The sensor size of the 50d is 3.32 sqcm and the 5d MKII is 8.64 sqcm thats 2.6 times bigger that has amazed me, and its got
full HD video and 22 mil pix the bit that lets it down for me is the frames per sec 50d is 6.3 5d MKII is 3.9
How much does this full frame sensor give in overall Quality above the 50d
I don't know why you are amazed at the difference in area. With a crop factor of 1.6X the area should differ by 1.6X1.6 = 2.56X, as you have found. It's simple maths and not amazing at all.
As for the IQ difference, it really depends. I don't have a 5D2 but look at it this way - the pixel density of the 5D2 is equivalent to the pixel density of the
30D. Let's say you have a 400/5.6 lens and you are struggling to fill the frame with your subject on your 50D. If you switch to the 5D2 with the same lens all you will accomplish is a load of extra wasted space around the subject. The sensor will be bigger but the image projected onto the sensor by the lens will be no bigger at all. The 21MP will be of no value to you if you have to crop like a mad thing to get a matching composition. Once you have cropped to create equal compositions you will find that the 5D2 gives you only half as many pixels on your subject as the 50D. Sure, the quality of each pixel will be superior in terms of noise and DR, but the detail will be less well defined. At low ISOs I would think the 50D would comfortably outgun the 5D2 with a lens that was not really long enough. At higher ISOs I'm not sure which the winner would be.
To really gain big benefits from the 5D2 you will need a lens with a focal length 1.6X greater than the 50D, to fill the frame to an equivalent field of view. Well 400mmX1.6 = 640mm. Since you can't go smaller than f/5.6 and still retain AF, never mind good AF, what are your options to get close to 640mm? A300/2.8L IS + 2X? A 400/4 + 1.4X? A 500/4 + 1.4X? A 600/4? Whichever option you pick you'll be looking at a bill of around £4,000 and upwards for glass vs the ~£1,000 for the 400/5.6. Add on the extra £1,000+ for the body and it's going to be an expensive solution to improving IQ.
If, on the other hand, your glass is long enough to fill the frame on the 5D2 then the IQ should be killer, so long as the AF can match the 50D.
Like I say, I don't have a 5D2 but I do have a 30D, 40D, 50D and 1D3. I performed a little test recently to try to determine which camera would give me the most useable "reach" for a small subject at a distance, taking into account detail captured and noise. I left the 30D out of the test but compared the other three at full stop ISOs from 100-3200. I produced three albums showing the results as....
- full frame images, resized to 800x533;
- crops to the size of my subject of interest;
- 100% crops.
The album that matters is the second one, since that represents a real world crop to a composition/subject that is equal across the board. Pixel peeping at 100% is really irrelevant, as is staring at full frame images that don't give me the picture I want. IMO, for any given ISO, the camera that produces the best looking images overall within that second album is the camera to use for perched birds. Once things start moving around then the AF performance and photographer skill becomes far more significant but I can't really engineer controlled tests of those things, so static targets will have to do.
Here are the albums. The images were shot raw and processed in DPP without any fancy editing. DPP was allowed to choose default NR for each image/ISO, as it saw fit. All other parameters (picture style, sharpening etc.) are identical for all images.
Full Frame Album
Crop to Subject Album (the important comparison)
100% Crop Album
I also reprocessed the images in Lightroom, just for the "Crop to subject" framing. All settings were at Lightroom defaults.
Lightroom Album
You can form your own opinions about noise, but to my eyes the 50D shows a clear edge in terms of pulling out detail from the subject. I used furry animals to simulate feather detail but the writing on the CD is the real giveaway. Comparing the 800 ISO examples, it just seems so much more crisp/contrasty in the example from the 50D. If the £800 50D can best, or at least equal, the £3,000 1D3 it's probably fair to say that it can best/match the 5D2 as well, when mated to the same glass.
EDIT : p.s. since the pixel densities are very similar, what might be interesting is to pitch a 5D2 against a 30D at 100 ISO, with the same lens and same subject/lighting,cropped to give an equivalent field of view, and see whether there is an obvious winner. As I don't have a 5D2 I'm afraid that's not a comparison I can perform.