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

[jonmanjiro]

The original design used a combination of different kinds of crown- and flint glass. Hard to imagine this can be imitated without any flint glass.

Nikon did a great job with the reissue Nikkor-S 50mm f1.4 and W-Nikkor 3.5cm f1.8. The original W-Nikkor has glass elements that contain lanthanum, but there's no way Nikon used lanthanum in the W-Nikkor reissue. Long story short, I think it would be quite easy to replicate the Summicron V1 using modern non-leaded glass. Going to the hassle of using flint glass containing lead oxide is just an extra level of geekery to make a lens as close as possible to the original. Not necessary at all, but if you can, why not?!

 

[yossi]

Nikon did a great job with the reissue Nikkor-S 50mm f1.4 and W-Nikkor 3.5cm f1.8. The original W-Nikkor has glass elements that contain lanthanum, but there's no way Nikon used lanthanum in the W-Nikkor reissue. Long story short, I think it would be quite easy to replicate the Summicron V1 using modern non-leaded glass. Going to the hassle of using flint glass containing lead oxide is just an extra level of geekery to make a lens as close as possible to the original. Not necessary at all, but if you can, why not?!

If memory if serves me right, you tested both original W-Nikkor 35/1.8 and the reissued version and the original one is better. If so, it seems that not even Nikon could replicate its own lens completely with different glass type.

 

[jonmanjiro]

If memory if serves me right, you tested both original W-Nikkor 35/1.8 and the reissued version and the original one is better. If so, it seems that not even Nikon could replicate its own lens completely with different glass type.

I did indeed do that test (here). Not sure why you thought the original was much better though. IMO the rendering of the two lenses is very very close (much closer than the Summicron V1 and LLL 35mm, which are quite similar anyway) and "better" comes down to personal preference. The difference in colour rendering is the only easily discernible difference - the original renders colours with a warmer tone. That's caused by the lanthanum in the glass and can be rectified easily (link), after which the colour rendering of the two lenses is pretty much identical.

 

[rscheffler]

@rscheffler,

Just to confirm once and for all if there is focus shift, I had gone out and found a tree, put it dead centre of the frame and took 4 shots at f2, f2.8 f4 and f5.6.

Shot with M9 Monochrom on tripod. jpeg untouched. last 2 digits of the filename indicates the f-stop. You can view them HERE. (click on "view original" at the left top of the page to see image in full-size).

Thanks!!

I downloaded and had a look at them. My conclusion differs a bit from Dave's.

I see focus shift in your images, but the big difference with your copy is that the center point of focus mostly remains within depth of field for acceptable sharpness.

I think your lens is focus optimized for f/2.8, depth of field masks most of the shift in the center and there is so much SA wide open that it masks the slight front focus (notice how the grass in front of the bus stop sign is in focus wide open, but not the grass directly under the sign - at f/2.8 it's sharp).

In the tree sequence, the bark on the near side of the tree is softer in the very center of the frame at f/5.6 than at f/4, but is sharp on the sides of the tree, and also at the base and top of the tree. This is what I see in my copy - the very center goes softer.

Notice the field curvature in the bus stop sequence: the grass behind the sign get sharper and sharper, until it reaches the far background at f/5.6 in the center of the frame, but not the mid zone areas on the sides, while the railing remains in focus along the sides (the field curves from the background at the center of the frame towards the camera and along the intended plane of focus along the railing).

This is exactly what is shown in my 'tree with benches' sequence. But with your copy, the point of focus in the center remains within depth of field at f/4 (mine does not) and is acceptable, if not critically sharp at f/5.6. This is more noticeable in the tree bark than the text in the sign.

I would guess with your copy, by f/8 depth of field has recovered sharpness in the very center of the image and slightly front focusing at f/5.6 would probably be enough to bring the center back to better sharpness.

Thanks again for taking the time to test your copy!

Ron

I checked your link and found out that shot at F4 is sharper than shot at F5.6. Does it mean that there's a focus shift here? In fact, At F5.6 the lens is back focus.

I think so, based on my comments above.

I received my lens two weeks ago. BP version with uneven f-stop spacings. #500138.

Yossi

Mine is BP with even spacing, number 500151.

 

[D&A]

rscheffler, I do concur with your observations in your post above. There a bit of focus shift in Yossi's images but was trying to emphasize the point that it was minimal (as far as I could determine compared to your lens). We also must keep in mind, these are manual focus lenses and depending which way each of us turn the focus barrel and then stop where our eye sees focus is a bit different for each of us. In addition no Leica digital body, no matter how well the rangefinder is calibrated, may indicate proper focus at a slight different point....so when I see a bit of front or back focus using a rangefinder, compared to someone else's rangefinder, slight differences may or may not be attributed to the lens vs. the body. In your case, the obvious differences are quite noticeable and thus can be attributed to your lens.

Now if everyone had live view to test out their lenses, that would be a different story. That's one reason I didn't call out (mention) the slight back focus seen with Yossi's lens (covered by the depth of field)...simply due to the variances I mentioned.

