Michael: 1Ds crop vs 10D?

[Joe Hardesty]

Isn't the whole point of such discussions to make one feel that they spent less and got more? .

Joe,
No it's not. It's about being objective, truthful and fair-minded. There are a lot of status seeking activities in this world which fudge the issues and cause confusion and trouble.

The beauty of technical discussions, is that they hopefully avoid all this crap.

OK Ray, I can accept that, but I am still unclear about what such as discussion could/should achieve. Is the question purely hypothetical/academic, or does it have a practical application?

I am honestly not trying to be contrary or argumentative. I simply don't understand why it matters.

[jwarthman]

If I was a newbie to The Luminous landscape, and trying to learn something about DSLRs in general, or the 10D and 1Ds in particular, and posted a question, and received a two-word reply of:

   "who cares?"

you can bet I'd be offended!

We all bring our own background to these discussions. I think we should be more tolerant to other folks who may not have the knowledge about the topic that we've accumulated over many months or years. We might probe a bit more, if we need to understand the basis for their questions. Or, if it's just something we don't care to address, our silence would be less insulting than a terse reply like "who cares".

*stepping off soapbox*

Enjoy!

-- Jim

[BruceK]

I'm tiptoeing into this thread with information that is probably in another thread elsewhere but I was too lazy to look.    :O

I took the CCD specs for the 10D and 1DS from the Canon data sheets and proceeded to try to guesstimate what the pixel size was for each camera. I did this by taking the sensor size and dividing the horizontal and vertical dimensions of each sensor by the number of effective pixels in each dimension. For the 10D this results in a pixel size of roughly 0.0074mm square; for the 1DS they're about 0.0088mm square.

I then took the sensor size of the 10D and figured the number of pixels that the 1DS would have in the same dimentions and wound up with 2580 x 1716 pixels. The 10D has 3088 x 2056 pixels in the same area. This would imply to this unworthy individual that you might be able to squeeze a slightly bigger print out of a 10D photo, all other things being equal (which they won't be).

Running for cover...

     Bruce

[BJL]

Is the question purely hypothetical/academic, or does it have a practical application?

Here is one possible application for a certain group of photographers: you like photographing birds, and to that end have bought the longest, fastest lens that you can afford; but often it is still not enough and you have to crop substantially: which camera does the best in that situation?

My "geek" answer: the one that has good sensor MTF (say 50%, following Norm Koren) at the highest possible number of line pairs per millimeter. So maybe the best choice will be a Sony 828 with telecoverter

[AJSJones]

Hey Ray!
How does your discussion change if you take the anti-aliasing filter into account?  Rhetorical, I know, because there's no information on it. So here goes a bit of speculation...

 I'd assume it performs an (analog) function similar to Gaussian blur, so the photons from each of your grains is spread through several sensor pixels and, among other things, thereby yields color infomation which would not be the case if only one sensor element had been in the path of the light ray (no relation  ).  The AA filter also does what its name suggests by this same "spreading" function.  Therefore, I'd go as far as suggesting the actual pixel size per se is not a very important factor, although we have shown that the spacing (pitch) does indeed affect captured detail with the 10D getting a bit more per sq mm of sensor.  Pixel size, sensor design and fabrication are clearly contributors to the resulting sensitivity/signal to noise etc, so in that sense pixel size is important,  for different reasons.  If all the sensed photons at a pixel have arrived after having been blurred by at least a diameter or two (I'd guess, for the AA function to be worthwhile) of the pixel pitch, it likely won't matter much whether the pixels are full-sized or half-sized, as long as they sample the photons well and have good S/N.

What do you think?

Andy