X1 Full Review posted at DPReview

[barjohn]

Here are the graphs with both lenses set to their optimum f stop for MTF. The Panasonic is clearly out performing it in both MTF and CA.

 

[Terry]

John,

The Panasonic will ALWAYS outperform other cameras on CA. I said it before, Panasonic is removing the CA and baking that into the RAW files. This is not to say that you don't get CA from the lens, you just don't see it. Olympus doesn't correct the RAW files for CA. That is exactly why I put the Olympus comparison in my post. If you are going to make a statement about the lens you need use Apples to Apples. I'm OK with software based corrections and looking at the output you get from the combo hardware and software but in this case you have the ability to look at the lens uncorrected for CA.

Now you talk again about the optimal f stop for the MTF. I thought the optimals were F8 on the Leica and F5.6 on the Panasonic. Your picture is at Leica f5.0 and Panasonic f3.3. Why have you chosen those? Here is the same picture at f8 and f5.6

View attachment 25902

Not sure why the screen print is staying so small but you can recreate this larger by following my link above and moving the scale at the bottom of the widget to the same f-stops I am showing.

 

[barjohn]

Terry, you are right and I missed your point earlier. However, it appears to me that Leica is baking in corrections to CA and performing some noise reduction to their RAW files. There is no way to measure the Leica lens sans sensor and processing since it is not removable. Attached are two magnified views of their noise test at ISO 800 and 1600. The loss of detail in the X1 images compared to either the Nikon or the GF1 is evident. The last image is a raw image magnified from the DPR comparison between the X1 and the GF1 at ISO 100 of their studio scene and the difference in detail is virtually non-existent.

 

[peterm1]

I had the DP1 and the DP1 and absolutely loved the image quality at lower ISOs, but the slow performance of those cameras killed them for me and I sold them. And it sounds like the X1 has similar issues at a super high price. I was really hoping to like the X1, but I think I am going to have to wait for the next generation or wait and see what other companies come up with. In the meantime, my S90 isn't bad at all for a compact that fits in my jeans pocket...

 

[Audii-Dudii]

I was really hoping to like the X1, but I think I am going to have to wait for the next generation or wait and see what other companies come up with.

One of the problems with this approach is that if you don't buy an X1, the X2 you really want may never be made. If the X1 isn't a robust success, then it's likely Leica may decide not pursue the concept any further, as proved to be the case with the Digilux 2 and Digilux 3 that preceeded the X1... :-/

For what it's worth, I preordered an X1 but recently canceled it after realizing the E-P1 / 20mm combo is good enough for my everyday photography needs and the GF1 and G1 in combination with Olympus' F2 zooms are good enough for my "serious" photography needs. It will be a nice change of pace to step aside from the camera wars for a while (a year, maybe, until the next big salvo is fired?) and focus on upgrading my photography instead of my photography equipment.

 

[georgl]

I'm not really convinced by the Leica-lens-performance in comparison to the 2.8/24Asph (which is of course expensive due to it's much larger image circle, mechanics and production scale) for the M, either. That makes me wonder if this is really a Leica-made lens or just a Leica-design with bigger tolerances and lots of mechanical plastic parts sourced from elsewhere... (which is crucial because everybody can put an APS-C-sensor into a small P&S but not design/build a Leica-lens...)

But the samples look better than the GF1 + 1.7/20mm anyway. And another question: how do they measure MTF for the lens (not the whole system) with a non-removable lens...

I love the layout and design of the dpreview-tests but especially the resolution results beyond nyquist always make me laugh (the X1 for example cannot resolve more than 2136lpph)...

 

[barjohn]

They measure the system, lens, sensor and raw processing combination for non removable lenses.

 

[georgl]

That's pointless, at best you can create a system-MTF when you can rule out any internal processing after the sensor within the RAW - they can't do that.