Previous versions of brainiac's message #6482085 « Canon 5D Mark II master thread »
brainiac Offline [X]
Re: Canon 5D Mark II master thread
Tariq Gibran wrote:
Brainiac, in order to achieve the expected 29% finer detail, would not the per pixel quality have to be the same between the 5DII vs 5D.
Yes, and to test it you would need to ensure that what filled the 4368 by 2912 pixels in some part of the 5D2 frame was exactly what filled the 4368x2912 pixels of the 5D frame, i.e. a 29% wider lens on the 5D2. It\'s very hard to assess per pixel quality without looking at the same image data, because of moire and psychology, so the best way to try to assess the 5D2 your way would be to use a zoom which had consistent and adequate performance at both focal lengths at f5.6. You would also need to find some way to ensure that you had zoomed to a 5616/4368 increase in magnification, which is quite tricky really. Once you\'ve done all that and pixel-peeped, then you can declare that the per pixel quality of the 5D2 is 3% worse than that of the 5D or whatever, and therefore the 5D2 is really only 27% better than the 5D, instead of 28.6% better and everyone will understand precisely what you mean and will be able to ignore the thoroughly misleading practice of uprezzing crops like the ones you have just linked to, which show very adequately the real scale of the improvement.
And note, btw, that the 5D2 reveals that this van model is called \'expert\', not \'repent\', not \'emport\', and not \'rupert\'.
Dec 13, 2008 at 01:24 PM
brainiac Offline [X]
Re: Canon 5D Mark II master thread
Tariq Gibran wrote:
Brainiac, in order to achieve the expected 29% finer detail, would not the per pixel quality have to be the same between the 5DII vs 5D.
Yes, and to test it you would need to ensure that what filled the 4368 by 2912 pixels in some part of the 5D2 frame was exactly what filled the 4368x2912 pixels of the 5D frame, i.e. a 29% wider lens on the 5D2. It\'s very hard to assess per pixel quality without looking at the same image data, because of moire and psychology, so the best way to try to assess the 5D2 your way would be to use a zoom which had consistent and adequate performance at both focal lengths at f5.6. Then you can declare that the per pixel quality of the 5D2 is 3% worse than that of the 5D, and therefore the 5D2 is really only 27% better than the 5D, instead of 28.6% better and everyone will understand precisely what you mean and will be able to ignore the thoroughly misleading practice of uprezzing crops like the ones you have just linked to, which show very adequately the real scale of the improvement.
And note, btw, that the 5D2 reveals that this van model is called \'expert\', not \'repent\', not \'emport\', and not \'rupert\'.
Dec 13, 2008 at 01:20 PM
brainiac Offline [X]
Re: Canon 5D Mark II master thread
Tariq Gibran wrote:
Brainiac, in order to achieve the expected 29% finer detail, would not the per pixel quality have to be the same between the 5DII vs 5D.
Yes, and to test it you would need to ensure that what filled the 4368 by 2912 pixels in some part of the 5D2 frame was exactly what filled the 4368x2912 pixels of the 5D frame, i.e. a 29% wider lens on the 5D2. It\'s very hard to assess per pixel quality without looking at the same image data, because of moire and psychology, so the best way to try to assess the 5D2 your way would be to use a zoom which had consistent and adequate performance at both focal lengths at f5.6. Then you can declare that the per pixel quality of the 5D2 is 3% worse than that of the 5D, and therefore the 5D2 is really only 27% better than the 5D, instead of 28.6% better and everyone will understand precisely what you mean and will be able to ignore the thoroughly misleading practice of uprezzing crops like the ones you have just linked to, which show very adequately the real scale of the improvement.
And note, btw, that the 5D2 reveals that this van model is called \'expert\'.
Dec 13, 2008 at 01:11 PM
Previous versions of brainiac's message #6482085 « Canon 5D Mark II master thread »
