On Fri, 14 Mar 2003 13:57:45 GMT, "ralford" <ralford@bigfoot.com> wrote:

>Actually Sanman has an interesting point. Given the same opening area of a

>lens (and the performance f-number) the lens focuses the incoming light into

>a cone the same total flux density in any area as the original light. (first

>order but all this discussion needs.) Given a fixed number of pixels over

>some area of the CCD the same light density would be in the area subtended

>by placing any size ccd at the correct density in the cone. There are

>certainly nonlinearities in the ccd's that may come into play to account for

>the sensitivity vs size.

[...]

Perhaps I'm missing something here, but it appears to me

all issues of lens and subject lighting can be and should

be removed in the discussion of relative large vs. small

CCD sensitivity... If you specify that the lens relative

apertures and the subject illumination are the same for

both cases, and that the lenses are capable of covering

the CCD areas involved (certainly these specifications

are reasonable - and you could go further and specify

that there is no lens, but just an equal and even

[diffuse] illumination of both chips with the same

light...), then the issue comes down to sensor size vs.

output, or sensitivity, where it is fairly obvious that

the larger sensor "sees" more light (having more area),

and between two CCDs of equal pixel count and sensor

type, but different area (permitting larger individual

sensing areas), the larger CCD will be more sensitive...