It's a torture scenery with a complex background and high contrast lighting. I don't think many 35mm lenses would do any better though. I'm sure the RX1 Sonnar would have smoother transition zone but perhaps even worse CA.
It's a bit disappointing to see this after what they accomplished with the 45/2.8 but I could already imagine the complaints about AF-C accuracy and soft images up close if they went that route.
Sigma had a great opportunity to design a 35mm that excels in rendering but it looks like they preferred to play safe and make a very sharp lens which most shooters prefer anyways.
Wow, that's a tough scene!
But it has a really nice 3d ring to it . The corners look pretty scary though, I wonder if f1.2 would make them look significantly better...
j4nu wrote:
Wow, that's a tough scene!
But it has a really nice 3d ring to it . The corners look pretty scary though, I wonder if f1.2 would make them look significantly better...
The corners and transition zone. Zoom in to see it at pixel level.
This is also a characteristic of the Voigtlander 35/1.2 and 40/1.2 lenses. The transition zone is also pretty harsh but the stronger blur at f/1.2 usually saves the scene.
It's a torture scenery with a complex background and high contrast lighting. I don't think many 35mm lenses would do any better though. I'm sure the RX1 Sonnar would have smoother transition zone but perhaps even worse CA.
It's a bit disappointing to see this after what they accomplished with the 45/2.8 but I could already imagine the complaints about AF-C accuracy and soft images up close if they went that route.
Sigma had a great opportunity to design a 35mm that excels in rendering but they preferred to play safe and make a very sharp lens which most shooters prefer anyways. My hope is on the 24/3.5 now....Show more →
The axial CA in that shot is pretty awful, but I wonder how much of it is caused by the harsh lighting and the near blown highlights in the middle of the shot caused by the sun being just outside the frame in the upper middle of the shot. I would love to know how this shot would look with a slightly different framing or at least with a shorter exposure. We see as we move to left edge of the frame the axial CA is much less prominent and the exposure is less hot in that area. It makes me wonder whether a stop or a stop and half of less exposure might have meant a lot less axial CA. I wouldn't write off the lens from this one example without know how some simple changes might have affected the shot.
Steve Spencer wrote:
The axial CA in that shot is pretty awful, but I wonder how much of it is caused by the harsh lighting and the near blown highlights in the middle of the shot caused by the sun being just outside the frame in the upper middle of the shot. I would love to know how this shot would look with a slightly different framing or at least with a shorter exposure. We see as we move to left edge of the frame the axial CA is much less prominent and the exposure is less hot in that area. It makes me wonder whether a stop or a stop and half of less exposure might have meant a lot less axial CA. I wouldn't write off the lens from this one example without know how some simple changes might have affected the shot....Show more →
I have not seen many 35mm lenses that would do any better in respect to axial CA in high contrast scenes like this. I'm sure the Voigtlander 35/1.2 and Sony 35/1.8 would not do better even at f/2. The RX1's sonnar would also show nasty fringing as well. Perhaps the Sigma 35/1.2, Tamron 35/1.4 and Zeiss 35/1.4 ZA would do better.
What was revealing to me is the harsh rendering at transition zone. It takes an image shot at mid-distance + a complex OOF area to see it (like bark on the floor). The lens' rendering is fine at close distance but when when the subject is focused at 2-5m, things like this may happen. This is just an example of such scene and a case where the Sigma 45/2.8C would have a more pleasant rendering.
Steve Spencer wrote:
The axial CA in that shot is pretty awful, but I wonder how much of it is caused by the harsh lighting and the near blown highlights in the middle of the shot caused by the sun being just outside the frame in the upper middle of the shot. I would love to know how this shot would look with a slightly different framing or at least with a shorter exposure. We see as we move to left edge of the frame the axial CA is much less prominent and the exposure is less hot in that area. It makes me wonder whether a stop or a stop and half of less exposure might have meant a lot less axial CA. I wouldn't write off the lens from this one example without know how some simple changes might have affected the shot....Show more →
Do you think we may actually be looking at some sensor blooming in this image? That could make it difficult to judge how much colour fringing is due to sensor blooming and how much is due to axial CA.
