Re: Full-frame or 1.5 DSLR?

In message <1123172046.451278.129260@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
"wilt" <wiltw@xxxxxxx> wrote:

>John,
> << The mention of frame size was to bring out the point that none of
>Canon's high-MP (11 and
>16.7) FF cameras can resolve subject detail as well as the 20D (or
>RebelXT, for that matter), with the same lens. >>

Please learn to quote properly. Your method is messy and irresponsible.
In a medium that is unformatted-text-based, the only acceptable method
of quoting is to preface each line with a special character.

>But the lens has to deliver its resolution to the sensor, at a FIXED
>resolution of the lens. If a lens can resolve 60 lines/mm in air, and
>if the subject fills a 3.6mm section of the FF sensor and a 2,2mm
>section of the 1.6 format sensor,

Stop right there. This is where your train of though jumps the tracks;
they are *EXACTLY* the same size, in mm, on the different sensors!
Imagine a tripod-mount lens on a tripod, inside a dark room, pointed at
the window. You don't need a camera body at all to get an image; just
put the palm of your hand, a *** of paper, or even a dirty piece of
glass in the focal plane, and you will have an image. The device that
you put in the focal plane to focus the image on is irrelevant, as the
the size of the image. When you *do* put a sensor there, the image of
the subject is the same size on the surface; the one with the finer
pixel pitch (350D or 20D, or better yet, D2X) resolves the focused image
to more pixels, meaning the same or greater resolution (it will be the
same if the lens itself is the limiting factor, in the fading tail of
the MTF).

>what the sensor sees is 3.6x60= 216
>lines on FF, and 132 lines on the 1.6 crop. The image is LESS
>resolution for the same subject in the 1.6 crop frame! It does not
>matter that the sensor has more pixels per mm in the 20D sensor, as the
>same section of the total image fills 350 pixels (20D) vs. 490 pixels
>(1Ds MkII). So you lose with 1.6 crop on TWO counts...delivered
>resolution to the sensor and in terms of number of pixels used to
>represent the same subject!

Nope; not at all. You use your longest fl, highest-res lens, and the
20D is going to use more pixels to represent the subject. The 100-400
zoom at 400mm on the 20D is not going to become an 800mm when you put it
on the 1Ds mkII; it's still a 400mm, and it will render the same subject
with less pixels.

><<The lens projects an analog image on the sensor. The pixel pitch of
>the sensor determines how well the lens is resolved. >>
>
>But as I have shown above, 3.6mm x 60 line pairs/mm = 216 lines pairs
>on FF, and 2.2mm x 60 line pairs s/mm = 132 line pairs on the 1.6 crop
>to show the same subject in air. *Less subject resolution on sensor*
>for the 1.6 crop with the same 60 ll/mm resolution lens on both
>cameras.

No; you are assuming things which are not real. You do not get a
different size image ion the focal plane by changing camera bodies with
different frame sizes.

>For a 4x5 print, you would enlarge the FF by 4x, which results in
>216/4.4 = 49 line pairs/mm of subject on paper. You would enlarge 1.6
>crop by 6.7x for final 132 / 6.7 = 20 line pairs of subject on paper.
>*Less final subject resolution on print* for the same magnfication
>print from both cameras.

No. When you have a larger frame, you will be cropping away more of the
image as with the 20D, and you will wind up with less total pixels. A
large frame full of grass with a dot subject in it somewhere is usually
totally worthless. You are thinking more about the number of pixels
than what those pixels contain. Many pixels can be totally worthless
garbage.

--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@xxxxxxx>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
.