Ray,
I have some reservations about your calculation.
To begin with, you seem to assume that DR is the most important factor in image quality, but this is not necessarily the case.
On the Canon the size of each pixel is 4.3 microns. To resolve a line pair we would need to have two pixels, so the maximum resolution of the sensor is 116 lp/mm. Let's assume that we have a very good lens that transfers about 20% contrast at 116 lp/mm, even if such a lens would be hard to find.
At 15 MP the pixel size on the P65 + is 12 microns. So limiting resolution would be 41 lp/mm. We may assume that a good lens may transfer 60% contrast 41 lp/mm. This is achieved by many top class lenses.
If we need to stop down f/13 the Airy ring would be 17.3 micrometer, so diffraction would probably reduce MTF a bit, but not very much. It would act as a weak OLP filter, which the Canon probably also would have. The Canon lens would at f/4 would have an Airy ring of 5.3 microns diameter. So the Canon lens would also be affected by diffraction, even at f/4.
So to sum up.
The contrast between two adjacent pixels on the Canon would be about 20% at f/4 and about 60% on the Phase One P65 on binned pixels at f/13. So the "binned" pixels on the Phase one would be much sharper. You could also say that fine detail contrast would be three times higher.
Best regards
Erik
Erik,
No! no! no! You seem to have overlooked one very significant fact, old chap. In my example where I believe the D60 would perform better than the P65+, at least in respect of dynamic range and by a significant margin, I'm comparing the D60 at base ISO to the P65+ at ISO 800, for the sake of equal shutter speeds and DoF.
How's the resolution of your Sony A900 at ISO 800? A bit compromised? I believe MFDBs are even more compromised at high ISO. Even my Canon DSLRs with CMOS sensors and analog amplification above base ISO, begin to suffer at ISO 800 resolution-wise.
Referring again to the DXOMark website, I see that Phase exaggerates the ISO of the P65+. The D60 at ISO 100 is actually ISO 94, almost exactly spot on. The P65+ at ISO 1600 is actually less than ISO 800, close to ISO 700.
The base ISO of the P65+ appears to be an actual ISO 44. It's downhill all the way from that point on, as one increases ISO. At the manufacturer's nominated ISO 1600 each pixel receives a maximum of 1/16th of full well capacity. Four binned pixels together, adding the individual signals, will receive a maximum signal of just 1/4th of the full-well capacity of one unbinned pixel at base ISO.
I really don't think that such a 15mp P65+ image would be as good as the D60 image in any respect, except for DoF and the degree of freezing of subject movement.
Furthermore, I don't think we'll ever get the chance to see such a comparison. I'm certainly not going to hire a P65+ with camera and lens just to prove my point
.