The other possible interesting fact is lens with #500138 has uneven f-stop spacing and yours with #500151 has even spacing. Too soon to make a pronouncement but maybe even spacing was introduced sometime between these two numbers?

Dave (D&A)

 

[james.liam]

Looks like there's an official European vendor. €1,500 incl. 20% VAT

 

[raid]

We did well at a price of $500.

 

[james.liam]

We did well at a price of $500.

You should have bought two.

But at 1500€ the soon-to-be-announced 35 Summaron f/2.8 becomes more appealing.

 

[raid]

You should have bought two.

But at 1500€ the soon-to-be-announced 35 Summaron f/2.8 becomes more appealing.

I nearly did so, but why would I need two replicas? Make money? This was not my goal. I am waiting for the CV 50/2 APO.

 

[tamibss]

Above all, i personally appreciate efforts of Kevin and Zhou team etc.

Having said that many of members are still waiting over a year having promised Batch 1 etc,

I also understand there could be business element involved too, still how could ebay seller get 20 copies of batch while our understanding was limited run and long waiting list for many of us here 🙂

 

[rscheffler]

Without doubt, IMO, Mr. Zhou is channeling some production to the open market at a higher price. Given the demand for the lens and the entrepreneurial nature of most people in business, this would seem to be a logical decision. Especially considering how this has morphed from a small pet project to one with apparently strong worldwide demand.

At least those on the waitlist are still receiving their lenses at the originally promised price. The maker could have easily claimed increasing materials/labor costs, complications due to the pandemic, etc., and raised the price. But they didn't, which is commendable.

rscheffler, I do concur with your observations in your post above. There a bit of focus shift in Yossi's images but was trying to emphasize the point that it was minimal (as far as I could determine compared to your lens). We also must keep in mind, these are manual focus lenses and depending which way each of us turn the focus barrel and then stop where our eye sees focus is a bit different for each of us. In addition no Leica digital body, no matter how well the rangefinder is calibrated, may indicate proper focus at a slight different point....so when I see a bit of front or back focus using a rangefinder, compared to someone else's rangefinder, slight differences may or may not be attributed to the lens vs. the body. In your case, the obvious differences are quite noticeable and thus can be attributed to your lens.

Now if everyone had live view to test out their lenses, that would be a different story. That's one reason I didn't call out (mention) the slight back focus seen with Yossi's lens (covered by the depth of field)...simply due to the variances I mentioned.

The other possible interesting fact is lens with #500138 has uneven f-stop spacing and yours with #500151 has even spacing. Too soon to make a pronouncement but maybe even spacing was introduced sometime between these two numbers?

Dave (D&A)

Yes, certainly there is a lot of variability combined with the nature of the rangefinder system.

I feel the inherent design of the lens, with a fair amount of field curvature, means it's a balance between placement of ideal focus for the center of the image versus maintaining an overall good level of sharpness across the desired plane of focus. As a result there is some central focus shift, but am guessing that a properly calibrated copy would normally keep the shift within depth of field.

Based on my copy with live view, keeping the center of the image in focus as the lens is stopped down has a dramatic effect on the peripheral plane of focus. It shifts considerably closer to the camera in the outer areas relative to the center. Therefore the results would be considerably different than for example those seen in yossi's images and likely would be worse for certain types of scenes.

But the way the lens is set up now, for rangefinder focusing, it's a compromise balance between maintaining central and peripheral sharpness. Whether or not someone will see focus shift in their copy will likely depend on image content, the characteristics of their copy, whether shooting film or digital, their expectations, etc. In other words, lots of variables.

 

[james.liam]

Above all, i personally appreciate efforts of Kevin and Zhou team etc.

Having said that many of members are still waiting over a year having promised Batch 1 etc,

I also understand there could be business element involved too, still how could ebay seller get 20 copies of batch while our understanding was limited run and long waiting list for many of us here 🙂

What’s to understand? This thread is 100 pages long dating back 18 months. A well thought-out and effective marketing strategy. Not to mention, virtually free. It’s about the profit (a good thing), not the kindness of Mr Zhou’s heart. Or his investors. Those first few hundred delivered at cost ($500) and demos sent to LHSA are worth 1000-fold in market interest generated.

Unlike Mitakon, 7TT-fArtisans and the rest, LLL worked to establish a reputation with its first model, as that of a higher-quality, high-cost producer of a niche item. Genius.

 

[rscheffler]

Did some casual shooting with the replica today during a 'lockdown fitness walk' and also tried the 35 Cron lens codes available in the M240. Lens was focused at infinity to provoke strongest color shift and vignetting. White balance was set for the 35 Cron non-ASPH image and applied to all three for consistent DNG export in Lightroom. Any color differences you see are therefore due to the embedded lens code corrections at time of capture rather than due to raw conversion variables.

See watermarks in the images below for lens code details.

My observations: uncoded is strongest vignetting with blue shift in corners. 35 Cron non-ASPH code has most neutral corner colors. 35 Cron ASPH has greenish corners.

