Right now the photo market is full of a lot of highly competent DSLRs with slightly different specialties. The D700 Nikon digital SLR camera, for example, is a low-light king. The Alpha 900 Sony digital camera is a high resolution monster, Canon digital SLR cameras are famed for having great dynamic range across the ISO range.
And if Fuji can follow through on its promises, we might soon have a new breed of camera starting to arise that has to make no compromises and specialize in one of those fields while sacrificing a little of the others.
Announced recently for a compact camera, but with potential application in larger sensors that has a lot of advanced amateurs watching with interest, is the new EXR technology. At the foundation of this technology is a restructuring of the traditional Bayer filter. The Bayer filter is a filter over the sensor on basically every digital camera going and tells the camera how to see the world in color, and not just values of light.

On the left is a traditional Bayer filter. A row of alternating Red and Blue and then a row of Green (turns out our eyes favor green pretty heavily). On the right is the pattern reworked for their EXR technology. You can see there’s still twice as much green as red or blue, but the new trick is that there are two pixels of each color next to each other at all times.
Further beyond that is that one pixel is high gain, it absorbs light very quickly. The other of the two is low gain, absorbing light slowly.
What’s all this mean, especially in relation to low-light, resolution, and dynamic range? What Fuji is saying is that the sensor can be switched between three different “modes.”
On is resolution. You can tell the sensor to use every pixel traditionally, which gives you high resolution.
The next is dynamic range. Those two pixels, the high and the low gain? The high gain lets shadows expose quicker, the low gain protects highlights. It halves the resolution, but should give a fairly notable increase in dynamic range by treating each type (high/low) as a different image and then combining the data, sort of like in-camera HDR.
The third is low light. Since the two pixels side-by-side see the same color, the camera can “bin” them together. This halves resolution, but creates a pixel twice the normal size, which means the picture will have less noise by nature. It sounds like it can do further binning, effectively letting you get unbelievable ISOs at the cost of resolution, not noise.
Time will tell what impact this has on the market, but as far as photo tech goes I think this is one of the neater announcements I’ve seen in some time.




The Downtown Brain Center of the Photo Processing Operation. 

Switch To Mobile Site