35mm, full frame.
is that supposed to be the upcoming Nikon D3x Sensor?
Not that I care much, but am sure curious to know...
Georg
Posted by David Bostedo on January 30, 2008 at 4:32 PM
That's the rumor that I've been hearing on some forums (and all those blogs I read ). It makes sense, since Nikon and Sony certainly are collaborative when it comes to sensors. The other rumor is that versions of this chip will show up in both the D3X, and a Sony full-frame camera. That would be interesting, because in a short time, the number of full frame digital cameras being produced will have more than doubled I think (from 1DsMKII, 5D --> 1DsMKIII, 5D, D3, D3X, Sony - right?) with 3 competing companies rather than just 1.
Posted by Greg Barteluk on January 30, 2008 at 4:58 PM
Yes, and with Sony and Nikon boasting new 24MP full-frame censors, a minor boost to 16MP for the next gen 5D will certainly seem underwhelming.
Posted by Bill Chambers on January 30, 2008 at 8:07 PM
24MP = Wow! I wonder what the noise performance will be like? Never thought I'd think of 16 MP as underwhelming!
Whoa! That's amazing. Glad I didn't rush out and get a D3 right this minute!
Posted by David Bostedo on January 31, 2008 at 9:32 AM
Well Bill - from what I've read I think the 1DsMKIII has noise performance somewhat better than the MKII, but not nearly as good as the 1D MKIII. (Makes sense, given the pixel sizes.) I'd expect the D3X to not be nearly as good as the D3, noise-wise - more like the 1DsMKIII.
The bigger problems with cameras with this high pixel count is that you lose sharpness due to diffraction at moderate apertures, and they potentially have a strong AA filter to combat moire. Both of these can result in soft, though high-res, images. The 1DsMKIII has been getting some complaints about the strong AA filter causing softness. I wonder if Nikon will try something different?
Posted by Basil Greber on February 1, 2008 at 6:03 PM
David - this sensor will not have much more pixel density than a 10 MP APS-C sensor. So diffraction shouldn't be a huge issue (you may need to stop down one stop more compared to APS-C for same DoF, though - so maybe that sensor is one stop more diffraction sensitive than a 10 MP APS-C sensor). And while you may not always be able to get "24 MP sharp" images, you certainly can get better ones than "12 MP sharp". The lens becomes more and more limiting here - so in a way, you can see it as a device to always fully exploit the lens' resolution potential.
Basil
Posted by David Bostedo on February 1, 2008 at 9:35 PM
Basil - That's true, Basil, but that's still pretty dense compared to the D3 and 1D MkIII. So noise wise, it's probably more like an APS-C than the D3 or 1DIII. Of course, noise is always getting better at a given pixel size, so maybe they've got some new tricks.
Good point on the diffraction limit - guess I didn't think that through. Of course, the D3 and 1DIII would be better in this regard too.
And from what I've heard, top quality lenses can still out-resolve even 24 MP full frame. I don't have any real evidence of that though, unfortunately. And I wonder if the AA filter problem really matters. Maybe someone who actually, you know, can afford one of these things could do some testing.
Discussion
7 repliesis that supposed to be the upcoming Nikon D3x Sensor?
Not that I care much, but am sure curious to know...
Georg
The bigger problems with cameras with this high pixel count is that you lose sharpness due to diffraction at moderate apertures, and they potentially have a strong AA filter to combat moire. Both of these can result in soft, though high-res, images. The 1DsMKIII has been getting some complaints about the strong AA filter causing softness. I wonder if Nikon will try something different?
Basil
Good point on the diffraction limit - guess I didn't think that through. Of course, the D3 and 1DIII would be better in this regard too.
And from what I've heard, top quality lenses can still out-resolve even 24 MP full frame. I don't have any real evidence of that though, unfortunately. And I wonder if the AA filter problem really matters. Maybe someone who actually, you know, can afford one of these things could do some testing.