quote:
I can imaging a scenario where CMOS looses however. This is the BETA vs VHS scenario. 99% of the worlds imager R&D is going into CCD. That makes it cheaper and more rapid advancement towards meeting "consumer" perceptions such as 35mm equivalent sensor size. CMOS may never reach enough economy of scale to beat it.
Yes and no [img]images/icons/smile.gif[/img] For the most part, economy of scale is one of the big reasons why CMOS has an advantage over CCD processes. The same CMOS process used to generate the D30 images is being used for electronics fabrication in a lot of other industries. Every PC, scanner, photocopier, video camera and heck, most of the rest of CCD based cameras, all use dozens of CMOS components.
What is unique is the particular design of the image sensing CMOS chips and generating CMOS die sizes as huge as we are dealing with. However, the underlying process is the same as all of that other stuff and is one of the reasons why CMOS is such a promising technology - especially for companies like Canon with interests in a lot of different areas. Once they refine their processes, any microprocessor fabrication facility in the world should be able to make these imager chips for Canon - as opposed to the small handfull of facilities capable (and experienced) at producing CCD chips. Right now, however, CMOS facilities may be partially renescent of building such large imagers because of the low yields they'll get, and the unpredicability of such an arangement.
As for the quality of the imagers, the inherent differences between CCD and CMOS are much smaller than the differences between the designs of each particular imager family. The D30 doesn't produce great images soully because it has a CMOS sensor, it produces great images because Canon created an excellent imager that happened to use the CMOS process. All that extra logic does in fact all but eliminate noise generation along the readout stage, however it also takes up a larger portion of the sensor leaving less for light detection (and hence decreaseing the SNR and generally increasing noise in the image at any given ISO). Its all a matter of balancing all of the factors to get the result you want - which process you use is a significant part of that, however it is far from the only one.
If people are seeing similarities between D1 images and 1D images it isn't because they're both CCDs, its because they appear to be from the same imager family (likely from Sony). Rob pointed this out in his review, and it does seem to be true. They both exhibit similar noise problems, they both have an 11.5 micron pixel size, they're both Interline Transfer type, etc. The banding problem is a syncing problem that was caused by a bad oscillator in the original D1, so it should be fixed in the final product - but is something small that would be expected in a prerelease version of the camera. For the most part, the 1D sensor seems to simply be an expanded D1H array. Of course, the manufacturer has had several months to touch it up, so there is likely some other small improvements but it seems to share the same fundamental design.
Linear Mode
