On Tue, 18 Mar 2003 04:36:18 -0500, "Sanman" <me@you.com> wrote:

>The bottom line here is that we can't base a camera's light sensitivity only

>on it's chip size.

Actually, you can if the lens maximum apertures are the

same... If you specify the same type of sensors, the same

number of sensors, and that the sensor size will increase

relatively with the CCD size, you can do just that...

>The original poster avoided a 3 chip model because the

>CCDs were small and he thought that would render the camera bad in low

>light. If you aim a 1" CCD and a 1/8" CCD at the same light source without

>a lens, the 1" inch will obviously gather more light because it's bigger.

Yes. This is "the bottom line"...;-)

>But that's not how CCDs are used in cameras. The CCD is not free to gather

>all the light around it. It just sees a focused square of light patterns,

>coming from one direction. The lens is responsible for how the light

>reaches the chip, and the size of the chip determines the focal length that

>the lens has to have.

Yes - but if you remove the characteristics that are

(properly) set equal for both lenses, the lens drops

out of "the equation" (which is what I was trying to

point out), leaving the comparison of CCD

characteristics independent of the lens.

>Let's take two cameras, A and B. Both cameras have the same lens diameter,

>but B has a bigger chip. The optics in the B camera will have to be weaker

>in order to create a larger image to fill the chip surface.

This is not how the conditions should be set up...

Why would you specify that the lenses have the same

diameter??? This means nothing, since it does not specify

the relative aperture of the lenses, which can be the same

for two lenses of very different diameter - and only

relative aperture specifies how much light passes through

with a given light level and subject...

>(Take two

>*same size* magnifying glasses of different strengths. Which one makes the

>larger "focused" image on a wall? The weaker one does, however, the image,

>be it larger, is much dimmer!) If the B camera doesn't have a larger lens

>diameter, the optics needed to properly fill the larger chip will project a

>darker image onto the chip.

You are making the "apples" on one side into "oranges",

unnecessarily, making the comparison more difficult than

it really is. Understand that both lenses can be set to

the same aperture, and that the same illumination for

the same subject would be used for each. Doing this,

the light intensity hitting the CCD is *the same* for the

two different sizes of CCDs, even though the lenses may be

different. Once you see this, you see that the lens issue

should be removed in the comparison...

>Maybe the larger chip is capturing more light,

>but it's getting *less* from the lens in the B camera than the smaller chip

>in the A camera, which has same diameter lens, but stronger lens elements.

No - you would not set up the experiment this way.

It is like saying, which is faster, a Buick or a

Chevrolet - but, BTW, we removed the front wheels from

the Buick...;-) If the lenses are used at the same stop,

and the subject and its illumination are the same, then

the light intensity hitting the two CCDs (or film)

****WILL*** be the same (by definition, assuming the

same transmission efficiency for the two lenses, and the

ability of both to cover their CCDs - and these are

reasonable assumptions...).

>If the B camera had a larger lens diameter (but the lens elements were kept

>the same thickness), then that would do the trick because as you stretch the

>diameter of the lens without changing the thickness, you are in affect

>making the lens elements weaker by reducing their curvature. This makes the

>lens weaker, which is what we need to fill the 1" chip, but it is a larger

>diameter, so it is gathering MORE light.

NO!!!! Learn basic optics, PLEASE! ;-)

Assume that both CCD sizes are illuminated by identical

lenses (one that would cover the larger chip, though),

and that they are set at the same relative aperture.

This works, and is the same as (for illumination) as

specifying a 3mm FL for the smaller chip, and a 6mm FL

lens for the larger, both set at the same stop. The FL

is irrelevant to the illumination, as is the coverage

(by specification) - only setting the same stop and

providing the same light level from the front is

relevant, but once these conditions are met, you can

also remove the lens and simply illuminate the CCDs

directly...

>NOW that we have both cameras projecting different sized squares of light

>onto their chips, but at the same brightness, the B camera (larger chip)

>will have better low light capability.

Yes, as in illuminating both with an even light source,

as I said, and throwing out the irrelevant lens issues...

If we are comparing CCD sensitivities only, then ONLY

the CCDs under the same light should be compared.

This is basic...

>You can't just talk about the CCD alone. The lens plays solo in getting the

>light to the chip, and different sized chips need lenses that will provide

>them with the light they need. Optics play a much larger part in light

>sensativity than most people think. Most low light ratings are derivatives

>of lens type and size, rather than the type of CCD.

This is nonsense (within some moderate variation...;-).

If the maximum aperature of the lens is f1.2, f1.6,

or f2.2 *does* matter for ***camera-system*** low-light

range (along with CCD sensitivity), BUT, when comparing

***one aspect***, all equal-value items can and should

be removed (as in, all lenses for camcorders can be set

to the same relative aperture, regardless of their FL or

diameter, and THIS is what counts in a simple CCD-size

comparison issue...). And, BTW, lenses with "larger front

elements" do not necessarily pass more light than those

with smaller fronts - it is the *relative aperture*

that counts...