[New test photos released] Leica Summicron 35/2 Eight Element copy made in China

[rscheffler]

In these two images of a sign in the middle of a park field, the first was rangefinder focused on the sign and shot at f/4. The second was live view focused on the sign at f/4.

In the rangefinder focused image, the area of best central focus has shifted behind the sign and is out of 'sharp enough' depth of field, at least on digital. Refocusing with live view on the sign at f/4 shifts the wavy plane of focus significantly forward.

 

[rscheffler]

Here the attached images are of a row of houses across the park field. Focus was pretty much at infinity according to the rangefinder patch. The first image was rangefinder focused on the house in the center and shot at f/4. The second image was live view focused at f/4 on the same house.

In the rangefinder focused image, the darker areas in the field and on the houses in the outer third of the image are well focused while the houses in the center are out of focus.

In the live view focused image, the central third of the image is sharp as well as much of the field in that area. The houses in the outer third are now out of focus.

I found in later tests that I had to stop down to f/11 to get even focus across the frame. In such an image, with this Find Edges effect applied, the alternating light and dark areas due to field curvature become an even gray field.

 

[raid]

I find your use of Find.Edges very useful. This is new to me.

 

[rscheffler]

Here is an example where this field curvature can be useful.

In this image of a bench in the park, I rangefinder focused on the near side of the bench and recomposed. At f/2 the near side is in focus and the rest of the bench is soft due to the shallow depth of field. At f/5.6 the wavy field curvature and focus shift pushes focus in the center of the image back and the plane of focus fully covers the bench.

Another scenario where this could be beneficial is something like a scene looking down a street or alley flanked by buildings, trees, etc., on both sides. Focus could be set on something nearer to the camera and placed in the rule of thirds zone, while the field curvature pushing away from the camera in the image center would extend the plane of focus down the street/alley.

 

[rscheffler]

I find your use of Find.Edges very useful. This is new to me.

Thanks. I can't take credit for it though. This technique is recommended by Roger Cicala of Lens Rentals for checking focus plane tilt and field curvature characteristics.

But I agree it is very useful at learning a given lens's characteristics and then using that information to figure out how to use it optimally for your typical use cases.

I also don't want to come across as bashing the replica. I think they've done an amazing job, both in the physical feel and use of the lens, but also 'cloning' the optics.

The prototype samples were very difficult to tell apart from the original Cron samples and this is what sold me on the project. I asked Kevin if anything has recently changed with the optics of the replica and he said he was told the optics are unchanged and continue to use flint glass, but refinements have been made elsewhere. For example my copy has the equally spaced and half-stop aperture clicks that wasn't available in the earlier production run. That said, production 'refinements' could also mean cost optimization that might affect aspects of optical performance. We may never know...

 

[D&A]

Someone a page or so back who also just received their replica lens from Wen from SH Cameras, was told the lens samples being released with equal spacing of f-stops on the aperture ring, also had the optics tweaked for a bit less contrast and sharpness and better shadow detail. Whether this is accurate or not, I can't say. One thing that's certain in general regarding optics...when you change one parameter, it will affect others (unless they too are corrected for and addressed).

Aside from all this, there were noted differences between the early samples of the replica and Lecia 8 element..such as the replica has a much more gradual falloff of the depth of field. This results in the OOF areas being more defined and less diffuse that the original 8 element. Again just an observation for what is an otherwise fine lens.

rscheffler, did you examine the two pics I posted above at f2 and f4 with the replica? I think it comes down to either there might have been a change in optics or simply the one sample you have. If someone with a recent sample can perform a simple series of test shots, I think you'd have your answer.

Dave (D&A)

 

[raid]

It would be useful to know which version replica I actually own. Are we over-analyzing the lens maybe? I am starting to get concerned about QC and design integrity. The price was right, but why are some people paying $1000 for the replica?

 

[valdas]

In the three attached images, even at the max 500 pixel size allowed by the forum, I think you can see how field curvature develops with my copy of the replica.

This is just a can placed on a gravel covered area and focus was on the can. The Find Edges filter was applied in Photoshop to essentially produce a 'focus peaking' effect to show the area of highest contrast, and therefore most likely plane of focus.

At f/2, rangefinder focus is accurate (enough) on the can.
At f/2.8 the wave becomes more apparent
At f/4 the wave is very evident and focus at the center of the image has shifted behind the can enough that it is out of focus when the full image is viewed on a 27" non-retina iMac.

I will reiterate that the MTF and field curvature documentation included with the replica indicates some waviness of field curvature is inherent in the design.

That said, and as previously stated, the severity of field curvature I'm seeing in images from my copy was not evident in images posted earlier in this thread from prototypes, nor do I see the effect in images from Ashwin Rao's copy.

Not included here, but also tested was my Canon 35/2 LTM. Yes, it too has similar field curvature. The big difference though, is that as the lens is stopped down, the majority of the curvature remains in the depth of field and the center never becomes obviously soft due to focus shift.

I'm still collecting suitable sample images to present at full resolution, for which I will provide a link in the next day or so.

Very interesting. I analyzed my shots (on film), also tried to apply this filter and I cannot see such an extreme waviness (if any). Probably my shots are not "representative" in terms of focus and aperture, will need to test more under conditions described by you.

 

[D&A]

Raid... rest easy. Your lens I would guess was produced around the same time frame as mine and is a superb performer. So far few if any issues have arisen with the lenses so far released and even among major manufactures a few samples leave the factory sub optimal. It's only when comparing the replica to the original 8 element that some differences are noted. This is bound to occur as time goes by. Price increases are also to be expected as word got out how this lens performed and was a close facsimile to the very pricy original 8 element. For the price, I couldn't ask for anything more and although I may show differences from the original or mention an anomaly or two I might come across... it doesn't diminish my enthusiasm for the replica.

Dave (D&A)

 

[rscheffler]

Someone a page or so back who also just received their replica lens from Wen from SH Cameras, was told the lens samples being released with equal spacing of f-stops on the aperture ring, also had the optics tweaked for a bit less contrast and sharpness and better shadow detail. Whether this is accurate or not, I can't say. One thing that's certain in general regarding optics...when you change one parameter, it will affect others (unless they too are corrected for and addressed).

Aside from all this, there were noted differences between the early samples of the replica and Lecia 8 element..such as the replica has a much more gradual falloff of the depth of field. This results in the OOF areas being more defined and less diffuse that the original 8 element. Again just an observation for what is an otherwise fine lens.

rscheffler, did you examine the two pics I posted above at f2 and f4 with the replica? I think it comes down to either there might have been a change in optics or simply the one sample you have. If someone with a recent sample can perform a simple series of test shots, I think you'd have your answer.

Dave (D&A)

Thanks for the additional info. Yes, I did look at those two photos. In the link I will post soon, you will find similar photos where at f/4 the tree, or other central object, is sufficiently out of focus to be noticeable (on digital). I didn't notice the same effect in your f/4 image.

I would be fine with the wavy field curvature IF depth of field covered the focus shift at f/5.6, but my copy doesn't on a 24MP digital camera. I guess based on the old circle of confusion calculations for an 8x analog enlargement, it probably does.

 

[rscheffler]

It would be useful to know which version replica I actually own. Are we over-analyzing the lens maybe?

Yes, to a degree we are overanalyzing. For me it's a matter of understanding what is happening in my images and finding the limits of the lens's performance that I consider acceptable for my expectations. IMO, this is pretty much always the case with "test" photos that prioritize technical aspects and results.

So, someone using it primarily wide open would be happy with my copy, as I am, too.

However, I would also like to be able to rangefinder focus it at f/4 or f/5.6 and get a sharp result in the center of the image without having to significantly front focus.

As it stands now, I need to learn the lens more and if I want good across-frame focus, will have to stop it down to f/11. Or, I will have to test further and determine how much I need to front focus at f/8 to pull the point of focus in the center of the image back into an acceptable plane of focus covered by depth of field while likely sacrificing some absolute mid zone sharpness.

Through feedback from forum members with this lens, I would also like to determine, if possible, whether my copy is typical or atypical. If it's atypical and could potentially be modified to achieve results closer to my initial expectations, then I would like to pursue that option. This is where Kevin and Mr. Zhou will come into play.

 

[rscheffler]

OK, here is a link to a Google Drive folder with a bunch of images.

Images are watermarked to indicate whether the lens was rangefinder focused or live view focused and the taking aperture.

For rangefinder focus, it means focus was set once and not changed as the lens was stopped down or the scene recomposed.

For live view focus, each image was refocused at the taking aperture to optimize focus at the center of the frame.

The first 9 images have Find Edges applied in Photoshop to give the 'focus peaking' effect to highlight the plane of best sharpness and illustrate field curvature characteristics as the lens is stopped down. This sequence is in half stops from wide open to f/8. The object in the center of the image is a small can I used to more easily focus the rangefinder and to give a reference point for focus shift as the lens is stopped down.

The next set is the sign in the park that was posted earlier, but including both the Find Edges version and normal version. In this set, note how focus extends behind the sign in the center of the image and the text on the sign loses critical sharpness. In the live view focused set the text remains sharp but the plane of focus in the outer 2/3 of the image is pulled significantly forward.

The following set is the row of houses opposite a park field, also posted earlier. Again, both the Find Edges version and normal version. Note how sharpness changes along the row of houses depending on whether the lens was rangefinder or live view focused.

The next set of the large white house/mansion is basically a repeat of the row of houses test, but includes the full aperture range. You'll note that f/11 is the point where across-frame sharpness is best.

Then a set of a tree flanked by two benches. This set was only rangefinder focused. You'll note that as the lens is stopped down, the tree goes out of focus and the small white cabin and area around it far in the background becomes sharp. Meanwhile the benches stay in focus. This IMO is a prime example of the "W" field curvature/focus shift exhibited by my copy.

The next set of a small building placed in the left third of the image: it was RF focused on the front of the building and the scene recomposed. As the lens is stopped down, you'll note the building remains consistently sharp, though the trees in the center of the image lose sharpness until recovered by depth of field. In this case, with the primary point of interest off center, the loss of central sharpness is less disturbing to my eyes.

Kitchen Garden sign set is similar to the earlier sign set, though includes f/11 and f/16.

The wooden fence set is similar to the earlier tree flanked by benches. Note how the fence goes out of focus in the center as the focus shifts to the background. And when live view focused, how the outer thirds of the image become very soft because placement of best focus in the center has pulled the peripheral field curvature forward of the fence.

Finally a few of a tolerant spouse to add a human element to the tests. The first two at a farther distance were both at f/4. The first was rangefinder focused and the second was live view focused.

The head & shoulder distance set includes rangefinder focused from f/2-5.6 and live view focused at f/4 and f/5.6 to highlight the sharpness difference at the two aperture settings I find most affected by the combined field curvature and focus shift of my copy. This would also be a more typical use case for me where normally I'd be shooting hand-held via the rangefinder (though these were locked down on a tripod for the purpose of the test). Were I desiring the reduced background blur of f/4 or f/5.6, it would be very difficult for me to nail sharp 'on the fly' results via the rangefinder with a centrally placed subject.

I recommend downloading the images and browsing them with your preferred image viewing/editing app rather than viewing in Google Drive to more easily flip between and compare images.

I also realize there are a lot of images and for those happy with their results with the lens, likely won't be of much interest. But I hope these at least illustrate what I've experienced with my copy and I would like to know if others see similar behavior in their images with this lens.

Thanks!

Ron

 

[Erik van Straten]

The Canon 35/1.5 has been called "a dog" for many years. It depends how you use this lens. If your main focus point is in the center, the image comes out nicely. Do you agree, Erik?

Raid, I've been always fascinated by this lens, but I've never bought or used one. My opinion is based on the pictures taken with this lens I've seen on the internet.


Erik.

 

[raid]

I have used this lens once in a while:

 

[D&A]

Yes, to a degree we are overanalyzing. For me it's a matter of understanding what is happening in my images and finding the limits of the lens's performance that I consider acceptable for my expectations. IMO, this is pretty much always the case with "test" photos that prioritize technical aspects and results.

So, someone using it primarily wide open would be happy with my copy, as I am, too.

However, I would also like to be able to rangefinder focus it at f/4 or f/5.6 and get a sharp result in the center of the image without having to significantly front focus.

As it stands now, I need to learn the lens more and if I want good across-frame focus, will have to stop it down to f/11. Or, I will have to test further and determine how much I need to front focus at f/8 to pull the point of focus in the center of the image back into an acceptable plane of focus covered by depth of field while likely sacrificing some absolute mid zone sharpness.

Through feedback from forum members with this lens, I would also like to determine, if possible, whether my copy is typical or atypical. If it's atypical and could potentially be modified to achieve results closer to my initial expectations, then I would like to pursue that option. This is where Kevin and Mr. Zhou will come into play.

Much agree. Whenever I purchase a lens, regardless of format or manufacturer, I put it through many controlled tests (all major f-stops and at least three different distances)...so I can fully understand the performance of a lens. This also lets me identify strengths and weaknesses of a lens so in actual use I will know how and when to use it.

This is just my guess about your lens. Its certainly different than those received months ago. Whether changes were made in the lens recently, its hard to say, but if they haven't...then I would venture a guess your lens may be atypical. Aside from field curvature, your lens seems to display considerable focus shift. Even at almost min distance, mine doesn't or minimal at best.

When the Zeiss 50mm f1.5 Sonnar lens in M mount was introduced, it was adjusted for best focus wide open. There would be massive focus shift at f2, f2.8 and f4 till the growing depth of field caught up. I purchased a late copy where Zeiss told me they adjusted for best focus at f2.2...which brought focus shift to a minimum. At f2.2 there was a little front focusing at f1.5 and at f2.8 and f3.5 a little back focus..but that's it. You lens focus may for whatever reason was set at a suboptimal f-stop and that's why you see your focus shift. This is not related to the field curvature though.

No lens should only be useful wide open or f11 and don't believe the replica lenses as a whole were designed this way.

I should also note, my test shots all were taken on a Leica M9...so no live view. I simply focused via Rangefinder and all my M9's are calibrated perfectly (and so was the replica lens. Sometimes, even Leica lenses come from the factory out of adjustment. Your lens may be nothing more than that. I hope you get it sorted out soon.

Dave (D&A)

 

[Erik van Straten]

Lovely shots, Raid! I like this lens very much! Like impressionist paintings!


Erik.

 

[rscheffler]

Much agree. Whenever I purchase a lens, regardless of format or manufacturer, I put it through many controlled tests (all major f-stops and at least three different distances)...so I can fully understand the performance of a lens. This also lets me identify strengths and weaknesses of a lens so in actual use I will know how and when to use it.

This is just my guess about your lens. Its certainly different than those received months ago. Whether changes were made in the lens recently, its hard to say, but if they haven't...then I would venture a guess your lens may be atypical. Aside from field curvature, your lens seems to display considerable focus shift. Even at almost min distance, mine doesn't or minimal at best.

When the Zeiss 50mm f1.5 Sonnar lens in M mount was introduced, it was adjusted for best focus wide open. There would be massive focus shift at f2, f2.8 and f4 till the growing depth of field caught up. I purchased a late copy where Zeiss told me they adjusted for best focus at f2.2...which brought focus shift to a minimum. At f2.2 there was a little front focusing at f1.5 and at f2.8 and f3.5 a little back focus..but that's it. You lens focus may for whatever reason was set at a suboptimal f-stop and that's why you see your focus shift. This is not related to the field curvature though.

No lens should only be useful wide open or f11 and don't believe the replica lenses as a whole were designed this way.