Dec 03, 2020 at 07:20 PM
Steve Spencer Offline • • • • • • • Upload & Sell: On
Fred Miranda wrote:
I have not seen many 35mm lenses that would do any better in respect to axial CA in high contrast scenes like this. I'm sure the Voigtlander 35/1.2 and Sony 35/1.8 would not do better even at f/2. The RX1's sonnar would also show nasty fringing as well. Perhaps the Sigma 35/1.2, Tamron 35/1.4 and Zeiss 35/1.4 ZA would do better.
What was revealing to me is the harsh rendering at transition zone. It takes an image shot at mid-distance + a complex OOF area to see it (like bark on the floor). The lens' rendering is fine at close distance but when when the subject is focused at 2-5m, things like this may happen. This is just an example of such scene and a case where the Sigma 45/2.8C would have a more pleasant rendering....Show more →
I am not sure that the axial CA in the shot isn't leading to a lot of the transition zone busyness. You can see that in the OOF highlights that the edges are quite green and the centers are quite magenta in this scene. I think we are seeing the the different wavelengths are blurring at different distances and I think that might be contributing to the busy transition zone bokeh. I would love to see if it was exposed a stop or a stop and half less with a faster shutter speed, if the bokeh didn't look as busy as the axial CA was less prominent. I am not at all sure that the prominent CA and the mid zone busyness in the bokeh are unrelated. I think they may well be.
Dec 03, 2020 at 07:22 PM
Steve Spencer Offline • • • • • • • Upload & Sell: On
bjornthun wrote:
Do you think we may actually be looking at some sensor blooming in this image? That could make it difficult to judge how much colour fringing is due to sensor blooming and how much is due to axial CA.
I suppose that could be, but I am not sure I understand sensor blooming enough to know how to identify it.
Steve Spencer wrote:
I am not sure that the axial CA in the shot isn't leading to a lot of the transition zone busyness. You can see that in the OOF highlights that the edges are quite green and the centers are quite magenta in this scene. I think we are seeing the the different wavelengths are blurring at different distances and I think that might be contributing to the busy transition zone bokeh. I would love to see if it was exposed a stop or a stop and half less with a faster shutter speed, if the bokeh didn't look as busy as the axial CA was less prominent. I am not at all sure that the prominent CA and the mid zone busyness in the bokeh are unrelated. I think they may well be....Show more →
I just looked quite closely, but even in the areas that aren't fringing, seems like some extremely minor doubling or general nervousness, which make me think it's the structure of the transition zone.
Honestly, I don't think the ZA 35 1.4 would do better here for LoCA, or most other well performing 35mm's. I would be quite surprised if the RX1, Pentax 31 or CV 35/1.7 or 35/1.2 did better. The Samy 35 1.8 and FE 35 would be way worse.
Steve Spencer wrote:
I suppose that could be, but I am not sure I understand sensor blooming enough to know how to identify it.
I don’t think it’s sensor blooming, the colour fringes are magenta ang green on different sides of the focal plane, which suggests axial CA (but no worse than on any other small 2/35)
Fred Miranda wrote:
I have not seen many 35mm lenses that would do any better in respect to axial CA in high contrast scenes like this. I'm sure the Voigtlander 35/1.2 and Sony 35/1.8 would not do better even at f/2. The RX1's sonnar would also show nasty fringing as well. Perhaps the Sigma 35/1.2, Tamron 35/1.4 and Zeiss 35/1.4 ZA would do better.
Thanks, Fred. Such comparisons matter as I try to avoid a 'perfect is the enemy of the good' scenario. I appreciate your direct comparisons that aid the inevitably personal decisions regarding which compromises are acceptable. None of my lenses are perfect, but I've been able to choose some that I like thanks in part to your work and to the information others have posted here.
Fred Miranda wrote:
The corners and transition zone. Zoom in to see it at pixel level.
This is also a characteristic of the Voigtlander 35/1.2 and 40/1.2 lenses. The transition zone is also pretty harsh but the stronger blur at f/1.2 usually saves the scene.
Here is a crop showing the area:
Nice finding, thanks for sharing. That is indeed bad looking.
This is a good example that you can fail almost any lenses with right combination, lighting, distance, subject to transition ratio, transition area detail frequency
This is definitely not a good looking bokeh. Stop down to f2.8 may help it. I'd love to see your future review comparison to 45mm at f12.8 to reveal that. Treat this 35mm's f2 aperture as a bonus compare to 45mm. Having said that, expect 35mm perform as good as 45 is also a little unfair. I see so many decent rendering 50mm glasses but so few for 35mm so far.
