I don't think that's right Humi. I think its the number of hot pixels that group together. There should still be zero column defects.
Its my understanding that a column defect is a more serious error that can only be fixed by a map of the chip that you use when callibrating your images. The standard for class 1 is defined on the Kodak Sensor Solutions site I am pretty sure. It has to do with only so many hot pixels in a group (6 or 8 as I recall). More than that then it becomes class 2. There is also engineering grade which is lower again. Class 2 is fine for imaging.
The KAF11002 chip (as can others in the KAF family) develop a hot pixel which because of the readout process where the lines are read and then moved down to the line below, causes the hot pixel to drag a white line down the image. This isn't though a defective column.
These lines fade with cooling. By the time you get to -35C they get quite faint. They can be a bugger in processing though.
Class 2 is fine. Darks just need to be accurate (they do anyway).
I had a class 2 and it must have just fallen into class 2 category as it wasn't that bad at all.
Its actually a combination of the 2 but I believe defective columns are the key. I just purchased a new STL1001e and Mike from SBIG was very forthcoming on the type of defects I could expect to see on a Class 2 chip (and defective columns were the key issue). He supplied me with a series of sample images from these cameras so I could gauge the typical position and extent of the errors before I purchased.
If you are taking 'pretty pictures' then these defects are very easy to paint over - more problematic for scientific images though.
Yes, a monochrome will only take a black and white image in native format. To produce a colour image, you need to take images through different filters (usually red, green and blue) and then combine them with an image processing program.
Colour chips take colour images out of the box. The chip has pixels with different responses as if they were covered by an R, G or B filter. These are then combined to form a colour image.
The resolution of colour CCDs are lower than monochrome - with mono, all pixels are devoted to each colour in successive images. In a one shot colour camera, one image serves to provide all 3 so the number of pixels to take blue for an image, for example, are about 1/4 the total pixels available.
I have just recieved my STL1001e CCD with Class 2 chip. It has a partial column defect and this is easily removed with a bad pixel map (averages the value with surrounding pixels). More than fine for precision photometry.