Has anyone here used any of the Canon tilt/shift lenses with the original 1Ds? Or, for that matter, with the Mark II/III models? If so, what have your experiences been? Thanks!
Beni wrote:
Had to process a 1Ds file for an order yesterday from a wedding I shot 4 years ago. I'd forgotten how much I hated the files from the 1Ds. ISO 200 and ugly as heck in the shadows.
I've seen so many beautiful images from the original. I'm curious, was the original underexposed? If so, it's no surprise that the shadows could be ugly. Still the case for many bodies, even today..
Daniel Heineck wrote:
I don't agree with most everything you just wrote there. Put as strong a color filter as the 1Ds on a 5DII and then we'll talk. Pure and simple.
Alignment of the microlenses is not a big deal--in comparison to mask alignment in chip manufacturing (e.g. a 45 nm node or so) the minimum feature size in a DSLR chip is huge (>250nm or so). Heck, Kodak just made a new sensor for the M9, amazingly enough it's a full frame sensor, which was poo-poo'ed earlier by Leica as being impossible due to telecentricity.
If Sony, Canon, Panasonic, etc could make a scaled up 24x36 version of their tiny P&S sensors, the game would be over. The little sensors are much much better per unit area.
On a per unit area, the pns sensors seem to be better. Why? Because manufacturers can be a lot more aggressive with the design spec than a ff or even APS-C sensor. The model for a 1/1.8 sensor is much much closer to a cmos design (even if they're making ccd's ) than the big chips are. FF Yield, which is probably not great as it stands would suck even more than it does now using the same design rules/fabrication as a pns sensor.
If you have a look around chipworks blog, I think they have the photon efficiencies and well density per area of these little sensors, and they better the bigger sensors by a lot. Of course, bigger sensors win on pure physics and I'm not going to deny that. Pixel density doesn't matter too terribly much and there doesn't seem to be a lot of cross-photodiode leakage in PnS. I could be mistaken. Higher photon efficiencies would allow ff sensors to either have better iso response or put more aggressive color filters in yielding higher color separation
I don't know why Leica said that going larger than 1.3x was impossible, but that was the official (or am I mistaken) alibi given when the M8 was released: they couldn't get rid of the vignetting. That and the whole m4/3 campaign on smaller sensors. Whoops. Microlenses do a pretty great job of focusing the light into the well, that's all.
