<ralford@bigfoot.com> wrote:

>David and David, et al.

>

>First, my concept of a lens isn't derived from "intuition" :-) but on a

>simple and well known figure that is shows the geometrical optic rays from

>an infinite light source. These generally are used to explain telescope

>etc, al la Galileo. A little search on geometrical optics turned up a

>really cool site, imho:

>http://micro.magnet.fsu.edu/primer/java/microscopy/variablelens/index.html

>with an interactive lens ray tracer! Fascinating but it only traces

>parallel rays.

>

>The issue becomes more complicated in that we are not imaging parallel rays

>form infinitely distance points. A relevant figure for that situation can

>be found midway down

>http://ublib.buffalo.edu/libraries/projects/cases/zoom_lens/zoom_lens.html

>indicating the imaging of an object. The figure indicates the imaging of an

>object on the focal plane. It is clear that the same lens cannot be used

>for different sized ccd's.

This does not follow...

While most lenses for small formats are designed to just

barely cover the format they are designed for, they will

also (obviously...;-) cover smaller formats with the image

plane placed similarly. For large formats, it is common to

design lenses with extra coverage, and the photographer may

then choose to use a lens on, say, 8"x10" film (with little

sideways movement possible [in our example]), or on, say,

4"x5" film, where the angle of view on film would be

reduced and the permissible side-motion increased (for

"perspective control"), or even on 35mm film... On

these different formats, the lens "speed" does not change

(a given f-stop remains constant in effect for all formats

the lens is used for), but one could say the "unit-area

light exposure" also remains the same, and therefore the

total light hitting the larger film area is greater...;-)

But, as I pointed out earlier, the lenses used with

different-sized imagers, once "equalized" (which is the

only reasonable thing to do, if one is comparing only

CCD size for sensitivity, and not whole systems...),

then become irrelevant to the comparison... Or, if

AX + B = 3AY + B, then X = 3Y, and B and A become

irrelevant to the comparison...;-) It appeared that you

were trying to make the lens "do tricks" with light

intensity (and cones, etc.), which, uh, well..., errrr....;-)

>A slightly different (and better discussion) can

>be found at http://cord.org/step_online/st1-3/st1-3frameset2.htm under

>chapter 3C. What isn't clear is why the light density on any ccd isn't the

>same if the same sized lens could be used.

This would not be clear to me, either...! ;-)

I don't believe this would be true.

>The more accurate figures don't resolve the ccd size/sensitivity issue.

>Because the same total incoming object light area is focused across the

>total image area of the ccd, assuming the same lens opening and f-number.

>( Neglecting the change in shape to focus on a different sized image area.)

>

>The answer, mention numerous times, likely involves the improved quality of

>the ccd elements which is not a simple linear relationship with the area of

>the ccd element.

Ah, but this is changing a condition other than size!

If we take a 1/4" "HAD" chip and compare its sensitivity

with a 1/3" IIT chip, then we are not comparing sensitivity

effect of size only (which is why we specify *all but size*

as being equal - otherwise the comparison rules are changed

in a way that makes the original comparison of sensitivity

differences with size alone more difficult...;-).

>Interesting, but as I have pointed out, more of academic interest than

>practical. Now to a more pragmatic problem - how to edit in this new

>program....

>

>Cheers,

>

>Richard

>> I think "RA's" intuitive lens concept is wrong - and

>> as I pointed out in the "unfocused" thread, once all

>> the reasonable specifications of lighting and lens

>> characteristics are made, for "apples and apples"

>> comparisons, the lens and subject lighting can be

>> dropped from "the equation", and a single even

>> diffuse source could be used to illuminate the CCDs

>> for comparison of sensitivity - and then it is obvious

>> that the larger area of the equal-numbered sensors

>> of the larger CCD will give it greater sensitivity...;-)

>> David Ruether