OK. Here is what I am saying. On a 30D you have a sensor that 
is 22.5mm x 15mm with 8.2 million effective pixels.
The total area of that sensor is 337.5 square mm. 

On a 5D you have a sensor that is 35.8mm x 23.9mm with 12.7
effective pixels. The area of that sensor is larger at 855.62 square mm. 
However,
the 5D has an area that is 2.5 times greater than the 30D yet has only 55% more
pixels. So the density is less. 

If you look at the newer Rebel XTi with 10.2 effective
pixels in a sensor that is slightly smaller at 22.2mm x 14.8mm you have an area
of coverage for an image of 328.5 square mm. This means again the 5D will have
an area that is 2.6 times larger but has only 25% more pixels than the XTi. 
Again,
the density of pixels for a given area is greater in the XTi.

So if I have a small object in the image I would have more
pixels covering that area of the image with a 30D or XTi than I would with a
5D.

Now I am not talking Dynamic Range,
Noise reduction, or anything else other than pixel density. That is all I am
saying.

The main benefit is that the pixels are physically larger on
the 5D enabling it to offer improvements at higher ISOs, hence cleaner images. 
There
is something called pixel pitch which is the size of the pixels on an image
sensor. The 5D is terrific with an 8.1 micron pixel pitch, while the 30D has a
6.4 and the XTi smaller at 5.7. The result is a cleaner image overall with the
5D v. the others when using higher ISOs. 
According to Canon about a year or so ago the pixel pitch should be no smaller
than 6.4 microns. But since then they seem to have found a way to do this since
the XTi has a pitch of 5.7 microns, smaller than what Canon claims, and I quote,
“would result in lower image quality at high ISO speed settings compared to our
current design.”

Peter K


----- Original Message ----

From: Tom Pfeiffer <[EMAIL PROTECTED]>

To: [email protected]

Sent: Thursday, June 21, 2007 2:19:25 PM

Subject: RE: EOS A slightly silly query







> -----Original Message-----

> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Austin

> Franklin

> Sent: Wednesday, June 20, 2007 5:31 PM

> To: [email protected]

> Subject: RE: EOS A slightly silly query

> 

> Hi Peter,

> 

> > No Austin. If you and I take an image and I use a 10MP sensor on

> > say a Rebel XTi and that is half frame camera. And you use one

> > with 12.7 that is full frame, and let us say we take the same

> > exact image of a group of people. I will have more pixels

> > covering a small face in the crowd than you will.

> 

> That's not right, and I would really like you to explain the math behind

> that, because if you do, you'll find out it's wrong.

> 

> 

> Austin

> 



Here's the math as I see it, using the specs of a 400D and a 5D as the 10 &

12mp cameras.



Hopefully, we can all agree that if we use an identical 50mm lens on each

camera (mounted side by side and facing the same subject), that we'll get

two identical image circles, except that the lens with the smaller sensor

will capture only a cropped part of the image.



Now, our 12 mp sensor is 36x24mm with a resolution of 4368x2912, which works

out to a nice even 121.333 sensors/mm in each direction. Using THAT density

of pixels, if we were to "chop off" the sensor to make it the same size as

our APS sensor (22.4x14.8mm), it would have a resolution of 2693x1795. At

that size, both sensors would have the same image coverage, but the 12mp

camera would have a 2693x1795 image versus a 3888x2592 image from the 400D.

And it seems to me that 10,077,833,935 pixels in an image is more dense than

one with 4.833.935 pixels.



tomp






       
____________________________________________________________________________________
Boardwalk for $500? In 2007? Ha! Play Monopoly Here and Now (it's updated for 
today's economy) at Yahoo! Games.
http://get.games.yahoo.com/proddesc?gamekey=monopolyherenow
*
****
*******
***********************************************************
*  For list instructions, including unsubscribe, see:
*    http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************

Reply via email to