http://www.flickr.com/photos/jimbobuk/7658911146/in/set-72157630781100484

This shot I think shows the curvature of sharpness quite clearly. The centre in focus trees are fairly distant, the hedge top and fence is perpendicular and square to the lens. It is soft at the centre, sharp at the right edge, and even the top of the building on the right has some sharpness on its roof, even though this is really quite close to me.

I've written more notes in the image description, but the amount of curvature on this "plane" of focus is rather extreme, I almost imagine visualising it like looking into a bowl, with perhaps quite steep walls.

Dofmaster shows that lens at f1 on the gh2 is hyper focal at 20m, with sharpness from 10m to infinity. To get this degree of unfocusedness at the corner would need to have some serious curvature or other defect to overcome such a large wedge of what should be focused.

Anyways, hopefully they're all the same in this respect, and let's face it we will be shooting 98% of stuff wide open and in that sweet spot of 0.5m to 2m where the shallow DOF is most noticeable and causes the subject to pop out of the scene in such a nice 3d way!!

@nomad - eek. If it's not normal i'm going to be pretty miffed. 2 lenses, i thought would imply that its pretty much just a feature of the lens.

I don't think a large evaluation chart will let you test anything other than curvature up close.

Did you try with the steps i outlined in my earlier post, just using live view and something 100m+ away, as close to actual infinity as possible... focussed centrally, out at the edges and fixed by focussing beyond infinity.

You're making me question whether or not my GH2 could contribute. I have access to an olympus pen, i'll see if can try the same test on that to see if it has the same results.

I'm also going to email Voigtlander directly. They should be able to say one way or the other whether this is "normal" for this lens. If not, i'll presume they could perhaps exchange it and test the replacement, or perform a fix. whatever. And get one that was much better or perfect.

Thanks for trying Nomad, look forward to hearing any more results. Would ask that if you've not tried something v distant as outlined before that you could try.. the key is infinity... I don't know how far away a roof opposite your house is :) Use my shots as reference for what i'm talking about for real distant focussed subjects.

Cheers

Thanks Nomad.

Tbh you are pretty much distant enough at that i'd of thought. But perhaps the further you go the more noticeable it is. I'd appreciate it if you could upload some full res pics, though it is a real pain to keep track of what you're doing with no aperture in the EXIF. So you would say that fully zoomed in with the focus assist the subject is focussed (perhaps not quite as sharp, its at the edge of the lens after all) at the edge with the same focus setting as you had with central focussing. Centrally focussed seems to be pretty much on the money for the infinity lens marking for me. Edge focus achieved by going to the rotation stop of the lens past infinity.

I've just had a quick go with the pen, and the issue is much less noticeable.. i'm just looking at the images i got now. I think the main thing though is the resolution difference may make the anomaly harder to see.

Yes, I found a much better location and I can confirm your observations now. What's annoying me is that my sample seems to be badly centered, since the effect is stronger on the right side than on the left.

What some people perceive just as soft corners is more like a curved field on this lens.

I'll investigate this further with a chart too.

Well, I googled around a bit and found that the Canon 24mm 1.4 which got positive reviews otherwise, is plagued by field curvature too. Maybe it's inherent to wide designs with fast speed. BTW, the 1.4 by Canon is about 4 times the price of the 2.8. One remark "The 24 f/1.4 II does have some field curvature which I think is its only notable weakness when stopped down."

Other users complain about field curvature with the Samyang 24mm 1.4 too.

I don't think such lenses are made for landscape, they are made for interior scenes at very low light. Every lens is a very complex compromise, and if you shift it to one side (like very wide and very fast) other parameters suffer. But some seem to like it, like Bjorn Utpott: "Field curvature can be a problem for those who spend most of their time photographing test charts and brick walls; the corners will look worse than on a lens with a flat focal plane. But it's also possible to take advantage of field curvature. In the real, 3-dimensional world, field curvature can be used to increase the apparent depth of field or to isolate a subject. I think Zeiss specifically designed the Sonnar with these 3D effects in mind." See here: http://www.bmupix.com/journal/tag/field-curvature

After all, he's talking about a Zeiss Sonnar!

I like your idea about testing with a curved target, should be possible to print a test chart on stiff paper and place it on a table bent to varying degrees. I won't have the time, though, for the next week, we are preparing a huge workshop right now – including lenses ;-)

@nomad

What he said.

RBD

Add my name to the list of those for whom it's a non-issue.

I paid a lot of money for manual focus fast glass, and used mine for the first time this morning shooting interior video at the local cafe, @ 160.

A really great match to the GH-2.

I'm ecstatic.

RBD

colour me unimpressed but i've had no reply from Voigtlander about this problem.. I'll leave it a little longer before trying again. Not exactly great treatment when you've thrown this much money at a lens!

Still not overly worried about the issue, but would appreciate and official comment on it.

To be continued...

@jimtreats Character versus technical perfection. It seems the two are, more often than not, mutually exclusive.

So true!

Get a Voigtländer Heliar 15mm 1:3.5 if you want incredible sharpness on a plane and low distortion in a wide lens.

It won't shoot in the dark, though ;-)

I had the Voigtländer Heliar 3.5/15 and sold it again. Now I have the Nokton 17.5 and am quite happy with it. Here is why: The Heliar does distort significantly more than the Nokton, despite various claims on the net that it has next to zero distortion. My copy, shot on a GH2, distorted a lot. Secondly the lens is tough for mounting any filters. Version 1 has no filter thread, so you need to manufacture something to mount filters to it. Secondly the controls, though smooth and nice, are very small. For video work it gets fiddly. Thirdly you need to get an adapter. The M39 to m43 adapter that I got on ebay did not allow me to get infinity focus. The new version requires a Leica M to m43 adapter which costs more. It has a 52mm filter thread but with a build in lens shade. So forget about mounting any filter adapter with a larger diameter so you can use any of your larger diameter filters. The Nokton 17.5 does have the flaws as described, but it has less distortion, beautiful bokeh, step-less aperture ring and a normal filter thread. The only thing I find a cumbersome is the very short focus ring travel from infinity to one meter. I am not using the lens for macro, and I think it is not designed as a macro lens which would deliver its best performance in close range. The Nokton just offers the short focal length as a bonus that may come in handy. Last but not least you get an aperture of 0.95 (which is closer to a T 1.2 in reality). This is about three stops faster than the Heliar. f 3.5 is not enough for indoor work, at least on a GH2 due to the noise. The downside is the higher price. If you can't afford the Nokton, get a Panasonic 2.5/14. Or get the cheap 14-42 kit lens. Negligible distortion at 17mm at essentially the same f stop as the Heliar and for the cost of the m43 adapter that you will need for the Heliar. If you don't need the lens now than wait for Photokina for new lens announcements. As for Zeiss lenses I have no personal experience with them but reading about their very high contrast I would be cautious. The GH2 has a very limited dynamic range, so high contrast lenses can cause blown out highlights or lack of shadow detail. For the GH2 lenses with less contrast are better as one can boost contrast in post, unlike recovering blown out highlights.

Cheers nomad.

Voigtlander never bothered to get back to me, but i think its settled that this lens has more quirks than its longer brother. I've got busy with my day job since so will have to wait to do anymore significant shooting with mine.

It's the sharpest at f/4 or f/5.6. That's my preferred f-number at infinity focus.