... it will crop the image so it appears to be closer but really isn't... in all reality. ...It outputs the exact same info it would send to ANY camera but depending on if the camera see's all of the info or not will depending on how the camera "sees" the image.
Also keep in mind given the high density of the 7D,
But that higher pixel density does result in apparent magnification. For example, if you choose FF vs. APS-C with the same pixel size/density, e.g. comparing a 200mm lens on a 5DII vs. a 20D (21 MP vs. 8 MP, but same pixel size), and take pictures of the same subject at the same distance then view them 100% on your computer monitor, that subject will be the same size in both images (although the 20D image will show less of the background). But if you do something similar with the 5DII vs. the 7D, the subject at 100% will appear much larger on your computer monitor, because that subject projected onto the 7D's sensor will cover a lot more pixels.
It appears as magnification the same way a digital P&S applies digital zoom... (but digital P&S sensors are not as sophisticated and up to date in processors) so the digital zoom looks crappy as a result.
I disagree with your analogy - they're different. With digital zoom on most
P&S cameras, the apparent angle of view is narrowed (that part of the analogy holds true), but then the cropped image is interpolated
back up to the pixel dimensions of the whole sensor (i.e. a 10 MP P&S with digital zoom applied will still give you a 10 MP image). It's the interpolation that degrades the image, not the cropping. Processing power of the camera is a factor, in that your computer can use better algorithms (e.g. bicubic in PS) than the weaker in-camera processor can support at sufficient speed. But even done with the best possible algorithms, upsampling is going to negatively impact IQ, whereas cropping without upsampling does not affect IQ.
Note that some P&S cameras can apply digital zoom by cropping only, i.e. that 10 MP image with digital zoom yields an 8, 5, or 3 MP final image instead of the full 10MP. Canon calls it 'safe zoom,' others call it by other names. If the rumors of a merging of 1D/1Ds lines are true, and a FF sensor has a 'crop mode' for APS-H and/or APS-C simulation, that's how it would work - you'd get a cropped image with equivalent IQ (unless you upsampled it in post).
Actually thinking of your comparison, you are wrong in your analogy and i'll tell you why... The 20D despite having the same density will be slightly cropped than the 5D taken at the same distance and same focal length and settings...
Slightly cropped? I'd say more than slightly - it will be cropped by a factor of 1.6x. I was assuming that the subject chosen for the scenario was small enough to fit within the APS-C portion of the image circle.
Now assuming you are shooting a 20D and 7D at the same settings... If you take that file and PRINT it as is at 8x10 for each file adjusting DPI so you are printing the full frame within the 8x10, there will NOT be any magnification, they should be the same because the SIZE of the sensor has not changed.
Sure, but you'd have downsampled the image from the 7D. Here, you're comparing APS-C to APS-C - obviously no optical magnification, but yes, apparent magnification in that the higher resolution image will display 'larger' at a given % size on a computer monitor.
Now if you take the 7D vs 5D mII and use different lenses or compensate so they are the same perspective and relation, the image overall, printed the same way at 8x10 within the same restraints will still be the same image, no magnification. Level of detail maybe a lot closer because the effective resolutions are so similar 18mp vs 21mp and in theory you should have a hair more detail from you 21mp than the 18mp.
I've actually done that
, both with the 5DII + 135mm f/2L
vs. 7D + 85mm f/1.2L
II (same picture, different body and lens) and later in that thread, with both bodies and the 70-200mm f/2.8L
IS II, using the zoom to compensate for the crop factor. The upshot was that the 5DII wins (no surprise), but it's never going to be an apples-to-apples comparison (the 85L is softer than the 135L with both shot wide open, even with the same lens the anti-aliasing filters on the two sensors are different meaning differential sharpness even with the same lens, etc.).