You are very right about VC 35 and 40 f1.2 in this case. Actually, 3-5M, harsh lighting, close background (with lots of detail) to subject ratio is a recipe fail most "fast" glasses, especially for wide to standard focal. Longer lens will tame it better, shorter lens deeper DOF will cover it, though for shorter lens, problem zone will shift to closer range.
I personal think most lenses will fail in this case include RX1 actually Do you have similar images like these from RX1 before? 58G maybe can survive, 45mm another maybe but 75lux and 80lux will fail even not as bad due to longer focal, the problem zone usually shift to different distance.
Don't get me wrong, I don't think this 35mm have super smooth bokeh but even neutral bokeh is a luxury for most 35mm I have seen so far. Don't expect it can perform under all cases. As for CA under this condition, this is not bad TBH. (and I know RX1 have worse CA performance than this for sure.)
It is all about right balance of individual performance spec including price and size.
zhangyue wrote:
Nice finding, thanks for sharing. That is indeed bad looking.
This is a good example that you can fail almost any lenses with right combination, lighting, distance, subject to transition ratio, transition area detail frequency
This is definitely not a good looking bokeh. Stop down to f2.8 may help it. I'd love to see your future review comparison to 45mm at f12.8 to reveal that. Treat this 35mm's f2 aperture as a bonus compare to 45mm. Having said that, expect 35mm perform as good as 45 is also a little unfair. I see so many decent rendering 50mm glasses but so few for 35mm so far.
You are very right about VC 35 and 40 f1.2 in this case. Actually, 3-5M, harsh lighting, close background (with lots of detail) to subject ratio is a recipe fail most "fast" glasses, especially for wide to standard focal. Longer lens will tame it better, shorter lens deeper DOF will cover it, though for shorter lens, problem zone will shift to closer range.
I personal think most lenses will fail in this case include RX1 actually Do you have similar images like these from RX1 before? 58G maybe can survive, 45mm another maybe but 75lux and 80lux will fail even not as bad due to longer focal, the problem zone usually shift to different distance.
Don't get me wrong, I don't think this 35mm have super smooth bokeh but even neutral bokeh is a luxury for most 35mm I have seen so far. Don't expect it can perform under all cases. As for CA under this condition, this is not bad TBH. (and I know RX1 have worse CA performance than this for sure.)
It is all about right balance of individual performance spec including price and size. ...Show more →
Great summary. I do think the Sigma 35/2 does better than the competition when it comes to axial CA correction, including the RX1. Whenever I have it in my hands, I will be able to understand a little more its compromises and priorities.
The transition zone nervousness is what I thought was revealing but this definitely needs more testing. I also don't know how processed this image is. If the raw is available I will try downloading it.
Because of the distance and background complexity (The subject was at 3.2m), it's the first sample image I came across that shows the focus transition with no blur to hide behind.
Actually all samples from this photographer (Carey Rose) are torture tests which is great when evaluating a lens. Thanks Carey!
I am far less bother by the CA than the transition zone. I really only associate super clean CA correction with monster lenses in this FL.
Maybe Sigma chose this very balanced lens at this common FL to appease the common desire (or obsession) with sharpness and it will test well, garnering good press. And maybe, just maybe, Sigma will put out a 28mm (f2 or 2.8) with the SA and rendering emphasis of the 45mm. Just a thought. A 28/45mm duo with similar rendering pairs well.
I could see Sigma putting out a slightly longer rendering based tele as well, given they now have the highly corrected 65mm, 85 1.4 DG DN and 105 macro at this point.
nehemiahphoto wrote:
I am far less bother by the CA than the transition zone. I really only associate super clean CA correction with monster lenses in this FL.
Maybe Sigma chose this very balanced lens at this common FL to appease the common desire (or obsession) with sharpness and it will test well, garnering good press. And maybe, just maybe, Sigma will put out a 28mm (f2 or 2.8) with the SA and rendering emphasis of the 45mm. Just a thought. A 28/45mm duo with similar rendering pairs well.
I could see Sigma putting out a slightly longer rendering based tele as well, given they now have the highly corrected 65mm, 85 1.4 DG DN and 105 macro at this point.
I think this is the kind of lens they wanted to make with the 45; their own story on their Sein page basically goes like this: management gave them size limits. The engineers tried to squeeze slow but art like quality into the aperture and size limits laid down by marketing. They couldn't manage, in part because no room for floating elements. But some people looked at the results and said "hey, all that SA looks kinda cool" so they went with it.
