical... My new Sigma SD-14 camera arrives today! They were $1,400 a year ago, but some retailers have been essentially wholesaling them this spring because the new SD-15 body is coming any day now. I bought mine for just $349 from Cameta Camera in New York. I've dealt with them before on the purchase of my Nikon D300 body, and they always have great prices it seems, coupled with fast service. Just don't expect to hold a conversation with them!
So why am I, a Nikon owner, thrilled to be receiving a $350 camera body I don't even have a single lens for yet? Simple. The Foveon X3 sensor! Right now, there are just three types of digital sensors; CCD, CMOS, and the Foveon. For the CCD and CMOS sensors use what's called a "Bayer filter", which is a screen of microlenses to split the incoming light into it's respective primary color frequencies, red, green, and blue. The internal software in the camera uses logic to deduce the exact average color value to assign to each "pixel photosite" based on the "brightness" at the red, green, and blue sensors. It does not actually see the color, just levels of brightness or darkness. In the process of recombining the data, sometimes there are these little bright blotches in the photos. I call them "bright jaggies". They typically appear at points where the sensor has a difficult time determining the exact color to assign to the pixel. I actually suspect they're a logic error in the extrapolation software, but that's just a theory.
Anyway, once you know what to look for, you can see them everywhere in photographs taken by CCD and CMOS imagers. The troublesome thing is this; they don't really exist. They're the result of camera error!
So why is the Foveon X3 sensor different? Because it doesn't use a Bayer filter! It senses light directly at each photo site, at three different depths in the silicon chip. Blue on the top, green in the middle, and red at the deepest point, because as the longest wavelength, it penetrates the farthest. The software required to output this information is far less complicated than the Bayerized method. Anyway, this "stacked sensor array" is extremely accurate for colors and sharpness, which is why I'm interested. The total number of pixels output from the sensor is smaller than other current imagers, but can be up-sampled in software like Photoshop very successfully.
And, it lacks those "bright jaggies" I hate... This is going to be an exciting camera to use!
Thursday, May 14, 2009
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment