That's the test I referred to when I wrote about decentering (4-corners) testing on my previous comment. I already tested with this method (at F2) rather than 1-stop down, but I will run the test a bit more carefully before sharing any corner samples from it. I always use variations of this test when checking my lenses right after getting them (at least if the lenses are not extreme character lenses).
Here are the extreme corner 500x500 pixel crops from today's run of decentering test on my 2 copies of 28/2 Apo for E-mount. I ran the test horizontally (handheld with Electronic Shutter) this time with following pattern:
IBIS enabled f2: left top, right top, right bottom (camera flipped upside down), left bottom
IBIS enabled f2.8: same sequence
IBIS off f2: same sequence
IBIS off f2.8: same sequence
I don't think all corners are 100% equal but the results are not indicating any noticeable decentering. There's also always a little bit of variation in the results between each run. New copy might have slightly crisper corners in this test overall.
Thanks Juha.
The following picture shows the crops of top left & top right corners @ f/2 for both the samples (original and new). The top left corners (first raw in the attached picture) had a different exposure (1 stop darker) so I increased their brightness acting on levels. Assuming no focusing errors, the original sample has (slightly) less crisp corners, as Juha already noted. I wonder if this might have some consequence on purple fringing. To answer this question I would try to replicate purple fringing with the new sample.
Both lenses are very good, IMO, but the new sample seems (under the assumption underlined above) even better.
Thanks Juha.
The following picture shows the crops of top left & top right corners @ f/2 for both the samples (original and new). The top left corners (first raw in the attached picture) had a different exposure (1 stop darker) so I increased their brightness acting on levels. Assuming no focusing errors, the original sample has (slightly) less crisp corners, as Juha already noted. I wonder if this might have some consequence on purple fringing. To answer this question I would try to replicate purple fringing with the new sample.
Both lenses are very good, IMO, but the new sample seems (under the assumption underlined above) even better.
Thanks! The exposure difference is due to use of auto exposure and different framing to get the same building top at the corner (much more sky in the frame for one shot and much more buildings for the other). I focused centrally at f2 in the same far distance target with both lenses before shooting each series and used magnification and took time to get it right so there should be no focus errors.
My original copy of the lens has ideal infinity focus at a point that is slightly further than midpoint of infinity symbol whereas new copy has it right at the midpoint (which I always think is a good sign of perfect calibration).
I'm not sure when I can try to reproduce similar scenario as the original shot with the new lens. Definitely can't go to same location... From next weekend I'll go on a one week business trip too so shooting opportunities will be limited in the coming weeks. Anyway, I'll keep it in mind.
I’ll probably give this lens a try, the new price for all the E mount APO Lanthars here in Korea are under $800 with the exchange rate so pretty good chance to pick one up and I've been wanting to try a 28mm prone for a long time! Been wanting to re-acquire a 50mm too. That was one of my favorites, shouldn’t have sold it.
Thank you Juha. It seems that lateral CA is also visible here (for example DSC08822 and 8932). So it belongs to this lens I suppose. It requires pp because otherwise you can have a purple glow over your pictures in the extreme situations.
ustjwenew wrote:
Thank you Juha. It seems that lateral CA is also visible here (for example DSC08822 and 8932). So it belongs to this lens I suppose. It requires pp because otherwise you can have a purple glow over your pictures in the extreme situations.
Thanks! Yes, I think it can show up in some extreme situations with this lens and it's not a lens copy issue or it might be very difficult to find a copy that behaves differently. I don't personally pay much attention to CA usually at all and I basically never try to remove it in PP. I appreciate the CV APO lenses for overall strong contrast rendering, flat field and great corner-to-corner sharpness. I also use a lot of wild character lenses for other purposes where I appreciate different aspects of rendering and I'm never concerned about CA with those.
I'm happy with both of my copies of the 28/2 AL lens but I feel like new one is maybe a little bit closer to a perfect copy overall, it seems to have a bit stronger bite in wide open shooting at close distances (as well as having optimal infinity right at the midpoint of infinity symbol). For now I'll be keeping both.
Juha Kannisto wrote:
... I think it can show up in some extreme situations with this lens and it's not a lens copy issue or it might be very difficult to find a copy that behaves differently.
Yes, the E-mount 28/2 APO does show a tendency to purple fringing on Sony sensors under extreme situations (e.g., tree branches on an overexposed sky in the corners of the frame).
It would be interesting to see if the VM 28/2 APO-Lanthar shows similar fringings on Leica M bodies, which have offset or uniquely shaped microlenses to address color shifts at the edges and/or thinner filter stack.
Ripolini wrote:
Yes, the E-mount 28/2 APO does show a tendency to purple fringing on Sony sensors under extreme situations (e.g., tree branches on an overexposed sky in the corners of the frame).
It would be interesting to see if the VM 28/2 APO-Lanthar shows similar fringings on Leica M bodies, which have offset or uniquely shaped microlenses to address color shifts at the edges and/or thinner filter stack.
Yeah... I did have the VM 28/2 for a couple of months to try it on my Sony cameras before the E-mount version got announced but I never had any Leica M bodies and I don't have the VM lens anymore. I didn't use it much since it wasn't really well suited for the Sony sensor (due to rather heavy sensor stack induced field curvature).
I think Leica Ms have quite many types of sensors (Sony 61 MP sensors on the latest generation and some other suppliers sensors before that). They advertised those microlenses around MP10 series but not sure if they gave up on them in later gen cameras and just rely on thin stack. The results might differ between different generations of digital Ms and it would be quite tricky to test comprehensively.
Juha Kannisto wrote:
I think Leica Ms have quite many types of sensors (Sony 61 MP sensors on the latest generation and some other suppliers sensors before that). They advertised those microlenses around MP10 series but not sure if they gave up on them in later gen cameras and just rely on thin stack. The results might differ between different generations of digital Ms and it would be quite tricky to test comprehensively.
Thanks Juha.
The following picture shows the crops of top left & top right corners @ f/2 for both the samples (original and new). The top left corners (first raw in the attached picture) had a different exposure (1 stop darker) so I increased their brightness acting on levels. Assuming no focusing errors, the original sample has (slightly) less crisp corners, as Juha already noted. I wonder if this might have some consequence on purple fringing. To answer this question I would try to replicate purple fringing with the new sample.
Both lenses are very good, IMO, but the new sample seems (under the assumption underlined above) even better.
Even if there are some minor differences, I find the performance of both lenses astounding.
just shows how amazing this new 28mm Apo Lanthar is.
PS:
Precise focusing is such a crititical point with this lens (due to it‘s extremely shallow plane of focus) that I fould testing my own example to be quite challenging. Just knocking the focus a bit off gives the impression, that the lens might be less than perfect, which just isn‘t true. I needed a lot of retesting to exclude the minutest errors I had made. Testing less perfect lenses is much less cumbersome, they reach a certain level and then you can stop, because there is nothing more to be gained. This lens ist just amazing in what it can achieve in the corners 😄
PS:
Precise focusing is such a crititical point with this lens (due to it‘s extremely shallow plane of focus) that I fould testing my own example to be quite challenging. Just knocking the focus a bit off gives the impression, that the lens might be less than perfect, which just isn‘t true. I needed a lot of retesting to exclude the minutest errors I had made. Testing less perfect lenses is much less cumbersome, they reach a certain level and then you can stop, because there is nothing more to be gained. This lens ist just amazing in what it can achieve in the corners 😄...Show more →
I dislike the perfectly silky-smooth focus action that Cosina seems to favor for their lenses. It feels nice in the showroom, but is a flaw in real world use. Requiring a little more effort to rotate the focus ring would be better.
Keith B. wrote:
I dislike the perfectly silky-smooth focus action that Cosina seems to favor for their lenses. It feels nice in the showroom, but is a flaw in real world use. Requiring a little more effort to rotate the focus ring would be better.
Which Voigtländer lens did you refer to? The focus ring on E-mount version of this lens requires a decent amount of torque to turn (more than average for CV lenses, perhaps due to floating element design), the focus action is not too light at all and definitely smooth/even is better than the unsmooth when it comes to focusing feel. This focusing ring will not turn accidentally with the slightest of touch. Most legacy lenses require a bit lighter torque than this.
Keith B. wrote:
I dislike the perfectly silky-smooth focus action that Cosina seems to favor for their lenses. It feels nice in the showroom, but is a flaw in real world use. Requiring a little more effort to rotate the focus ring would be better.
I have never once felt like this is a flaw, but everyone has their preferences.
"Which Voigtländer lens did you refer to? The focus ring on E-mount version of this lens requires a decent amount of torque to turn (more than average for CV lenses, perhaps due to floating element design), the focus action is not too light at all and definitely smooth/even is better than the unsmooth when it comes to focusing feel. This focusing ring will not turn accidentally with the slightest of touch. Most legacy lenses require a bit lighter torque than this"
My copy of the CZ 35/1 Apo is very smooth turning...I'd like it to have substantially more resistance. My personal taste, of course. I also have a CZ 15/4.5 III that's pretty much the same as the 35/2 Apo. Also have briefly owned the E-mount 21/1.4, 21/3.5, and Z-mount 28/1.5. If the new Z 28/2 Apo takes more effort to turn, I'll be pleased.
Keith B. wrote: "Which Voigtländer lens did you refer to? The focus ring on E-mount version of this lens requires a decent amount of torque to turn (more than average for CV lenses, perhaps due to floating element design), the focus action is not too light at all and definitely smooth/even is better than the unsmooth when it comes to focusing feel. This focusing ring will not turn accidentally with the slightest of touch. Most legacy lenses require a bit lighter torque than this"
My copy of the CZ 35/1 Apo is very smooth turning...I'd like it to have substantially more resistance. My personal taste, of course. I also have a CZ 15/4.5 III that's pretty much the same as the 35/2 Apo. Also have briefly owned the E-mount 21/1.4, 21/3.5, and Z-mount 28/1.5. If the new Z 28/2 Apo takes more effort to turn, I'll be pleased....Show more →
Thanks! I have owned and used quite many CV lenses, currently have 30+ across E-mount, X-mount and VM / LTM. I have had all 20 of their E-mount lenses, keeping 15+1 at the moment (since I have 2 copies of 28/2 APO). Among these, 28/2 APO E-mount is among those with relatively heavy resistance, and both of my copies feel exactly the same in this regard so I think it's the standard with this lens. First when I received my original copy I was a bit surprised about the level of resistance, but it's not a problem in use at all. I don't have experience with any of their Z-mount lenses so can't compare relative to those and can't know if 28/2 APO in Z-mount will be just like 28/2 E in this regard though.
Many of the older ones in E-mount are definitely lighter in terms of focus resistance, especially my 35/1.4 Nokton classic. I would say that all of the new E-mount designs they've released since 2024 have had relatively heavier resistance than some of the older ones, but for example 28/1.5 Nokton in E-mount focuses a bit more lightly than 28/2.