 

[rscheffler]

Here are some of the walk-around photos from today. Mostly the part where we walked through a cemetery.

The full set can be seen at this link: https://postimg.cc/gallery/pjmK6X4

A selection is below to give some immediate impressions. Images were mostly at f/2 but a few at f/2.5 might be mixed in. I no longer remember exact details for each. If you go to the gallery, file numbers up to 0028 were uncoded. Everything thereafter was coded as the 35 Cron non-ASPH.

I might compare it more, but I think I’m leaning towards f/2.5 for general purposes in that contrast picks up a bit and the central hotspot is reduced, yet some of the wide open character is retained. For people photos, with the face located centrally, f/2 might be the most flattering, especially for women. The lower contrast and touch of veiling combined with maximum SA opens shadows and masks skin imperfections (sorry, no examples of this included).

Images have been processed to taste, but I have generally withheld use of clarity. Contrast has been tweaked via curves adjustment. Also generally no cropping other than to level an image, or perspective correction (the plastic snowman, book shop window). No distortion correction has been applied.

















Notice the coma in the upper right corner Christmas lights. 🙂

 

[rscheffler]

In one of the above images, the out of focus lights of the building in the background have a hot-spot in the middle of the out of focus ‘bokeh’ ball. Wondering if this is also typical of the original 35 Cron? Here's a crop:

 

[TenEleven]

I gotta say that's less coma than I had with my original M/L 8-Element Summicron. The lens had been professionally cleaned and overhauled at Kanto Camera in Tokyo. However, when shooting wide open, the corners still were loaded with coma - especially at night.

I was told that this is just how it is.

 

[rscheffler]

Good to know. 🙂

I think we're learning that this replica is pretty faithful, but not an exact copy. Which IMO is fine.

 

[yossi]

As for me, through the Operation “Witch Hunt”, I have a much better understanding how this replica behaves. My feeling towards its performance is a mixed bag. While it retains most of the original 8E characteristics, it has some quirky behaviour, viz. abnormally sharp corners for distant objects while shooting at close range (super annoying), wavy curvature of field, hot spot within bokeh balls, etc. Knowing these and avoiding them where possible, it is still a fine lens to shoot. Exact 8E Summicron replica it is NOT, IMHO.

 

[D&A]

As for me, through the Operation “Witch Hunt”, I have a much better understanding how this replica behaves. My feeling towards its performance is a mixed bag. While it retains most of the original 8E characteristics, it has some quirky behaviour, viz. abnormally sharp corners for distant objects while shooting at close range (super annoying), wavy curvature of field, hot spot within bokeh balls, etc. Knowing these and avoiding them where possible, it is still a fine lens to shoot. Exact 8E Summicron replica it is NOT, IMHO.

To respond to both rscheffler's and Yossi's most recent comments.....yes, I too believe the replica lens does retain most of the 8E characteristics with some of the differences anomalies shown by rscheffler, Yossi, myself and others. For myself, those abnormally sharp corners at infinity of objects clearly out of the depth of field sometimes is quite noticeable and is clearly different than the 8E I shot with previously. I didn't seee evindence of thois in Yossi's test image. My copy (and from what I've seen of Yossi's and many others posted test images) displays little focus shift (except for what rscheffler has shown in his copy), so that generally hasn't been an overall issue. I haven't encountered the hot spots in the bokeh balls but then again I haven't shot anything that would have provoked it yet. As for wavy field curvature, that too seems from what I understand the providence of rscheffler's sample unless we are attributing the abnormally sharp "out of the depth" of field corners as an indicator of this. Also depth of field drop off is far more gradual than the 8E, lending OOF areas to appear sharper with more definition. That can be a plus or minus, depending on subject matter and individual preference. On at least the Leica M9, there is very noticeable cyan cast to images, especially on overcast days.

Obviously there seems to be differences in individual copies of the replica and it might come down to whether a individual notices these anomalies. shoots film or digital, the calibration of their camera and most of all..something we haven't discussed in length, but the consistency that each lens during construction and subsequent bench testing, is calibrated and adjusted within a set of fixed tolerances.

It would be unfair of me not to also say there are improvements in the replica compared to the 8E. Contrast is better, resolution seems to be higher as does definition of certain objects...and of course having new glass and mechanics. The replica as stated may not be a 100 percent faithful replica of the original 8E, but even if the lens becomes somewhat more expensive than its initial price point, it an exceptional lens in my opinion, when every aspect of the replica from construction, to performance is taken into account. Tweaks to the replica may eventually bring it even closer to the 8E. Even now, the lens has lots of character, a close representation to the original 8E and notable performance overall,,,and that's aside from the faithful construction to the original 8E.

The maker/manufacture should be proud of what he's achieved so far as well like Kevin and others who have been involved in getting this lens off the ground at such a reasonable cost and into our hands.

Dave (D&A)

 

[james.liam]

Wondering if anyone can comment on the original vs new 28 Summaron and whether differences/improvements appeared there.