On Wed, 19 Mar 2003 03:13:56 -0500, "Sanman" <me@you.com> wrote:

 

>But what I don't understand is how 2 lenses can output the same amount of

>light, even though one of them has to spread its light over a larger CCD

>area.  On an old film projector, when you turn the front lens to make the

>picture larger, it gets dimmer.  If there were a CCD instead of a screen,

>that CCD would be getting less light per pixel.  OK, the CCD and its pixels

>are large, so it is more sensitive, but doesn't that just compensate for the

>dimmer projected image?  The light sensitivity of this system would not

>change with a larger chip.  It would stay the same.

>

>Sanman

 

Sorry I didn't catch the "import" of this, from your earlier

post, when I answered it last night... (more, below):.

 

>I will admit that most of my optics knowledge is based on telescopes and

>monoculars, and cameras are a little more complex than those items, but if

>I'm buying a camera, chip size is NOT the be-all and end-all when it comes

>to light sensitivity.

>

>Sanman

 

You are doing a reasonable thing: observing things, deciding

what they mean, and basing opinions on related issues on

this. Unfortunately, you have missed some basics (and the

"quirks" that have mislead you), leaving you with incorrect

assumptions that still do work in the specific instances you

have observed (see the end of this long post...).

 

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

OK, the basics (or an attempt at them...;-).

--A lens is defined by its focal length (the distance

from its optical center {often, but not always where the

diaphragm is] to the film plane when correct focus is

achieved with an infinite-distance subject) and its

aperture (which is the clear area of the lens measured

when observed from a large distance from its front, along

the axis of the lens). These are the VERY basics...

(example: 100mm FL lens with a circular aperture 50mm

across).

--The lens has a "relative aperture", specified as

the lens FL divided by its aperture diameter (example:

100mm/50mm=f2, the maximum stop available for the lens.

Since this is a simple ratio, it is true that another

lens and aperture combination, or many, can have the

same relative aperture (examples: 50mm/25mm=f2,

200mm/100mm=f2, 10mm/5mm=f2).

--Most lenses are fitted with a "diaphragm", which allows

adjustment of the amount of light passing through the

lens, and this is calibrated so that as its area is

changed, a commonly-used set of names ("f-stops")

is used to describe the changes. These are based on

the same formula, FL/diam.=f-stop (example: 100mm

f2 lens has its diaphragm diameter cut in half, so now

100mm/25mm=f4). This specifies the brightness of

a given light that passes through the lens, *regardless*

of all other factors (f4 is f4 is f4, regardless of lens

FL or maximum clear aperture [with exceptions

occurring if transmission is unusual, if illumination

isn't even, if the lens does not open as wide as f4, and

especially if focus is radically changed from infinity

to very close - but in this last case, reworking the

figures by adding the FL to the increased lens-center-

to-sensor distance needed to change the focus gives

you the correct new "effective aperture" {example:

100mm f2 lens is focused very close, requiring 50mm

of extension to achieve correct focus, so then

100mm+50mm/50mm=f3, vs. 100mm/50mm=f2}]).

--The lens may be called a "wide-angle", a "normal",

or a "telephoto" (this last is common-use, but incorrect,

since it refers to a specific construction type),

depending on the relationship between FL and sensor

(film or active CCD area) diameter, with the convention

being that angle of acceptance for "normal" lenses with

a given sensor diameter have a FL roughly equal to the

sensor diameter (WA and "tele" have respectively lesser

and greater FLs compared with the sensor diameter [or

diagonal, if made rectangular]).

--Lenses designed for different angles of coverage

(which can be found with simple trigonometry, using

the FL as one side of a triangle, and the 1/2 the side

or diagonal of the active sensor area for the other side

to find 1/2 the angle of coverage of a particular FL

lens on a particular sized sensor) have different

requirements for exit and entrance lens element sizes

needed to avoid off-axis cut-off of light passing through

the lenses, depending on the angles of view required at

each end of the lens, but none of this specifies that any

given lens cannot cover an area different from that

of the intended sensor (example: a given 50mm f4

lens may well be used on a digital camera with a small

sensor area, a 35mm camera, or even a 6x6 camera, if

its design is sufficient - and while the angle of coverage

will change with the sensor size, the relative aperture

[and therefore the relative brightness of the light

passing through] will *not* change).

--The lens may have differing construction types to

best accommodate its intended angle of view and

for allowing the passing of an SLR mirror - and the

main ones are "retrofocus" (used for WAs needing

greater than normal rear clearance - with a diaphragm

placed forward of the "simple lens" position, and with

generally much larger front elements than usual),

"symmetrical" (most often used for normal FL lenses),

and "telephoto" (where it may be desirable for the

diaphragm to be closer to the sensor than usual,

reducing the overall lens length). This, and specific

design requirements, make the front element diameter

a poor guide to "light-passing ability". Another type

of lens is the "zoom", with further optical problems,

including a changing aperture with FL change (the

brightness of the light hitting the sensor changes with

zooming...).

--Lenses normally project a circular area of coverage,

barring physical obstructions.

--Nothing in the above specifies the "coverage" of the

lens (except specific construction details independent

of FL and aperture), which can be larger or smaller

than the sensor it is used with (the latter is undesirable,

since it causes "vignetting", or corner-cutting, of the

image rectangle that is cut out of the lens circle of

coverage by the sensor.

--Lens resolution  can be said to be limited by

various undercorrected optical problems, which are

generally most evident at the widest stop (with

several improved by simple stopping down of the

lens diaphragm), and by diffraction, which increases

in effect as the lens is stopped down, and which is

independent of  the quality of the design and

construction of the lens (it is a universal resolution

limit, present in all lenses if stopped down enough).

Which is to say, the lens performance quality

improves with stopping down (and varies at wide

stops depending on lens quality) until the diffraction

limit is reached, after which further stopping down

reduces the lens performance (though "depth of

field" requirements may lead one to use a stop that

is smaller than optimum for resolution).

--Lens resolution is but part of a "resolution system",

which is limited by (but not to!) the lowest resolution

part of the system. The resolution product is always

lower than the lowest part, but can be improved by

improving any part, including ones that are already

far higher than the lowest. In other words, increasing

the resolution of a lens that is already very sharp,

but is used with a very low-resolution sensor, will

slightly increase the resultant system resolution,

but this will always be at least slightly lower than

the lowest resolution in the system...

--There is much more, but....;-)

      DR

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

 

NOW, for your specifics....! ;-)

For the projector zoom lens, the illumination of the

*slide* remains constant (the lens aperture does change

some, but not enough to compensate as you zoom the

image larger on the screen, making it dimmer (you are

distributing a given light amount over a larger area,

reducing the light brightness). This is not the same as

the reverse, where a constant-illumination *subject*

is projected onto any size sensor that the lens will

cover, with a brightness level on the sensor

determined by (almost...;-) nothing but the relative

aperture of the lens and the subject brightness...

The total light, though, will be greater for the larger

sensor than for a smaller, though the brightness does

not change...

For the binoculars and telescopes, since the angles of

view are so small, the "lens" entrance size can be

about equal to the aperture size (unlike for wider-angle

optics), so "lens size" for these does tell you something

about light-passing ability, but the size must *still* be

related to magnification (giving you "relative aperture"

again, which is hard to escape...;-).