02/01/2008, 4:30pm, EST
Friday, February 1stSony showcases 24MP flagship SLR prototype
Sony on Friday removed more of the mystery surrounding its 24.8-megapixel pro digital SLR by showcasing a prototype model that reveals some additional information about the camera. The Japanese firm refers to the new top-end model only as the "flagship" Alpha but notes that the actual camera will shoot at a slightly lower 24.6 effective megapixels due to sensor limits. Like other Sony Alpha cameras, it should still use Carl Zeiss Vario-Sonnar lenses; it will have at least the option of a lens with a minimum f2.8 aperture and a manual focus override toggle built into the lens.
The finished version of the flagship remains on track for release before the end of 2008 and is expected to sell at a price well above the $900 of the Alpha A350 or the $1,300 of the A700 series; battery grips and other accessories common to high-end cameras should be available on launch.

Filed under: gadgets, digital imaging
Other story tags: sony, Alpha
,
, 4
,
,
,
,
, 
subscribe to comments
for this article
Next step is to reach the dynamic range. Negative can handle 1000:1 contrast ratio. Digital sensors can resolve 256 levels of gray for each colour. I wonder how many more years until we move from 24-bit colour to something more along the lines of 96 bits?
Then again, once we print these images on paper, we squeeze them right back to 100:1 anyway (that goes for photo negative printed on photo paper, too). The analogue, film-quality contrast would still be nice for editing, though.
Odd that it was Sony - huh ?
First, a negative has an overall contrast ratio (luminance) of 1000:1, and not 1000:1 per color channel.
But even if the latter were the case, each additional bit in sampling depth doubles the dynamic range. So while 8-bit/channel give you 256 levels of brightness, 10 bits are enough to give 1024 levels.
The contrast range however only says something about brightness difference between brightest and darkest, but not necessarily about how many different shades between these two extremes can be distinguished. The bit depth talks about that, which means it's a quantity only indirectly related to the contrast.
Anyway, I think people would be hard pressed to notice more than 1000 distinct levels of brightness in a given picture, so 10-bit/channel is sufficient, which would result in 30-bit RGB images.
This however assumes a linear mapping. By the time gamma curves are used, the current 8-bit/channel are adequate for final image storage, while 16-bit/channel are great for processing. Given that you can combine multiple exposures into HDR digital pictures, the time when digital outdoes analog has long begun.
All that's missing is a camera that does HDR images by itself rather than as a post-processing step done by the photographer.