View Single Post
Old Apr 16, 2012, 6:54 PM   #15
Senior Member
snostorm's Avatar
Join Date: Jan 2004
Location: Chicago Suburb, IL, USA
Posts: 2,770

Originally Posted by VTphotog View Post

Scott: I beg to differ, strongly, with your contention. I still have my 5MP Minolta D7hi, as well as the Pentax *istD. The vertical resolution on the sensors is nearly identical (1920 pixels vs 2006 pixels) I took some test shots a few years back, when both cameras were fairly new, with equivalent focal lengths and under the same lighting conditions, There is much better detail with the *istD. Viewed at normal screen sizes, it is evident, and even when printed at 4"x6", the difference is visible. Sensor size is still probably the most significant factor in IQ.

Hi Brian,

I'm not trying to be contentious, but your test, while certainly valid for your purposes, cannot really be generalized when attempting to compare the differences in IQ in the current larger sensor formats. It's really apples and oranges, IMO.

Your Minolta uses a 2/3" sensor (@ 8.8 x 6.6mm with @ 58.1 mm˛ area and a 3.9x crop factor). Your Pentax D uses an APS-C sensor (23.5 x 15.7 mm, with an area of @ 369mm˛ with a 1.5x crop factor) so the Pentax sensor is larger by a factor of @ 6:1.

The difference between 135mm (24 x 36mm, with an area of 864mm˛) and a crop 645 (44 x 33mm and 1452mm˛) is considerably less (1.68:1), and differences in IQ are also mitigated in some respects by the technical differences between CMOS and CCD sensor design.

Compare 4/3 with APS-C or APS-C with 135mm Digital, (both have @ 1.5:1 sensor area differentials), and the IQ differences are considerably smaller, and technical development levels in the sensors can make for more difference than sensor size. Since the APS-C sensors are more advanced than the other two formats, the gap is wider compared to 4/3 and smaller compared with 135mm Digital.

I don't consider DXOMark scores as the definitive measure of IQ, since image quality is subjective and really can't be quantified, but their tests show that the D800 sensor can produce 25.3 bits of color depth with 14.4 Ev (stops) of DR. The 645D, tested at 24.6 bits of color depth and 12.6 Ev of DR. Given only a 10% difference in resolution (36 MP for the Nikon, and 40 MP for the Pentax), The differences in IQ should not be that significant, from my experience (after going from 6MP to 10, 14, and now 16 MP in APS-C format bodies), unless one's primary aim is very large format prints or high magnification pixel peeping.

Given the relatively small difference in IQ potential on a sensor only basis, and now knowing that Dawg is looking at landscape performance, I'd go with the 645D. Lens performance will make the difference though, not sensor performance. The 645D is a crop sensor Medium Format, so will use the "sweet spot" from all the lenses, making every lens perform better. The 800D will suffer with glass not optimized for digital at the edges and corners since it uses the full image circle. Also, the 645D will be able to use the SR advantage with all lenses, while with Nikon, you'd have to get premium current VR lenses to get this level of performance, and the extreme expense for the glass would close the price gap for the total system quickly.

Personally (and theoretically only), I'd still go with the Nikon, but my performance and size priorities are different. . .


Last edited by snostorm; Apr 16, 2012 at 7:06 PM.
snostorm is offline   Reply With Quote