The specifications for the STF-8300M/C say the full-frame image size is 3326 columns x 2504 rows. But the Camera Information from the STF-8300 itself says the full-frame image size is a little bit bigger at 3352 columns x 2532 rows. I'm writing an interface to the STF-8300 and calling sbigudrv.dll from my program. To acquire a full-frame image, do I use the larger 3352 x 2532 size for my image buffer? If that's the case, then why does the marketing information say the full-frame image size for the STF-8300 is smaller than it really is? Thank you in advance.
Check out the data sheet for the sensor: http://www.onsemi.com/pub/Collateral/KAF-8300-D.PDF You'll see there are extra pixels around the edges, and the arrangement is different slightly for the mono vs colour versions of the chip.
Thank you so much for your quick reply, Colin. I understand that there are non-imaging pixels around the perimeter of both the monochrome and color KAF-8300 CCD. When calling sbigudrv.dll from my program to acquire full-frame images, I want to make sure to avoid memory access errors. If the camera itself reports an image size of 3352 x 2532 for the full-frame readout mode, then is it okay to allocate my image buffer as 3352 x 2532 in size without causing a memory access error? In other words, can I readout a full-frame image that includes all the active and inactive pixels without causing an error accessing the camera's internal memory buffer?