Originally posted by Kunzite I'm sorry, but I don't get the "body and mount should match the sensor" idea. [...] Can you explain
The Pentax K mount roughly has a mount diameter which is almost the same as the image circle diameter (43mm). And the register distance is such that the mirror (which has a size proportional to the image circle size) is able to flip up.
As a consequence, mount diameter and register distance are proportional to the sensor diagonal. Therefore, a DSLR body
MADE for APS-C should have roughly 1/1.53^3 or 28% the volume of a body made for FF. It does have 100% its volume. Therefore, in the long run, either the mount for APS-C will change or the sensors will become FF. I opt to keep the K mount and go for FF. In an intermediate time, I can live with APS-C in FF bodies, though.
And I agree, the FourThirds mount specification is flawed (too big). They try to fix it now.
Originally posted by whatever7 Actually FF sense is 10 times more expensive than a APS sensor. This has to do the bigger size cut of the sicilon waffer has exponential higher chance of getting a bad sample.
Here is a link of the guy talking about a 200mm wafer contains 20 FF sensor but you can print 200 APS sensors on the same wafer.
Another legend. In some of my previous posts I derived that an FF sensor is about 3 times the cost of an APS-C sensor. It was later backed by other posts. I don't remember the thread. The guy above is right for the specific layout which is unfortunate because it produces a lot of waste (unused silicon area). The situation is different with 300mm wafers.
The yield rate does decrease with the following law, going from APS-C to FF:
y[FF] = y[APS-C]^2.34
So, an 80% yield rate for APS-C becomes 59% for FF. Note that yield rates for CMOS sensors aren't very bad because bad pixels are tolerated.
The price for electronics (even if it includes a sensor of FF size) will soon be neglegible compared to the mechanical costs of a DSLR camera. Which means that the cost of glass is everything we need to care about.