I should also note, my test shots all were taken on a Leica M9...so no live view. I simply focused via Rangefinder and all my M9's are calibrated perfectly (and so was the replica lens. Sometimes, even Leica lenses come from the factory out of adjustment. Your lens may be nothing more than that. I hope you get it sorted out soon.

Dave (D&A)

Thanks Dave, all really good points.

it would seem my lens is calibrated for RF accuracy at f/2, though I'm also not sure how perfectly my M240's RF is calibrated. In some of the test images, it might be slightly front-focused wide open, but still results in very nice images due to the leeway provided by spherical aberration in the plane of focus.

If it were f/2.8 optimized, then the problems I see at f/4 and f/5.6 would be reduced, though possibly at the expense of some mid zone performance where it is currently quite strong. It would probably make f/8 more usable than it is now; I might simply have to resort to some intentional front-focus when stopped down to try to find a compromise between the center and edges. If the central 'spike' of combined curvature and focus shift was less severe, field curvature would probably fall within depth of field at most apertures. Maybe this is can be tuned by the factory?

 

[D&A]

Thanks Dave, all really good points.

it would seem my lens is calibrated for RF accuracy at f/2, though I'm also not sure how perfectly my M240's RF is calibrated. In some of the test images, it might be slightly front-focused wide open, but still results in very nice images due to the leeway provided by spherical aberration in the plane of focus.

If it were f/2.8 optimized, then the problems I see at f/4 and f/5.6 would be reduced, though possibly at the expense of some mid zone performance where it is currently quite strong. It would probably make f/8 more usable than it is now; I might simply have to resort to some intentional front-focus when stopped down to try to find a compromise between the center and edges. If the central 'spike' of combined curvature and focus shift was less severe, field curvature would probably fall within depth of field at most apertures. Maybe this is can be tuned by the factory?

Only time for a short response now but if you are in the States, DAG can easily adjust for optimal focus at a given f-stop that will minimize focus shift. Still I believe you shouldn't have to go through this and if you are able to switch out for another lens, that would solve your problem...otherwise DAG. The replica I have is adjusted so I don't even have to think about what f-stop I'm shooting at and that's the way your lens should be.

Dave (D&A)

 

[Doug]

Only time for a short response now but if you are in the States, DAG can easily adjust for optimal focus at a given f-stop that will minimize focus shift. Still I believe you shouldn't have to go through this and if you are able to switch out for another lens, that would solve your problem...otherwise DAG. The replica I have is adjusted so I don't even have to think about what f-stop I'm shooting at and that's the way your lens should be.

Dave (D&A)

Agreed... referring to the C-Sonnar, mine when new was set to best focus at ~f/2.8 and DAG adjusted it to ~f/2.2, at which point it's good for general use without having to think about focus shift. I don't know that this kind of adjustment would resolve the odd behavior of the specific LLL replica.

 

[yossi]

Thanks for the high-rez sample image.

The MTF documentation provided with the replica indicates that there is some degree of "W" shaped field curvature inherent in its optical performance. I believe this is what is seen in your image and likely is not quite as severe as that exhibited by my copy.

My question to you is: if you were to put one of those trees in the center of the image, or a person, or any object, and photograph it at f/4 or f/5.6, is it acceptably in-focus?

With my copy it is not.

...

Sorry for the late response (busy at work and due to very wet weather these few days, no chance to go out to test shoot).

Tonight I took two test shots with the 8E Clone, M9M on tripod. One at f2 and the other at f4. Close range. Target at centre of frame. Result: both shots focus spot on. f2 shot only centre portion is sharp, further away starts to get softer but the bottom right corner is sharpen up. At f4, almost the whole frame is almost evenly sharp, but that the bottom right corner spot is unusually sharper than its neighbour. (Btw these test shots are too boring to be shown here. Unless there are some technical reasons to further examine them, I shall spare everyone from them LOL).

In short, my copy has that weird sharp corner but otherwise no focus shift issue like yours.