Freed from the limitations of making the "pancake" of the series, they revert to what they usually do. Which is pretty damn good - as a complete package I think I might prefer this to a 45 style lens at 35mm. But I totally understand someone who wanted something more like a faster 45 (but I'm guessing the quality in other respects at 35 would be much worse at f2.0 if it had a design like the 45 but faster; the 45s design worked partly because it was a bit slower)
nehemiahphoto wrote:
And maybe, just maybe, Sigma will put out a 28mm (f2 or 2.8) with the SA and rendering emphasis of the 45mm. Just a thought. A 28/45mm duo with similar rendering pairs well.
I could see Sigma putting out a slightly longer rendering based tele as well, given they now have the highly corrected 65mm, 85 1.4 DG DN and 105 macro at this point.
I only recently got the 45/2.8 and it immediately became a favourite. You’ve described what I’d very much like to see: a 28mm with the SA and rendering emphasis of the 45mm followed by a slightly longer tele (say 90mm) with the same rendering. A kind of reimagining of the 28/45/90 trio that worked so well on the Contax G.
DavidBM wrote:
but I'm guessing the quality in other respects at 35 would be much worse at f2.0 if it had a design like the 45 but faster; the 45s design worked partly because it was a bit slower)
I think you can get something not too dissimilar from the Sigma 45 even at faster aperture while retaining decent resolution. This is what the Nikon 85 Z achieves. But to do so you really need to eliminate nearly all on axis aberrations with the exception of SA and the PSF will be a little different, with a gaussian or so illumination in the centre gradually giving way to a flat disk with sharp edges (but zero "over" ringing) instead of a fully gaussian illumination with blurred edges like the Sigma. If the contrast between the detail you're looking at and its surrounding isn't too high you won't ever see the sharp edge and it will produce similar results as the Sigma.
This is the sort of balance that Nikon has been trying to achieve for more than a decade and I think that at least at longer distances they've totally nailed it with the 85 Z. It and the Sigma are to me both the most desirable lenses I've seen in a while. Too bad they aren't in the same system .
Dec 04, 2020 at 07:55 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
nehemiahphoto wrote:
I just looked quite closely, but even in the areas that aren't fringing, seems like some extremely minor doubling or general nervousness, which make me think it's the structure of the transition zone.
Honestly, I don't think the ZA 35 1.4 would do better here for LoCA, or most other well performing 35mm's. I would be quite surprised if the RX1, Pentax 31 or CV 35/1.7 or 35/1.2 did better. The Samy 35 1.8 and FE 35 would be way worse.
I saw the same thing with the minor doubling and general nervousness. This is just one sample, but I think what we are going to see when there isn't a lot of axial CA provoked as in this image is a transition zone that isn't terrible but isn't great either. I think it won't be as bad as the transition zone in the FE 35 f/1.8 or the Samyang 35 f/1.8, but it won't be as good as the transition zone in the Sigma 45 f/2.8. But again this is just one sample shot and we will know more as people here actually get to use the lens.
It seems to me, however, that what Sigma did with the 35 f/2 is different from what they did with the Sigma 45 f/2.8. I agree with DavidBM that this seems to be a lens about balance. They kept it sharp but it won't be in the very top on sharpness. The kept CA reduced but didn't totally eliminate it. The aimed for good bokeh, but they did not eliminate all flaws (there will be onion rings sometimes, there will be some busyness in the transition zone). They made it small, but not quite as small as some of the other options. I think it is a sound approach, but if you emphasize any parameter you will see its imperfections as well. I think the 65 f/2 take a similar approach and is probably a bit more successful because it is an easier focal length to reduce flaws.
I think the 45 f/2.8 took a different approach. Obviously they kept the size small and the aperture small and sacrificed wide open sharpness (especially at close distances) and emphasized excellent bokeh, while keeping CA quite low but not eliminating it. I personally like this approach a lot, but I think it is clearly a different approach than they took with the 35.
I still am very impressed by this new I series of lenses. I think the design approach of the 35 and 65 will likely be more common with these lenses and they will provide very nice options, but personally I would love to see more lenses with the design approach of the 45 f/2.8. I don't think this 35 and 65 are really that
. The corners look pretty scary though, I wonder if f1.2 would make them look significantly better...
