Tariq Gibran wrote:
What still concerns me though is color gamut and accuracy. This shot in particular looks like all the color has been drained out of the subject. Unless this shot has been post processed to give this look, it concerns me a bit.
The photographer confirmed in another forum that the look was achieved in Aperture with some post adjustments "nothing fancy" after converting from raw via Silkypix.
From his flickr stream, this looks like no colour was drained from it!
FlyPenFly wrote:
THe NEX-7 samples posted here for the CV15 look actually pretty good except for the color fringe.
Yeah, I'm finding that color shift is mostly the issue with the CV 15 on the Nex-7, so far.
FWIW, I don't think Sony or Fuji would be purposefully crippling these cameras for M lenses. I think they only really worry about their own lenses, and, if the sensor, AA filter and microlense arrangement happens to work ok with M lenses, it's mostly a happy accident. I think that's what happened with the 5N. Had Sony intended for M lenses to work great, they would have gone all the way, like Ricoh.
I'd not yet seen a view comparing the two this way yet. Visually, no other consideration, I prefer the one on the right.
To be honest I always thought Fuji cameras are ugly
Texas Leica, Natura, Klasse, GA645, even the XPan seems plasticy and utilitarian to me. This one is no exception, and I'm ok with that.
The M6 is obviously a great camera in many different ways, including looks, but it's also the only camera that has managed to give my thumb RSI. My problem is that I have large hands (I'm 6'5") and I only use wrist straps, which meant that I was spending hours holding the camera basically with 3 fingers, and the M6 is simply too small and dense/heavy for this. Add my frantic film advancing action and after a few months my thumb was gone. Leicas just dont agree with the way I like to hold and operate a camera, so now I know that out of these two I'd take the Fuji every day. Gotta pay to learn these things
El Presidente wrote:
Hi - I agree the smearing is not great with the CV 15mm wide open and Fuji X-Pro 1 combination.
I no longer have the M8 but will see if I have any pictures with the CV 15mm wide open
In the meantime, I'll put the CV 15mm on the GXR M-Module and shoot some wide open there as well and post at http://picabroad.com/
**update** - pics taken with the same CV 15mm lens on an M8 and M9 now included
I'm seeing similar corner softness in the CV 15 on the GXR M-Module as on the Fuji in these shots. It certainly would be helpful if he stopped the CV 15 down to F8 to show results closer to optimum for that lens. I wish someone with the CV 15 and Fuji would take the time to perform some meaningful controlled tests.
michaelwatkins wrote:
The photographer confirmed in another forum that the look was achieved in Aperture with some post adjustments "nothing fancy" after converting from raw via Silkypix.
From his flickr stream, this looks like no colour was drained from it!
I guess all his examples are now suspect since the one you post above has also been post processed and we have no idea to what extent. It is good to hear that he ruined the first image in post processing though.
I think we need to differentiate between corner sharpness fall off vs corner smearing with the CV15. On the former the corners will pick up on the GXR with stopping down, with the latter the corners will not pick up when the lens is stopped down.
Me thinks the adapter is not to blame, rather sensor layout/distance/topping/filter etc.
But it is full circle on that discussion, people keep insisting there is hope it is NOT the sensor but just about everything else.
The X Pro 1 is one of the most exciting new cameras in recent years, but it is not the holy grail in cameras, and especially not for alt wide angle glass. That one is yet to come.
I checked the web samples, thank you very much for the effort, much appreciated.
On the GXR 15/4.5 wide open sample (last on page) it is easy to see that it is lens sharpness fall off and not corner smear. My CV15 copy fares a little better wide open than that.
On the X Pro 1 there is a lot of corner smear going on, more than I would be satisfied with. It does not look nice as there is no gentle sharpness roll off, because it is mixed up with corner smear. The wide open samples are simply unacceptable to me, and the stopped down not good enough either.
If I were in the market for an X Pro 1 I would stick with the Fujinon lenses, for better and for worse.
Yes, thanks so much for posting all the F8 shots with wides. The CV 15 @ F8 is mislabeled by the way. In any case, this looks really bad as far as using the X-Pro 1 as a base for wide rangefinder wides and I even now suspect the issue with the Fuji 18 is more likely due to the X-Pro 1 sensor and not as much the lens given how terrible known lenses are performing.
michaelwatkins wrote:
The CV15 has been seen to deliver more consistent edge to edge clarity on some but not all other APS-C cameras; hopefully this isn't a foreshadowing of how other lenses < 50mm will behave on the X-Pro 1.
From my experience with CV15 and the NEX7, better corner and edge IQ can be obtained by focusing for the corners/edge. The CV15 has quite a bit of field curvature. It appears that the focus point is center frame.
One the positive side, the X-Pro1/CV15 doesn't show the color shift that the NEX7/CV15 does.
JimBuchanan;
I don't know about field curvature being that much of a problem on the CV15, to be honest.
Other images I have shot with the CV15 has indicated that the lens has no severe field curvature (i.e. not adversely affecting corner resolution for infinity shots) on APS-C crop sensor. I didn't have the lens when I had the M9. Is your experience with the field curvature on crop or full frame?
Just for the sake of it I put on my slippers, stepped out on the balcony and shot a few wide open and stopped down samples with the CV15. As it turned out even the wide open shots had very good corners as well as centers as indicated by magnified focusing/live view when shot and confirmed on the computer monitor.
I haven't tried closer range or mid distance (mid on the CV15 would be roughly 2<3,5 meters only, before reaching infinity) but for infinity shots I did find that it would not make a difference.
Samples converted but uncorrected, resized, not color matched, with subsequent mauling by Flickr conversions. Sorry. Crops are at 100% and all the way out on the image and unsharpened.
If this is what you can get with a random off the shelf CV15 and any other random wide angle I have put on the GXR then I can only assume the optics are not at fault when comparing with other camera platforms, in this case the X Pro 1.
Thanks again El P for taking the time to make and post these and the other series. It's a delight to be able to see the same lens on four bodies, plus the others you've shot on the X-Pro 1.
The CV 15 shot on GXR is similar to what I am seeing from mine at similar distance, focused at infinity; the M8 in the sample does look better than on GXR, but its distance is also further away.
For comparison, the ZM18/4 wide open, more subdued light.
(posting hickups...)
Maybe it's just me being obtuse, but I don't experience the same problems as many people seem to do, what with field curvature and sagging corners on wide angle RF lenses. Back in the films days I expected RF wides to be sharp all the way (granted, some needed stopping down, but this was full frame too, mind you) and on digital I know most camera/sensor implementations have trouble coping with these wide angle lenses. There are a few exceptions but I don't find this a surprise and I don't look for plausible explanations to that, other than the simple fact that the sensor/camera platform simply isn't designed for the legacy RF wides by default.
Blaming the adapter and lens sample variation all the time gets old when the major problems stem from current technology as implemented in camera sensor design and the simple fact we alt users are not the highest priority. We want it so bad and every new offering is another straw to cling on to, before the straw snaps.
