Spyro
Well-known
c'mon, most fast lenses at this price range will find some way to screw up point light sources when wide open. Besides I thought the offset microlenses BS is mostly to avoid vignetting?
Mister E
Well-known
It probably is just to avoid vignetting. As for fast lenses having aberrations, sure they almost all have bad coma, but whatever's going on with the X100 isn't coma. From what I've seen it doesn't disappear when stopping down.
rxmd
May contain traces of nut
Which would actually make it more difficult, because if you have a fixed lens you can tailor your microlenses to the sensor/lens combination. If the user can put any odd lens on the camera it takes a lot more engineering to get gut results.
Wadcha
Member
Sorry, but my point is that we've already seen tremendous results from the "cheap" Sony sensor in the NEX which to my knowledge is not esoteric in any way. The difference between M8/M9 comparisons with the NEX that are up on Steve Huff are negligible and I would wager that they are due more to the variance in crop factor than anything else. I would expect a full frame sensor to outperform an APS-C sensor.
But the argument that Fuji would have to spend ridiculous amounts of money on R&D or outsource the sensor manufacturing in order to make a system camera is nonsense.
Wadcha
Member
Where exactly are you getting your certainty from?
Fuji has been recorded as stating the sensor is their own. I'm not sure if I believe that but I have no confirmation of the contrary - please share with us if you do.
I also thought the X100 trounced the D90 sensor in the DXO Labs analysis. EDIT - I just checked DXO and the two cameras both scored a 73 so I can see where you may think they are the same. Did you have any other confirmation of this?
Last edited:
rxmd
May contain traces of nut
I guess it's mainly economies of scale, plus designing lenses for a smaller image circle. The approach in a compact camera is different. Fuji is designing a smaller lens, resulting in a compact, small and fast 24/f2. The smaller image circle makes a real difference in lens design - for comparison, a Leica 24/f2.8 is slower and costs twice as much as the whole Fuji camera; a Biogon 25/f2.8 costs about the same as the camera and is still slower; and both are a fair bit bigger and heavier than the Fuji lens. Whatever shortcomings the lens optically has, you can stick microlenses in front of the sensor and build extra optimizations into the sensor output processing logic. All these figure into the R&D expenses for the whole system, but by virtue of not having an interchangeable lens the whole system can be made both smaller and better.
The main difference between M8/M9 and NEX in those tests of yours is small is because the lenses they are testing it with are built for a larger image circle. Expensively engineer a lens for a large image circle and utilize only the sweet spot of it, and no wonder you get decent results. Look at what the NEX delivers for wideangles under 20mm with smaller image circles - suddenly you start to get vignetting. Some people may like that for its pictorial effect, but technologically impressive it's not.
If Fuji makes an interchangable lens model, they will most certainly make their own lenses and mount.