Megapixels are messing up the digital camera marketplace the same way MHz did consumer computing and ZOOOOMM did consumer camcorders. The problem with consumers is that they're human. Once humans get hooked on a metric (measure) that's all they can handle; humans can only solve single variable optimizations. So single variable metrics drive a consumer market --- at the cost of everything else. Ultimately we end up with something like the "Irish Elk" -- giant antlers, but a lousy Elk.
This article demolishes the megapixel myth (yet again). Most of all, though, it had some very interesting things to say about light sensitivity, and about why light sensitive digital cameras need electromechanical shutters and mechanical diaphragms.
I want to be able to take pictures indoors without a flash. That means the digital equivalent of ISO 400 film and f2.0 optics. Image stabilization would help too.
I figured I'd be able to get that camera next year in a G2 (non-dSLR) form factor. This article has convinced me that's unlikely. I'll have to go for the next generation of the Nikon D70 or the Canon Rebel.
One thing the author omitted is that another advantage of a larger sensor is that it can be manufactured using CMOS-type technology rather than CCD-type technology. My military optics/electrical engineer buddy tells me that at human-tolerable temperatures CMOS sensors are effectively more light sensitive than CCD sensors, but they require more surface area to provide an equivalent resolution. Bigger sensor means bigger camera. Once camera size moves into the range of a dSLR, it's hard for a vendor to justify producing anything else. Emphases mine:
You can fit roughly 16 of the 2/3-inch (11.08 mm diagonal) sensors into the same area as a 35mm film frame – or 2.56 of the Canon sensors. Naturally, in a 2/3-inch type, 8-megapixel sensor, the photodiodes have VERY small surface areas, which means their light-capturing ability will be limited. In dim lighting, this sensor will be struggling to collect photons and the resulting signal will need considerable amplification to produce an image. (And everybody knows that when a signal is amplified, the associated noise is also boosted!)
Because the sensors themselves are much larger, the photodiodes used in DSLRs have greater surface areas that can gather more light, leading to an increase in the sensor’s effective sensitivity without the inevitable noise associated with an increase in gain in the system. The downside is that they cost more to manufacture and take up more space in the camera, which means camera bodies have to be bigger...
...DSLRs use full frame transfer sensors, in which the entire surface area of each photodiode is exposed. This further increases their light capturing facilities but means they must be used with a mechanical shutter that controls each exposure (this also means the inclusion of adjustable aperture diaphragms).
Post a Comment